Master’s Degree Project in Innovation and Industrial Management 
Master’s Degree Project in Innovation and Entrepreneurship 
 
 
THE IMPACT OF THE INTERNET OF THINGS ON 
ESTABLISHED BUSINESS MODELS 
A multiple case study of Swedish insurance companies 
 
 
 
Author: 
Simon Emanuelsson 
 
GRADUATE SCHOOL 
Supervisors: 
Ph. D. Ethan Gifford, Gothenburg University 
Ph. D. Richard Tee, LUISS Guido Carli 
  
   
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
The Impact of the Internet of Things on Established Business Models 
A multiple case study of Swedish insurance companies 
 
By Simon Emanuelsson 
 
© Simon Emanuelsson 
School of Business, Economics and Law, University of Gothenburg 
Vasagatan 1 P.O. Box 600, SE 405 30 Gothenburg, Sweden 
 
All rights Reserved. 
No part of this thesis may be reproduced without the written permission by the authors. 
Contact: simonmoreira96@gmail.com 
  
ABSTRACT 
The insurance industry has long been suffering from profitability and growth issues due to the 
increasing commoditization of insurance solutions, which has forced insurers to engage in 
destructive “premium wars”. Digitalization and technology advancements are only set to 
increase these pressures even further by shifting market boundaries and increasing the level 
of “digital sophistication” expected by customers. Ultimately, these pressures are forcing 
insurance companies to rethink the business models they have employed for so long and look 
beyond “business as usual”. 
One technology that might allow for insurers to stand out in this commoditized market is the 
Internet of Things (henceforth: IoT). This paper aims to study the impact of the IoT on the 
business model of insurance companies. With the help of the business model canvas, this 
paper provides the reader with an in-depth understanding of how IoT applications affect the 
different blocks that constitute the business model of insurance companies. Moreover, this 
paper aims to provide the reader with an understanding of the challenges and risks related to 
the implementation of the IoT, and how these might be mitigated. In order to do this, a 
multiple case design was chosen where semi-structured interviews were held with four 
different Swedish insurance companies currently working with different IoT-centred 
solutions. Additionally, the paper follows a qualitative research strategy with an abductive 
approach to theory generation. 
According to the research, the IoT provides insurers with the opportunity to collect and 
analyse real-time data on insured objects, allowing them to shift towards more dynamic and 
accurate risk and pricing models that are based on usage, rather than the static indicators 
conventionally used. Moreover, by detecting damages in their earlier stages or before they 
even happen, IoT sensors have the potential to reduce the severity and frequency of claims, 
which stand for 69 % of insurers’ expenses. Lastly, insurers are able to provide customers 
with personalized recommendations on how to take care of their insured objects, thus 
increasing the frequency with which they interact with their customers, which in its turn 
improves customer loyalty and retention. 
In order to successfully implement the IoT, insurers will need to address external and 
organisational resistance, as well as disruption threats from competitors. To address external 
resistance, insurers must increase customer willingness to share data through such incentives 
as premium discounts. Moreover, with the insurance industry being recognized for its 
conservatism, insurers will also need to make use of change management and enterprise-wide 
training to create a corporate culture more accepting of change and business model 
innovation. Lastly, the IoT also creates opportunities for new players to disrupt the insurance 
industry. For this reason, first-movers will have to develop data-driven capabilities that are 
difficult to replicate in order to create a sustainable competitive advantage. 
Keywords: Insurance, Internet of Things, IoT, Business Model, Business Model Canvas, 
Business Model Innovation. 
  
ACKNOWLEDGEMENTS 
I would like to thank my supervisors Ethan Gifford and Richard Tee, for their feedback and 
insight on the structure and content of this research. I am also grateful for the inspiration and 
guidance offered by the supervisors at First to Know, who also helped me to get in contact 
with relevant interviewees. Additionally, I would like to express my appreciation for the 
interview participants for giving me their time and energy to assist me with my research. 
Lastly, I would like to thank my colleagues in the Master’s in Innovation and Industrial 
Management programme for their thoughts and constructive critique throughout this process. 
 
 
Gothenburg, June 2019 
  
TABLE OF CONTENTS 
1. Introduction ........................................................................................................................ 1 
1.1. Background ................................................................................................................. 1 
1.2. Research Problem ........................................................................................................ 1 
1.3. Research Purpose ........................................................................................................ 2 
1.4. Research Questions ..................................................................................................... 3 
1.5. Delimitations ............................................................................................................... 3 
1.6. Disposition .................................................................................................................. 4 
2. Theoretical Framework ...................................................................................................... 5 
2.1. The Nature of Insurance .............................................................................................. 5 
2.2. The IoT and the Insurance Industry ............................................................................ 6 
2.3. Business Model Innovation and the Business Model Canvas ..................................... 8 
2.4. The Impact of the IoT on the Business Model of Insurance Companies .................. 11 
2.4.1. Financial Aspects ............................................................................................... 11 
2.4.2. Customer Interface ............................................................................................. 13 
2.4.3. Value Proposition............................................................................................... 15 
2.4.4. Infrastructure Management ................................................................................ 15 
2.5. Managing the Risks with the Implementation of the IoT.......................................... 17 
2.5.1. External Resistance ............................................................................................ 17 
2.5.2. Organisational Resistance .................................................................................. 17 
2.5.3. Disruption Threats ............................................................................................. 18 
2.6. Summary of Theoretical Findings ............................................................................. 18 
3. Methodology .................................................................................................................... 21 
3.1. Research Strategy ...................................................................................................... 21 
3.2. Research Design ........................................................................................................ 21 
3.3. Research Method ....................................................................................................... 22 
3.3.1. Secondary Data Collection ................................................................................ 22 
3.3.2. Primary Data Collection .................................................................................... 23 
3.4. Data Analysis ............................................................................................................ 27 
3.5. Research Quality ....................................................................................................... 28 
4. Empirical Findings ........................................................................................................... 30 
4.1. Cases.......................................................................................................................... 30 
4.1.1. Moderna Försäkringar ........................................................................................ 30 
4.1.2. Paydrive ............................................................................................................. 30 
4.1.3. Folksam .............................................................................................................. 31 
4.1.4. Länsförsäkringar ................................................................................................ 31 
4.1.5. IoT Expert – Talkpool ........................................................................................ 32 
4.2. Impact of the IoT on the Business Model of Insurance Companies ......................... 33 
4.2.1. Financial Aspects ............................................................................................... 33 
4.2.2. Customer Interface ............................................................................................. 36 
4.2.3. Value Proposition............................................................................................... 39 
4.2.4. Infrastructure Management ................................................................................ 40 
4.3. Managing the Risks with the Implementation of the IoT.......................................... 43 
4.3.1. External Resistance ............................................................................................ 43 
4.3.2. Organisational resistance ................................................................................... 44 
4.3.3. Disruption threats ............................................................................................... 45 
5. Analysis and Discussion .................................................................................................. 47 
5.1. Impact of the IoT on the Business Model of Insurance Companies ......................... 47 
5.1.1. Financial Aspects ............................................................................................... 48 
5.1.2. Customer Interface ............................................................................................. 52 
5.1.3. Value Proposition............................................................................................... 54 
5.1.4. Infrastructure Management ................................................................................ 55 
5.2. Managing the Risks with the Implementation of the IoT.......................................... 58 
5.2.1. External resistance ............................................................................................. 58 
5.2.2. Organisational resistance ................................................................................... 59 
5.2.3. Disruption threats ............................................................................................... 60 
6. Conclusion ....................................................................................................................... 62 
6.1. Answering the research questions ............................................................................. 62 
6.1.1. Impact of the IoT on the Business Model of Insurance Companies .................. 62 
6.1.2. Managing the Risks with the Implementation of the IoT .................................. 65 
6.2. Implications ............................................................................................................... 67 
6.3. Suggestions for Future Research ............................................................................... 68 
7. Bibliography .................................................................................................................... 69 
8. Appendixes ...................................................................................................................... 71 
Appendix A – Contacting Respondents ............................................................................... 71 
Appendix B – Interview Guide ............................................................................................ 71 
Appendix C – Empirical Findings ....................................................................................... 73 
 
TABLE OF FIGURES 
Figure 1: Disposition of report ................................................................................................... 4 
Figure 2: The Business Model Canvas (Osterwalder & Pigneur, 2010, page 44) ................... 10 
Figure 3: BMC summary of theoretical findings ..................................................................... 19 
Figure 4: Interview Respondents ............................................................................................. 27 
Figure 5: BMC summary of empirical findings ....................................................................... 33 
Figure 6: Comparison of theoretical and empirical findings ................................................... 48 
 
1 
 
1. Introduction 
This section presents the reader with the problem background and empiric setting that led to 
the research questions identified in this thesis. Moreover, the research purpose, delimitation 
and disposition of the thesis are presented.  
1.1. Background 
The world is amid a fourth industrial revolution that is set to alter the way humans interact 
with one another and their environments. At the centre of this revolution is the Internet of 
Things (henceforth: IoT), a network of devices and sensors that collect data from the physical 
world and transmit it in real-time, with the aim of creating more attractive and safe 
environments, products and services (Talkpool, n.d.). 
With more than seven billion IoT devices installed worldwide, there are currently more 
objects connected to the internet than there are humans on Earth. This number does not take 
into account smartphones, tablets, computers or other conventional computing platforms. 
Instead, the IoT allows for traditional non-internet-enabled objects and machines to be 
connected to the internet (Lueth, 2018).  
Even though this subject has become of increasing interest to both academics and industry 
leaders, the phenomenon of the IoT is nothing new. The first known IoT application was a 
Coca-Cola vending machine at the Carnegie Mellon University in Pennsylvania, altered by a 
few engineering students in the early 1980’s to monitor the availability and temperature of 
the beverages in real-time (Teicher, 2018).  
So, what has triggered this increased interest as of lately? Lower sensor costs, increased 
processing power, advanced connectivity and the miniaturization of components, among 
others (Ammiot, 2015). Many industry leaders, especially within manufacturing, have 
understood the potential of the IoT and are already experiencing the increased quality, cost 
efficiency and speed that it brings to their supply chains. Indeed, with an average growth rate 
of 17 % per year, the number of installed IoT devices worldwide is expected to reach 21.5 
billion by 2025 (Lueth, 2018).  
1.2. Research Problem 
The topic of this paper is an industry generally recognized for its conservatism and static 
landscape, the insurance industry. Specifically, this paper is concerned with addressing the 
2 
 
impact of the IoT on insurance companies’ business model (henceforth: BM), as the 
increasing adoption of such technologies are set to fundamentally alter customer 
expectations, and therefore the insurance landscape (Reifel et al., 2014). 
The insurance industry has long been suffering from low organic growth and profitability 
issues. Increased competition and a lack of differentiation have turned insurance into a 
commoditized product, forcing players to engage in “premium wars” (Canaan et al., 2016; 
Reifel et al., 2014). In fact, during the last ten years, there has been a 10 percent decline in 
inflation-adjusted car insurance premiums. This number is expected to decrease even further 
with the development of new technologies such as Car Sharing Services and ADAS 
(Advanced Driver Assisted Systems), thereby reducing the number of insurable objects and 
traffic collisions, respectively (Reifel et al., 2014). 
The rate of digitalization is not slowing down and will only continue to increase industry 
pressures even further as entry barriers for outside players are reduced, creating a shift in the 
market boundaries (Ammiot, 2015; Reifel et al., 2014). Whereas large amounts of risk-related 
data used to be a big entry barrier in the past, companies like GAFA (Google, Amazon, 
Facebook, Apple) have now the capacity to gather relevant data at a faster rate and create on-
demand value propositions (Ammiot, 2015). As an example, Tesla has recently decided to 
disintermediate insurers by providing their own car insurance for customers that are 
unsatisfied with the prices provided by insurers (Lambert, 2018). 
These pressures are forcing the insurance industry to rethink the business models they have 
employed for so long. By leveraging the IoT, first-movers could turn digitalization from a 
threat to an opportunity. For example, sensors embedded into a buildings’ infrastructure have 
the potential to detect safety breaches such as smoke, toxic fumes, moisture or mould and 
alert the user before the fire or water damage escalates. This not only allows for insurance 
companies to adopt an offensive position, shifting their focus from restitution to actual 
prevention of avoidable damages, but improves upon their value proposition by reducing the 
risk that the policyholder1 loses an irreplaceable object or incurs costly deductibles (Reifel et 
al., 2014; Ammiot, 2015; EY, 2016). 
1.3. Research Purpose 
In short, insurers cannot afford to be left behind and need to look beyond “business as usual”. 
The rate of change is not slowing down, and the connected customer expects an increasing 
                                                 
1 A person who holds an insurance policy. Also referred to as customer or insured throughout the paper. 
3 
 
level of “digital sophistication” from the businesses they choose to engage in. The IoT 
provides insurers with the opportunity to address the increasing industry pressures by shifting 
their attention from restitution to the actual prevention of avoidable damages. Among others, 
the IoT will allow for first movers to differentiate themselves from a commoditized market 
and improve on customer relationships (Canaan et al., 2016; Reifel et al. 2014). 
This paper aims to study the impact of the Internet of Things on the Business Model of 
insurance companies, providing the reader with an in-depth understanding of how this 
technology affects the different building blocks that constitute the BM. Moreover, Business 
Model Innovation, such as the implementation of the IoT, tends to bring forth challenges and 
risks, especially for such a conservative industry such as insurance. For this reason, this paper 
also intends to study the risks and challenges related to the implementation of the IoT, and 
the potential solutions to eliminate them, with the help of Business Model Innovation theory.  
Additionally, most academic literature is focused on the technological and technical aspects 
of the IoT. Whilst managerial research about the impact of IoT on established BMs is 
increasing, several authors (Dijkman et al, 2015. Metallo et al., 2018) have mentioned that 
understanding the mechanisms employed to create and capture value are still of critical 
importance. Among these managerial research papers, most of them focus on the 
manufacturing industry, also known as the Industrial Internet of Things. For this reason, this 
paper contributes to academic literature with a novel perspective on the impact of the IoT on 
the BM of a conservative service-centred industry such as insurance. 
1.4. Research Questions 
The following research questions summarize the specific goals of this study: 
 How will the implementation of the IoT affect the Business Model of Insurance 
Companies? 
 How can insurance companies manage the challenges with the implementation of the 
IoT? 
1.5. Delimitations 
This paper seeks to provide the reader with an understanding of the impact of the IoT on the 
business models of insurance companies, from a managerial standpoint. For this reason, the 
technological or technical aspects of the implementation of the IoT are disregarded.  
4 
 
Interviews were held exclusively with players in the Swedish insurance industry that have 
either experimented or implemented IoT in parts of their supply chain. Additional insights 
were also gathered from an IoT provider, to add an external perspective from a key partner in 
the development of IoT-centred insurance solutions. 
1.6. Disposition 
The thesis is divided into six sections that follow the following structure: 
 
Figure 1: Disposition of report (own elaboration) 
  
Introduction 
•Problem Background 
•Research Purpose & Questions 
Theoretical 
Framework 
•The Nature of Insurance 
•Business Model Innovation 
•The Impact of the IoT 
•Managing the Risks with the IoT 
 Methodology 
•Research Strategy & Design 
•Data Collection & Analysis 
•Research Quality 
Empirical 
Findings 
•Findings from Interviews 
Analysis 
•Comparison of theoretical and empirical findings 
Conclusion 
•Answering the Research Questions 
•Implications 
•Suggestions for Future Research 
5 
 
2. Theoretical Framework 
In this section, the theory and frameworks used later in the analysis are explained and 
discussed. This section aims to give the reader an overview of the previous managerial 
literature on the applications of the IoT on the insurance industry and its implications. 
Moreover, the author explains the choice of this theoretical framework by explaining how the 
Business Model Canvas and Business Model Innovation theory will be used along the thesis. 
2.1. The Nature of Insurance 
An insurance policy is a contract where the insurance company guarantees to cover the cost 
of an object or subject in the event of a loss, up to a certain amount. Hence, in exchange for a 
premium payment to insurance companies, policyholders receive financial security and 
stability in the case of a loss. These premiums are calculated through actuarial methods of 
analysing and quantifying risks. Moreover, insurance consists of pooling funds from a large 
number of policyholders to pay for the damages suffered by some of them. This is because it 
is far more accurate and less challenging to estimate risk for a group of people, than it is for a 
single individual (Desyllas & Sako, 2013. Derikx, De Reuver, & Kroesen, 2016). 
Insurers’ profits come traditionally from two sources: underwriting, which is the difference 
between the premiums paid by policyholders and the payments made to them for incurred 
losses and the underwriting expenses; and by investing the premiums into common equities 
and fixed-income securities (Desyllas & Sako, 2013. Derikx et al., 2016). To help illustrate, 
Desyllas and Sako (2013) explain that for every 100 USD an insurance company receives in 
premiums, 69 USD are paid out in claims and 25 USD are spent on underwriting expenses. 
So, on average, insurance companies profit 5 USD from every 100 USD paid by 
policyholders. 
According to Desyllas and Sako (2013), there are three major ways in which insurers can 
improve performance, with the first and most obvious one being claim cost reductions. These 
costs stand for about 69 % of premiums and are defined by loss severity and frequency. Cost 
savings in this area could be achieved by improving underwriting capability and the 
subsequent analysis and pricing of risks. However, this solution is limited by the problems of 
moral hazard and adverse selection. 
Moral hazard and adverse selection are both principal-agent problems that arise due to the 
inherent asymmetry of information in insurance, whereby the insured has more knowledge 
than the insurer. Moral hazard arises when the insured provides misleading information to the 
6 
 
insurer, such as in the case of insurance fraud. Adverse selection, on the other hand, relates to 
the issue where the policyholder has more accurate information, resulting in ineffective price 
signals. A good example of the latter is when the insurance company sets a relatively high 
insurance premium to make up for the losses of risk-prone customers. This in its turn drives 
away the risk-averse customers whilst attracting risk-prone customers to whom the premium 
seems reasonable (Nickolas, 2019). At the same time, policyholders are more likely to be less 
concerned for losses than they otherwise would have been without an insurance policy. In 
order to counter this information asymmetry, insurance companies make use of different 
historical and socio-demographic factors in their risk and premium calculations. In car 
insurance, for example, insurers look at prior claim experience, demographic characteristics 
such as age and sex, and car model characteristics such as safety features and crashworthiness 
(Desyllas & Sako, 2013). 
The second way to improve performance includes increasing the efficiency of underwriting 
process through better policy handling, automation, lower commissions and economies of 
scale. The third way is through the improvement of return on investment, but this is highly 
dependent on stock market cycles (Desyllas & Sako, 2013. Derikx et al., 2016). As this paper 
will show, the IoT has the potential to reduce claim costs, as well as improve the efficiency of 
the underwriting process. 
2.2. The IoT and the Insurance Industry 
As previously discussed, digitalization is only set to further increase the pressures the 
insurance industry is currently experiencing. In short, the IoT allows insurers to collect and 
analyse real-time data in a dynamic environment, thereby improving risk selection, pricing 
and monitoring models (EY, 2016; Canaan et al., 2016; Ammiot, 2015).  
As a result from increasing competition and pressures to reduce costs for both policyholders 
and the insurer, some companies are currently offering Usage-based Insurance (henceforth: 
UBI) solutions (EY, 2016. Baecke & Bocca, 2017. Canaan et al., 2016; Ammiot, 2015). 
Examples of UBI include pay-as-you-drive (henceforth: PAYD) or pay-as-you-live models, 
which collect real-time customer behaviour data from IoT sensors embedded in cars or 
watches. According to an estimation by EY (2016), there are currently 5 million active UBI 
policies in 35 different countries, with a forecast of a 15 % market penetration by 2020 in 
Europe, Asia and Americas. 
7 
 
In their reports, Baecke and Bocca (2017) and Derikx et al. (2016) study PAYD solutions, 
where the policyholder attaches a sensor to their car’s On-board Diagnostic (henceforth: 
OBD) input that collects driving behaviour data such as speed, location, mileage and 
acceleration. This allows insurers to provide policyholders with more accurate and 
personalized insurance premiums that rely on dynamic usage-based data, rather than solely 
looking at historical and socio-demographical factors. For insurers, this allows for a reduction 
in incurred losses through more accurate risk estimations. Moreover, these services usually 
allow for extra integrated services such as automatic emergency calls, self-diagnosis, stolen 
vehicle monitoring and driving suggestions that increase safety and reduce fuel consumption 
(Baecke & Bocca, 2017. Derikx et al., 2016). 
Besides collecting data on customer behaviour, sensors attached to machinery and in-home 
sensors can detect early signals of fire, wind or water damage, and notify the user at the 
slightest hint of trouble. This allows insurers to prevent avoidable damages or at least reduce 
their impact, whilst simultaneously protecting the customer. In short, the IoT allows insurers 
to move from restitution to the actual prevention of avoidable damages (Canaan et al., 2016. 
Reifel et al., 2014; Ammiot, 2015; EY, 2016). 
Moreover, a better understanding of individual customer behaviour allows insurers to move 
from generalized offerings to increasingly personalized ones. This move from mass to 
segmented markets can be seen with the adoption of UBI models. The IoT provides insurance 
companies with the ability to focus their attention on the most profitable insurance pools, 
encouraging them to renew policies and reduce their exposure to risk. Better yet, it allows for 
first-movers to create more attractive and personalized value propositions and leave less 
profitable customer groups for competitors (Ammiot, 2015; EY, 2016; Reifel et al., 2014). 
Still, Ammiot (2015) recommends that insurers use caution with this kind of segmentation on 
risk-prone customers, so that they can still find an appropriate insurance solution. 
Another area in which the IoT has potential, is customer relationship management. Currently, 
customer interaction takes place mostly during claim management, which in certain lines of 
business can take an average of four to five years. The IoT offers the opportunity for insurers 
to not only make this process more pleasant and speedy, by having already collected data 
prior and during an incident, but to interact with customers more often by offering them real-
time information on their insured objects.  
8 
 
Additionally, industry experts expect a shift from simple risk insurance to advisory services, 
creating a new offering among traditional insurance solutions. This could potentially include 
recommendations on the advantages of proper maintenance and care, thus improving 
customer interaction, whilst influencing customer behaviour to prevent avoidable damages. 
Alternatively, insurance companies could use their data to assist customers in purchase 
decisions and advise them on how to best retain the value of their insured object, such as a 
house or car (Ammiot, 2015; Reifel et al., 2014). 
In summary, the real-time data collected from IoT sensors provides insurers with a better 
understanding of risk and customer behaviour, reducing the indirect costs incurred from 
information asymmetry through better fraud detection, segmentation and pricing models 
(Ammiot, 2015; Reifel et al., 2014; Canaan et al., 2016). Moreover, with a better 
understanding of customer behaviour, customers might be more accepting of personalized 
advertisements and relevant offerings. This allows for insurers to provide policyholders with 
tailored and personalized up-sales (Baecke & Bocca, 2017. Derikx et al., 2016). 
This transition will most likely squeeze margins in the short run, in exchange for a long-term 
sustainable advantage. These short-term costs will come from external promotions to increase 
customer’s willingness to be monitored by IoT sensors, and internal attempts at creating a 
corporate culture within insurance companies that is more accepting of change and 
innovation (Ammiot, 2015; EY, 2016; Reifel et al., 2014). Another issue that insurers will 
have to deal with is finding the right partners and external sources able to provide and 
maintain high-value data, as security and privacy are becoming bigger concerns in an 
increasingly digital world (Ammiot, 2015; Reifel et al., 2014; EY, 2016). 
Lastly, even though early adopters might get a good head-start, imitators will naturally follow 
route. First-movers are recommended to develop data-driven capabilities that are difficult to 
replicate, allowing for a shift from generalized product offerings to improved segmentation 
and personalization (Ammiot, 2015; EY, 2016). 
2.3. Business Model Innovation and the Business Model Canvas 
 “A mediocre technology pursued within a great business model may be more valuable than a 
great technology exploited via a mediocre business model.” – Chesbrough, 2010. 
A Business Model can be defined as a “consistent and integrated picture of a company and 
the way it generates revenues and profit” (Spieth, Schneckenberg & Ricart, 2014). According 
9 
 
to Chesbrough (2010), a Business Model must be able to explain the main value proposition, 
identify the market segment, the revenue generation mechanism, the structure of the value 
chain, the cost structure and profit potential. 
Spieth, Schneckenberg and Ricart (2014) find three major motivations for engaging in 
business model research: explaining, running and developing the business. When explaining 
the business, the target audience are the external stakeholders, and the aim is to explain how a 
certain company is able to generate profit. When running the business, the target audience are 
mostly the employees and managers in the company, and the aim is to understand the 
processes and structures that need to be performed for the day-to-day operations. When 
developing the business, the target audience is the strategic function, and the aim is to support 
management in the definition and development of a company’s strategy.  
In this case, this paper aims to discuss how insurance companies can develop their business 
and strategy with the help of the IoT. In other words, how insurers can innovate their business 
models to enhance their customer experience and insurance protection. In order to study the 
implications of the implementation of the IoT for insurance companies, the author chose to 
make use of business model innovation theory, which studies the creation and transformation 
of new and established business models, and can be defined as “the discovery of a 
fundamentally different business model in an existing business” (Markides, 2006, p. 20).  
This theory was chosen since it allows to look at the common pitfalls and challenges 
companies face when innovating their business model, and how insurance companies can 
address them in order to successfully implement the IoT into their business models. 
The most popular tool for Business Model Innovation and research is the Business Model 
Canvas (henceforth: BMC) by Osterwalder and Pigneur (2010). Business modelling and 
mapping tools, such as the BMC, allow companies to visualize and experiment with 
alternative business models before fully committing to them (Chesbrough, 2010). This will be 
the framework used to aid in the construction and visualization of the impact of the IoT on 
Insurance Companies’ BM.  
The BMC was chosen since it provides the author with an established structural framework 
on which to build the theoretical framework, data collection and analysis. Additionally, this 
framework allows for the study of a single IoT application through multiple lenses, thus 
providing the author and reader with a better understanding of the impact of the IoT on the 
different blocks that constitute a business model. The empirical findings and analysis follow 
10 
 
the same structure in order to facilitate the structured comparison of theoretical and empirical 
findings. 
In their framework, Osterwalder and Pigneur (2010) identify nine critical elements to the 
business model, called the nine building blocks of the Business Model Canvas, that allow 
managers to understand the critical factors in creating, delivering and capturing value. 
 
Figure 2: The Business Model Canvas (Osterwalder & Pigneur, 2010, p. 44) 
The customer segments, channels and customer relationships constitute the Customer 
Interface and identify who the company is selling to, how they are delivering the product to 
them and how they build strong relationships with these customers. The key partners, 
activities and resources, constitute the Infrastructure Management, and identify how the 
network/supply chain should look like, what activities need to be performed, and finally what 
assets are required for the company to produce and deliver the offering to the consumer. The 
Financial Aspects identify how value is captured (revenue streams) and how much it will 
cost the company to provide these products (cost structure). Finally, the Value Proposition 
defines what product or service the company is offering, and how it solves the problems of 
the customer in mind (Osterwalder & Pigneur, 2010). 
  
11 
 
2.4. The Impact of the IoT on the Business Model of Insurance Companies 
“The IoT is here to stay, the rate of change is unlikely to slow anytime soon, and the 
conservative insurance industry is hardly impervious to connectivity-fuelled disruption—both 
positive and negative. The bottom line: Insurers need to look beyond business as usual. In the 
long term, no company can afford to engage in premium price wars over commoditized 
products. A business model informed by IoT applications might emphasize differentiating 
offerings, strengthening customer bonds, energizing the industry brand, and curtailing risk 
either at or prior to its initiation.” – Canaan et al., 2016. 
The quote above properly summarizes the path the insurance industry is heading for if it 
continues looking at “business as usual”. In this section, the author will discuss the different 
ways in which insurance companies are able to leverage the IoT to improve on their business, 
by studying the potential future impact of the IoT applications mentioned in 2.2. on the 
different blocks of the BMC. 
2.4.1. Financial Aspects 
2.4.1.1. Revenue Streams 
A company can capture value through different revenue streams. A revenue stream can be the 
result of one-time customer payments. such as asset sales, or recurring revenues from 
ongoing payments, such as usage or subscription fees, lending/renting/leasing, licensing, 
brokerage fees or advertising (Osterwalder & Pigneur, 2010). 
Each revenue stream might have different pricing mechanisms: fixed menu or dynamic 
pricing, where the former is predefined and based on static variables and the latter changes 
based on market conditions. Examples of fixed menu pricing include list price, or can be 
dependent on product features, customer segment or volume. Examples of dynamic pricing 
include negotiation/bargaining, yield management (e.g. airplane tickets and hotel rooms), 
real-time market and auctions (Osterwalder & Pigneur, 2010). 
The main revenue stream of insurance companies are the premium payments they receive 
from policyholders in exchange for insurance protection. Conventionally, these premiums are 
calculated through actuarial analysis of risk, by looking at historical and socio-demographic 
indicators related to the policyholder (Desyllas & Sako, 2013). The IoT, on the other hand, 
allows insurers to receive real-time data on insured objects and customer behaviour, thus 
improving the accuracy of their risk and pricing models. This can be seen through the 
development of UBI solutions, where policyholders pay their premiums according to their 
12 
 
individual usage (EY, 2016; Canaan et al., 2016; Ammiot, 2015). This, in its turn, allows for 
a shift from fixed menu subscription fees towards more dynamic usage-based pricing 
mechanisms. 
Moreover, this increased understanding of customer behaviour allows for insurers to provide 
policyholders with targeted recommendations and up-sales, which affects revenue streams 
positively. On the other side, insurers will also have to provide benefits for policyholders to 
increase their willingness to share data. Such benefits include premium discounts, value-
added services, loyalty points, among others (Ammiot, 2015; EY, 2016; Reifel et al., 2014). 
These incentives, in their turn, constitute a revenue stream reduction for insurers. 
Lastly, the IoT allows insurers to provide policyholders with advisory and other extra 
integrated services (Ammiot, 2015; Reifel et al., 2014). Depending on how the offering is 
packaged, these might constitute new revenue streams or be included into the offering as a 
way to increase customer willingness to share their data. 
2.4.1.2. Cost Structure 
Virtually all processes and activities a company performs bring about costs, which is why 
companies need to bear this in mind when designing a viable BM. Companies can choose to 
position themselves as cost- or value driven, where the former one focuses on minimizing 
costs as much as possible through maximum automation and outsourcing, whilst the latter 
focuses on value creation through personalized services and customization (Osterwalder & 
Pigneur, 2010). 
The IoT is able to decrease claim frequency and severity in several ways, thus reducing the 
biggest cost for insurance companies. Firstly, by detecting hazards in their early stages, the 
IoT has the potential to prevent losses from escalating or from happening in the first place. 
Secondly, UBI solutions such as PAYD provide lower premiums for policyholders that drive 
in a responsible way, thus influencing customer behaviour in a way that reduces risk for both 
parties. The same is seen with the addition of advisory services, that give the policyholder 
recommendations and tips on how to reduce their risk for losses. Lastly, by moving towards 
more profitable and risk-averse customer segments, insurers are also able to reduce claim 
frequency. Ultimately, the IoT allows insurers to move from the restitution of damages 
towards an increased prevention of avoidable damages (Reifel et al., 2014; Ammiot, 2015; 
EY, 2016). 
13 
 
Moreover, with little differentiation between insurers’ offerings, players in the insurance 
industry are currently pressured to engage in “premium wars”. By helping policyholders 
avoid damages and moving towards more individualized and frequent customer interactions, 
the IoT allows insurance companies to move from a cost-driven BM to one that is more 
value-driven. 
2.4.2. Customer Interface 
2.4.2.1. Customer Relationships  
Companies should attempt to build strong customer relationships to increase sales through 
customer acquisition and retention. Companies can opt to do this personally or with the 
assistance of automation, ranging from dedicated personal service to automated self-service 
(Osterwalder & Pigneur, 2010). 
Personal assistance is based on human interaction, where the customer gets help from a 
representative during their purchase or post-purchase experience. Dedicated personal 
assistance consists of assigning one representative to an individual customer during a longer 
period of time. On the other hand, there is self-service, that provides the customer with the 
tools necessary to help themselves, and automated service, such as chat bots and 
recommendations. Lastly, companies can choose to create or encourage communities like 
forums for customers to interact and help one another, or incentivize co-creation 
(Osterwalder & Pigneur, 2010). 
Currently, insurers rarely interact with their customers. By providing policyholders with real-
time information on their insured objects and recommendations on how to reduce their risk 
for losses, the IoT provides insurers with the opportunity for more frequent customer 
interactions. Moreover, the IoT can potentially improve the customer experience throughout 
the claims process by having collected data prior or during an incident (Ammiot, 2015; Reifel 
et al., 2014). 
This improved customer interaction, together with the addition of advisory services, allows 
insurance companies to offer an extra level of personalization to their customers, thus 
facilitating customer acquisition and retention. This indicates a move from rare occasions of 
personal assistance to more frequent and individualized automated assistance.  
2.4.2.2. Customer Segments 
As Osterwalder and Pigneur (2010) mention, “without profitable customers, no company can 
survive for long”. Companies need to make an active decision on which customer segments 
14 
 
to focus on and understand how their customers provide them with a solution to a want or 
need.  
Customer segments might be mass or niche market, segmented or diversified, or multi-sided 
markets. Companies focused on mass markets provide for one large group with very similar 
wants and needs, without major segmentation, whilst niche markets focus on specific 
customers, often found in supplier-buyer kind of relationships. Segmented business models 
provide slightly different offerings to different customer segments with slightly different 
wants and needs, whilst diversified business models focus on one or more completely 
unrelated products to significantly different customer segments. Lastly, multi-sided platforms 
serve two or more interdependent customer groups, as the example of credit card companies 
or Google (Osterwalder & Pigneur, 2010). 
By capturing real-time data on customer behaviour, the IoT allows for more accurate risk and 
pricing models, as well as customer segmentation. This allows for insurers to shift from mass 
market to increasingly individualized offerings, whereby the policyholder receives tailored 
premiums and insurance solutions. This can also be seen through the development of UBI 
solutions (Ammiot, 2015; EY, 2016; Reifel et al., 2014). 
The IoT also allows companies to focus on more profitable customer segments, whilst 
leaving less profitable ones to competitors. Derikx et al. (2016) found that policyholders that 
drive longer distances are less likely to choose UBI solutions, since this would mean a higher 
fee due their higher exposure to risk. This means that usage-based insurance indirectly 
attracts policyholders with less exposure to risk, which in its turn reduces claim frequency 
and severity. 
2.4.2.3. Channels 
Channels include communication, marketing, distribution and sales channels, and these serve 
to raise awareness, allow for the customer to buy and receive the offering, and provide 
support and help the customer evaluate a company’s value proposition (Osterwalder & 
Pigneur, 2010). 
In order to implement the IoT, insurers will initially need to rely on partner channels to 
distribute the physical IoT sensors and provide policyholders with relevant information on 
their insured objects, since this is not their core competence (Ammiot, 2016; EY, 2016). 
Partner channels have lower margins but allow for companies to provide their offering 
without spending as much time and resources developing their own channels. Additionally, 
15 
 
partner companies can achieve better economies of scale and scope, thus achieving a cost per 
unit that a smaller company would not be able to reproduce (Osterwalder & Pigneur, 2010).  
2.4.3. Value Proposition 
“A good business model yields value propositions that are compelling to customers, achieves 
advantageous cost and risk structures, and enables significant value capture by the business 
that generates and delivers products and services” – Teece (2010). 
Each value proposition caters to the wants and needs of a specific customer segment, 
describing how the offering creates value for the customer. According to Osterwalder and 
Pigneur (2010), an offering can create value for the customer through different elements: 
newness, performance, customization, “Getting the job done”, price, design, brand/status, 
accessibility, convenience/usability and cost or risk reduction (Osterwalder & Pigneur, 2010). 
The IoT allows insurance companies to improve their value proposition on several 
dimensions. Firstly, by installing sensors onto infrastructures and other non-internet-enabled 
objects, policyholders can get notified whenever a potential danger or hazard is detected. This 
allows for insurers to avoid costly restitutions and policyholders to reduce the risk of losing 
irreplaceable assets, including their own lives. This real-time monitoring also reduces the risk 
of paying costly deductibles, thus producing an indirect cost reduction for policyholders. 
Additionally, with an increased understanding of customer behaviour, insurance companies 
will be able to move from generalized offerings to increasingly personalized premiums and 
solutions, as seen with UBI models. This segmentation allows for the improvement of both 
personalization as well as pricing for certain customer groups. Moreover, by adding new 
offerings such as advisory and other extra non-risk related services, insurers create a value-
added service that increases convenience/usability for policyholders. 
All these aspects will ultimately allow for insurers to improve the value-added of their 
brand/status. By differentiating themselves in a growingly commoditized market, first-
movers will be able to position themselves as innovative and responsive to customer needs 
(Reifel et al., 2014). 
2.4.4. Infrastructure Management 
2.4.4.1. Key Activities 
Key activities describe the actions and processes that a company must perform to create, 
deliver and capture the value from the customers. These can differ significantly depending on 
16 
 
the business model type, focusing on either production, problem solving or platform/network 
(Osterwalder & Pigneur, 2010). 
Insurance companies are highly dependent on the analysis and understanding of risk. By 
providing real-time data on insured objects, insurance companies can improve on their risk 
and pricing models by bypassing the information asymmetry problems and ultimately 
improve the underwriting process (Ammiot, 2015). Moreover, by extending their offerings to 
advisory and non-risk related services, insurance companies are developing new activities in 
their business models.  
2.4.4.2. Key Resources 
The key resources consist of the critical assets for the success of the business model, these 
can be either physical, financial, intellectual or human. Companies may choose to develop 
their own resources, or lease or acquire them from key partners (Osterwalder & Pigneur, 
2010). 
In order to sustain a competitive advantage as a result of the implementation of IoT-centred 
insurance solutions, first movers will need to shift their attention from financial resources 
towards developing data-driven capabilities and intellectual property that are difficult to 
replicate and allow them to stand out from the competition (Ammiot, 2015; EY, 2016). 
2.4.4.3. Key Partners 
Companies can choose to expand their network through partners for different reasons: 
reduction of risk and uncertainty, optimization and economies of scale, or for the acquisition 
of specific resources.  According to Osterwalder and Pigneur (2010), this can be done in four 
ways: strategic alliances (these happen between non-competitors), coopetition (these happen 
between competitors that aim to reach a common goal), joint ventures (two or more 
companies join assets and resources to create a new business) or buyer-supplier relationships 
(contracts that ensure a certain quantity is supplied to the buyer). 
One of the challenges in providing IoT-centred solutions, is finding the right partners to 
collect and maintain high-value data. Since data collection and sensor development is not the 
core competence of insurance companies, they will have to acquire these activities and 
resources from external partners. By building strategic alliances with IoT providers, insurers 
can get access to specific resources, at the same time as they reduce their exposure to 
potential cyber risks (Ammiot, 2015; Reifel et al., 2014; EY, 2016). 
17 
 
2.5. Managing the Risks with the Implementation of the IoT 
Adopting the IoT brings about some challenges. This section describes the challenges and 
risks related to the implementation of the IoT in the insurance industry and is divided into 
three main areas: external resistance, organisational resistance and disruption threats. 
2.5.1. External Resistance 
One major challenge, according to the literature, is increasing customer willingness to share 
data. In their study, Derikx et al. (2016) show that even though customers originally prefer 
conventional car insurance over usage-based insurance, specific privacy concerns about 
usage-based insurance services can be compensated by offering a premium discount. Other 
incentives include value-added services, customer loyalty points, and so on 
Additionally, when creating a BM, companies need to evaluate their value proposition from a 
customer-centric perspective, allowing for the creation of a clearer market segmentation and 
value proposition for each one of these segments (Teece, 2010). Therefore, a deep 
understanding of customers’ needs will be necessary to navigate through the uncertainty that 
is IoT, as not all innovations will be relevant or profitable (EY, 2016; Reifel et al., 2014).  
2.5.2. Organisational Resistance 
According to Chesbrough (2010), BMI by itself brings about a challenge, seen as 
experimentation usually conflicts with the traditional configuration of assets. Mapping tools 
such as the BMC give managers an easier time visualizing how the current and prospective 
BM differ, and how their decisions affect the different blocks that constitute their BM 
(Chesbrough, 2010). 
Moreover, the conservatism the insurance industry is known for will most likely make this 
transition challenging (Canaan et al., 2016). A survey by AT Kearney (Reifel et al., 2014) has 
shown that insurers lag in the market in terms of their ability to optimize long-term value, 
collaborate with customers and utilize new insights for long-term benefits. For this reason, 
insurance companies must start by setting the right governance and corporate culture in place 
to create a mindset that embraces change and innovation. This will require true leadership 
commitment to spearhead the transformation, followed by enterprise-wide training and 
change management initiatives (Canaan et al., 2016; Reifel et al., 2014). 
18 
 
2.5.3. Disruption Threats 
For innovators to be able to capitalize on an invention, they need to put mechanisms in place 
to capture value from each customer segment and hinder competitors from imitating their BM 
(Teece, 2010). As previously mentioned, insurers will need to develop intellectual property 
and data-driven capabilities that are hard to imitate, in order to create a sustainable 
competitive advantage. During this process of change it is still necessary to perform well in 
the current business, at the same time as leaders need to embrace low cost and small-scale 
experimentation in real life situations, using these opportunities to learn ahead of competitors 
(Chesbrough, 2010).  
A final concern is the self-disruption that such digital transformation brings about. A shift 
from restitution to prevention might itself make certain insurance policies obsolete. If the IoT 
sensors embedded in cars and homes effectively reduce the amount of damages or accidents, 
then there would be no further incentive to purchase insurance, due to the lower frequency 
and severity of said losses. This might call for a move from human error to product liability, 
or low-frequency, high-severity events that are harder to price (Canaan et al., 2016). 
2.6. Summary of Theoretical Findings 
In summary, the Internet of Things might allow for insurers to move from simple restitution 
towards the actual prevention of damages. A better understanding of risk and customer 
behaviour allows for the reduction of information asymmetry through improved fraud 
detection, segmentation and pricing models. This in its turn will allow for first-movers to 
move from generalized offerings towards increasingly personalized offerings, focusing on the 
most profitable customer segments, whilst leaving the rest for competitors (Reifel et al., 
2014). 
Additionally, the IoT creates an opportunity for insurers to improve customer relationship 
management by having more frequent advice-led customer interactions that provide 
policyholders with real-time information on their insured objects, including themselves 
(Reifel et al., 2014).  
The table below summarizes the theoretical findings on the potential future impact of the IoT 
on the BM of insurance companies: 
19 
 
 
Figure 3: BMC summary of theoretical findings (own elaboration) 
  
Key Partners Key Activities Customer Relationships Customer Segments
•Strategic alliances 
with IoT-providers
•Improved fraud 
detection
•Shift to more frequent 
interactions
•Shift to more profitable 
segments
•Outsource technical 
product development
•Improved risk and 
pricing models
•Shift to advice-led interactions •Shift to personalized 
offerings
•Move towards non-risk 
related offerings
•Speedier claim management
Key Resources Channels
•Shift to intellectual & 
data-driven capabilities
•Reliance on partner channels
•Shift from subscription to usage-based fees
•Shift from fixed menu to dynamic pricing
•New offerings/revenue streams
•Up-sale opportunities
•Premium discounts
•Claim frequency and severity reduction
Value Proposition
•Risk reduction
•Increased 
personalization
•Increased convenience
•Improved brand/status
Cost Structure Revenue Streams
•Improved pricing 
•Cost reduction
•Shift from cost- to value-driven BM
•Shift to more profitable segments
20 
 
As early adopters continue to experiment with Usage-Based Insurance models, a first-mover 
advantage could be critical, allowing for differentiation in an increasingly commoditized 
market with long-term growth and profitability issues. First-movers will be able to leverage 
the IoT to position themselves as innovative and responsive to customer needs (Reifel et al., 
2014). 
However, the conservatism that characterizes the insurance industry will not make this 
transformation easy. First-movers have to be prepared for short-term losses in favour of a 
long-term sustainable advantage, if they are to keep up with an increasing digitalized world. 
This includes fundamentally changing corporate culture, increasing customer willingness to 
have their data collected and setting barriers that hinder competitor from imitating their BM. 
  
21 
 
3. Methodology 
Below, the author describes the research strategy and design used to conduct the research, as 
well as the way in which the data collection and analysis was performed. Additionally, every 
choice is motivated and thoroughly explained to improve the validity and reliability of the 
research. 
3.1. Research Strategy 
The research question was created as a result of the author’s general interest in understanding 
how the IoT affected the established business model of a conservative service-provider, such 
as the insurance industry. However, due to the author’s limited knowledge on the research 
subject, and the lack of established theories within this area, an abductive approach was 
chosen. The abductive approach aims to make an unexplored area less puzzling by using 
previous theory, although limited, to facilitate in the generation of new theory. So, by 
combining theoretical and empirical findings, this approach provides a more thorough 
analysis of the impact of the IoT on the BM’s of insurance companies. Additionally, the 
ability to change the literature review after the data collection allowed for a more iterative 
process that was beneficial for the paper (Bryman & Bell, 2015). 
The explorative nature of the research questions and the limited knowledge on the subject call 
for a certain level of flexibility in the face of unexpected answers and data. For this reason, a 
qualitative research strategy was chosen. This research strategy is mainly concerned with 
words rather than the quantification in the data collection an analysis (Bryman & Bell, 2015).  
According to Bryman and Bell (2015), a qualitative research strategy is suitable when the 
research questions investigate “how” something is affected by a change, and/or when the 
researcher is concerned with creating an in-depth understanding of a certain context. For that 
same reason, quantitative research strategy was deemed unsuitable for this study, since it is 
usually concerned with the generalization, measurement and testing of previously established 
theory (Bryman & Bell, 2015). The main problems and critiques related to qualitative 
research strategies are found in 3.5. 
3.2. Research Design 
According to Bryman and Bell (2015, p. 40), a research design provides a “framework for the 
collection and analysis of data” and relates to the criteria used to evaluate business research. 
This paper follows a multiple case study design, looking at different revelatory cases of 
22 
 
Swedish insurance companies that are currently working with the IoT. The multiple case 
design allows for a cross-case analysis, comparing the similarities and differences between 
each individual case. Compared to a single case study, this allows for a more robust 
understanding of the impact of the IoT on the business model of different insurance 
companies (Yin, 2009). 
The multiple case study design is in some ways similar to a cross-sectional design, with the 
latter usually being employed with quantitative studies, since it aims to collect wide amounts 
of data from various cases, at a single point in time, in order to analyse the relationship 
between two or more variables (Bryman & Bell, 2015). However, since this paper is not 
necessarily interested in one single point in time or statistical generalizations, this design was 
deemed inappropriate. Moreover, according to Yin (2009), case study designs are suitable for 
answering “how” and “why” questions where the contextual conditions are considered highly 
relevant (Yin, 2009). 
A common critique to qualitative strategy overall, and case studies in particular, are their 
problems with generalization (Bryman & Bell, 2015). However, as previously mentioned, 
this paper is concerned with the deconstruction and in-depth understanding of the particular 
context and features of the research subject rather than the statistical generalization of 
previous theoretical propositions. Still, the fact that multiple case studies rely on multiple 
sources of data and can benefit from prior theoretical propositions was found helpful in 
guiding the data collection and analysis (Yin, 2009). For this reason, the paper still manages 
to create some generalizable findings on the impact of the IoT on the BM of insurance 
companies and the challenges related to its implementation, as well as potential solutions to 
eliminate them. 
3.3. Research Method 
According to Bryman and Bell (2015), the research method refers to the techniques used in 
the data collection and covers how the interview guide was constructed and the respondents 
selected. This section is divided mainly in two parts: primary and secondary data collection, 
where the former is generated from investigation, and the latter is gathered from external 
sources. 
3.3.1. Secondary Data Collection 
After having identified the research area, a literature review was performed in order to 
develop a better understanding of the research area and previously established theory. Due to 
23 
 
the limited knowledge of the author and the limitations of both time and resources, the choice 
was made to perform a narrative literature review, rather than a systematic one. 
The systematic literature review was developed mostly through quantitative research and can 
be defined as a transparent and highly structured process aimed at creating unbiased and 
extensive accounts of the existing literature. Narrative reviews, on the other hand, although 
being more prone to bias, tend to have a wider scope and are therefore more suitable for 
qualitative investigations, especially when the aim is to generate understanding of the 
research area rather than accumulating knowledge (Bryman & Bell, 2015). 
Additionally, this method was found to provide the author with the possibility to change their 
view of the theory as a result of the analysis of the collected data. As with the choice of an 
abductive approach, this method allowed for added flexibility in the face of unanticipated 
issues or topics (Bryman & Bell, 2015). 
The main research objective was to find information regarding the current applications of the 
IoT in insurance companies, as well as its impact on their business model. The search was 
initiated by searching on two different databases: Web of Science and Scopus. However, the 
author found that there was a general lack of existing papers addressing the research area. For 
this reason, the search was extended to Google, where the literature was mostly dominated by 
the research of Management Consulting firms. This provided the author with the theoretical 
framework that describes the impact of the IoT on the BMs of insurance companies and the 
challenges and risks related to its implementation. 
Additionally, some criteria were formed to assist in the literature review. Firstly, the articles 
had to be related to the application of the IoT in insurance-related settings, from a managerial 
perspective. For this reason, any articles that focused solely on the technological or technical 
aspects of the IoT, without taking into account its impact on the Business Model, were 
disregarded. The most frequently used keywords were: IoT, Business Model, Business Model 
Canvas, Business Model Innovation, and Insurance.  
3.3.2. Primary Data Collection 
In order to answer the research questions, qualitative interviews were held with relevant 
actors working with the implementation of the IoT in the Swedish insurance industry. 
Qualitative interviewing is able to capture insights that would otherwise not be detected 
through observation or such quantitative methods as surveys. These were primarily 
employees at Swedish insurance companies currently working with the IoT, as well as one 
24 
 
IoT provider that has worked with insurance companies to develop IoT-centred business 
solutions. 
The interviews were semi-structured with open-ended questions, thus granting the 
interviewer the freedom to delve deeper into relevant areas that were initially not part of the 
interview guide, at the same time as it allows the respondent to answer freely without being 
cut off or lead to an answer. Additionally, with the interviews being semi-structured, 
questions could be added or rearranged if appropriate, whilst still providing a path for the 
interviewer to ultimately get an answer to their research question.  (Bryman & Bell, 2015). 
3.3.2.1. Selection of companies and respondents 
The author chose the interview participants through a generic purposive sampling approach, 
in order to ensure the collection of valuable and accurate data. In a case study design, the 
participants should be selected on the anticipation of the opportunity to learn (Bryman & 
Bell, 2015). For this reason, the interviewees should have prior experience of the research 
area, that is, IoT-centred business solutions. 
The purposive sampling approach is a form of non-probability sampling, meaning that it 
disallows for the generalization of the findings since they might not be representative of the 
whole population (Bryman & Bell, 2015). However, as mentioned earlier, this paper is 
interested in the depth of the particular context, rather than the breadth and generalization that 
comes with larger samples. 
The interviewees were primarily found by searching on Google and LinkedIn for individuals 
that have worked directly or indirectly with the implementation of the IoT in the Business 
Model of Swedish insurance companies. The emphasis was on their experience, rather than 
their specific title. The individuals were then contacted through E-mail and LinkedIn InMail, 
where the author explained the nature of the research and the purpose of their contribution. 
All respondents were native Swedish speakers, so the author chose to primarily communicate 
in Swedish. See Appendix A. 
Additionally, the interview guide was sent to the respondents before the interviews so that 
they could prepare beforehand, or alternatively, refer the author to a colleague they felt was 
better suited to answer the interview questions. 
25 
 
3.3.2.2. Practicalities 
The interview guide was constructed based on the research questions and structured similarly 
to the Theoretical Framework, as to facilitate the coding and data analysis process later on. 
The interview guide started with an introduction of the research and the purpose of the 
interview. During this introduction, the respondents were also asked if they consented to 
being recorded, and whether they wanted to remain anonymous. Then, the interview guide 
was split in three parts: the first section asked for the respondent’s background and 
experience with BM and the IoT; the second section was designed to capture the impact of 
the IoT on the different BMC blocks; and lastly, the third section asked about the risks and 
challenges related to the implementation of the IoT and what insurers can do to eliminate 
those. See Appendix B. 
Due to limitations of time and resources, and the fact that most Swedish insurance 
companies’ headquarters are located in Stockholm, the author chose to conduct telephone 
interviews. One of the disadvantages with telephone interviewing, according to Bryman and 
Bell (2015), is the higher risk for technical difficulties and the ease with which a respondent 
can terminate the interview. This was not found to be a problem during the course of the data 
collection. Additionally, the interviewer cannot interpret the respondent’s body language. 
This, however, did not constitute a major issue seen as the research is not interested in 
capturing the reaction of the individual respondent. 
With all respondents being native Swedish speakers, the choice was made to hold all 
interviews in Swedish, thus making the respondents feel comfortable when answering the 
interview questions and ensuring the collection of valuable data. This makes it crucial to 
translate the answers as accurately as possible in the empirical findings, as to avoid linguistic, 
cultural or methodological problems with their interpretation. For this reason, the author 
made use of respondent validation, whereby the author provided the respondents with a 
summary of the interviews, in order to ensure a good correspondence between the empirical 
findings and the perspective of the interviewees (Bryman & Bell, 2015). 
Seven out of eight interviews were recorded, and no respondents requested to remain 
anonymous. The recorded interviews were transcribed and translated into English as soon as 
the interviews finished, in order to ensure that all relevant information was captured. The 
choice was made to perform a partial transcription, based on two major reasons. Firstly, since 
26 
 
they create larger amounts of data, full transcriptions are more time-consuming. Secondly, 
not all parts of the interview were considered relevant in the context of the research purpose.  
The table below depicts the respondent’s name, position, company and the date, length and 
method use to hold the interview. The interviews were on average 57 minutes long. 
  
27 
 
Company Interviewee Position Date 
Length 
(minutes) 
Method 
Folksam 
Stina Wärn 
Product Developer, Car 
Insurance 
03/05/2019 65 Phone 
Kent Lindmark 
Business Area Manager, 
Damage Prevention 
18/04/2019 67 Phone 
Länsförsäkringar 
Johan Nyman Chief, Property Insurance 26/04/2019 66 Phone 
Mats Arnström 
Product Manager, Health 
Insurance 
24/04/2019 45 Phone 
Justus Alholt 
Tobias Groth 
Business Developer  
Business Developer 
23/04/2019 63 Phone 
Paydrive Emma Jändel Chief Operational Officer 02/05/19 50 Phone 
Moderna 
Försäkringar 
Jonas Frid 
Head of Business 
Development 
17/04/2019 30 Phone 
Talkpool 
Stefan 
Lindgren 
Chief Technological Officer 08/04/2019 70 
Face-to-
face 
Figure 4: Interview Respondents (own elaboration) 
3.4. Data Analysis 
The data gathered from the interviews was coded and interpreted through a thematic analysis. 
According to Braun and Clarke (2006, p. 79), a thematic analysis consists of “identifying, 
analysing, and reporting patterns (themes) within data. It minimally organises and describes 
your data set in (rich) detail”. A theme, in its turn, represents a certain response pattern in the 
data, that is relevant to the research (Braun & Clarke, 2006). This process was chosen since it 
facilitates the breakdown and interpretation of the information gathered during the interviews 
into the different categories and blocks identified by the theoretical framework. Following an 
abductive approach, this was then used to perform a structured comparison of the empirical 
and theoretical findings, which provided a validating effect to the research.  
In order to answer the first research question, the data collected from the interviews was 
organised in the empirical findings according to the BMC structure, which allowed to study a 
single IoT application’s impact on the different blocks that constitute a BM. For each 
question, which corresponded to a specific BMC block, subthemes were created according to 
recurring or relevant subjects. In the financial aspects, for example, such subthemes included 
the reduction of claim frequency and severity in the cost structure, or premium discounts in 
the revenue streams. The risks with the implementation of the IoT were organised in a similar 
way without following an established framework, but instead bigger recurring and relevant 
28 
 
themes such as external resistance, organisational resistance and disruption threats, which in 
their turn had their own subthemes. See Appendix C for coding tables. 
With the paper following a multiple case design, the empirical findings aim to present the 
similarities and differences among the different cases and respondents. Together with the 
abductive approach to theory generation, whereby the empirical findings are compared to the 
theoretical, this allowed for a better understanding of the level of impact of every IoT 
application on the different blocks that constitute the BMC, as well as the challenges related 
to its implementation.  
In order to determine the level of impact of every application, a table (figure 6) was 
constructed comparing the theoretical and empirical findings. The impact of each factor was 
designed by taking into consideration the frequency with which a certain subtheme emerged 
during the interviews (see appendix C), as well as their presence in the theoretical findings. If 
a subtheme is mentioned by less than one third of the respondents, then the impact is low, 
whilst if more than two thirds of the respondents mention a subtheme, the impact is high. 
Finally, if a subtheme is mentioned by more than one third but less than two thirds of the 
interviewees, then the impact is medium. Moreover, whenever a subtheme is mentioned in 
the empirical findings, but not in the theoretical framework the impact is also considered low. 
3.5. Research Quality 
The most normal criteria to evaluate business research include reliability, validity and 
replication. However, many researchers have stated that these criteria are not as relevant to 
qualitative research, as they were initially invented to assess quantitative research (Bryman & 
Bell, 2015). For this reason, these terms have been adapted to fit a qualitative research 
strategy and are discussed below. 
Internal validity is concerned with whether the theoretical propositions developed by the 
researcher match the researcher’s observations. By providing the transcripts of the interviews 
and research findings to the interviewees, the researcher made use of respondent validation. 
Additionally, triangulation was used during the collection of both primary and secondary data 
to ensure that the findings were not influenced by one single source of information. This was 
done by interviewing more than one representative of a single insurance company and 
searching for academic literature from multiple databases (Bryman & Bell, 2015).  
29 
 
External validity refers to the degree to which the research findings can be generalized into 
other contexts. Qualitative research methods are usually employed in smaller samples, which 
makes generalization more difficult. The case is no different with a case study design, that 
usually looks at a few cases. However, this choice was made primarily because this paper is 
more interested in explaining the particular details of a certain situational context, rather than 
creating generalizable findings (Bryman & Bell, 2015).  
External reliability, similar to replication, relates to the degree to which a study can be 
replicated, that is, if another researcher would reach the same outcome if they were to 
reproduce the same procedures and processes. Seen as it is impossible to “freeze” a social 
setting and the conditions in which the research was performed, this becomes a difficult 
criterion for qualitative business research to meet (Bryman & Bell, 2015). 
Moreover, due to the interpretive nature of qualitative research, it is prone to subjective 
preferences and interpretations of the researcher, who is usually not interested in quantifiable 
measures. For this reason, the author has avoided including preferential bias or subjective 
opinions into the research to the best of his abilities. This subjective nature of qualitative 
research creates further problems with replication, seen as a paper can many times become an 
interpretation of what the researcher themselves find relevant or interesting (Bryman & Bell, 
2015). To avoid problems with replication and transparency, this paper goes into great 
lengths to describe in full detail the choices and the reasons behind the decisions made during 
the course of the project. 
Internal reliability relates to whether the members of the research team agree about what they 
see and hear. Seen as this study is performed by a single researcher, the author would 
regularly seek the advice from the thesis supervisors and other external points of view. 
Additionally, the use of respondent validation and triangulation also ensured a greater level of 
internal reliability. 
 
  
30 
 
4. Empirical Findings 
In this section, the author presents the relevant findings from the semi-structured interviews 
held with the employees of different insurance companies. This section aims to show what the 
respondents believed to be the impact of the IoT on the different blocks of the BMC, as well 
as the challenges related to its implementation and recommendations of potential solutions. 
4.1. Cases 
Below, the author shortly describes the IoT projects that each case company is currently 
working with. Eight interviews were held with employees at four different Swedish insurance 
companies, as well as an employee from the IoT provider Talkpool, in order to provide an 
external perspective to the results. 
4.1.1. Moderna Försäkringar 
In 2017, Moderna Försäkringar launched a Usage Based Car Insurance called Enerfy, in 
partnership with the insurtech company Greater Than. The offering consists of a plug-in 
sensor that is installed into a car’s OBD (On-Board Diagnostic) input, that then measures the 
driving behaviour of the policyholder. The collected data is then sent to a cloud, where it is 
analysed with the help of the Artificial Intelligence Greater Than developed to match the 
policyholder to one of the fifteen different premium levels provided by Moderna. Finally, the 
policyholder can view all the collected information, including their risk score, through an 
application on their telephone. Moreover, the service also provides other integrated services 
such as a fleet management tool, a driving journal, parking assistance, among others (Greater 
Than, n.d.).  
4.1.2. Paydrive 
Paydrive started in 2013 with the aim of providing customers with a car insurance solution 
that allows for drivers to affect their own premiums, with the help of the IoT. The sensor, 
developed and distributed by Telia, is connected to the car’s OBD port and measures the 
distance and driving behaviour of the policyholder. The policyholder can then view 
information on their driving behaviour through the mobile application Telia Sense, as well as 
utilize other features such as car tracking, parking assistance, driving journal and internet 
connection. Like Enerfy, the customer pays a base premium, that covers the lease of the Telia 
Sense sensor and application, and a usage-based premium, that depends on the distance and 
driving style of the insured. Every trip is given a driving score from 0-100, where 70-100 is 
31 
 
calm, 25-70 is medium and under 25 is tough. If the customer receives a 100-point score, 
then they only pay for the base premium itself (Paydrive, n.d.). 
4.1.3. Folksam 
Folksam, a mutually owned insurance company, is currently testing a few different IoT 
applications, with one of them being Köra Säkert, a PAYD insurance solution. According to a 
study by Folksam, speed is one of the most normal reasons for traffic accidents, and by 
providing real-time feedback on their speed, customers have been shown to drive over the 
speed limit less frequently. Like Paydrive, Folksam has partnered up with Telia Sense, 
meaning that the customer can choose their PAYD insurance solution without having to buy 
another OBD sensor or changing applications (Folksam, n.d. b). 
Customers at Folksam can still choose to buy a traditional car insurance. However, if they 
choose to subscribe to Telia Sense, they can reduce their car insurance premium by 20 % 
depending on their average driving speed (Folksam, n.d. b). The user installs a light indicator 
visible behind the wheel that lights green when they are under the speed limit, yellow when 
they are a bit above it and red when they are driving too fast. The light indicator itself is 
developed by Folksam and is connected via Bluetooth to the Telia Sense sensor. So, if the 
policyholder never drives over the speed limit, they receive a 20 % premium discount 
(Folksam, n.d. b). 
Another area where Folksam is implementing the IoT is property insurance. According to 
Lindmark, Folksam has partnered up with Anticimex to provide water leakage protection to 
their policyholders’ properties, with the help of IoT sensors. Policyholders can install water 
leakage sensors in high-risk areas such as bathrooms, kitchen sinks and other water-
dependent appliances, that then signal the user if the humidity levels are too high. A water 
leakage circulation breaker is also beginning its testing phase. These smart sensors would be 
installed on high-pressure pipes close to the water meter and be able to turn off water 
circulation in the case of a water leakage, regardless of its extent. In return, the policyholder 
receives a 10 % premium discount on their home insurance (Folksam, n.d. a). 
4.1.4. Länsförsäkringar 
Länsförsäkringar consists of 23 local mutually owned insurance companies and is currently 
working with three separate IoT-centred solutions: water circulation breakers, Ride Safe, and 
Aifloo. 
32 
 
In an attempt to reduce the damages resulted from over 100 000 yearly water leakages in 
Sweden, Länsförsäkringar is currently selling water leakage sensors as well as water leakage 
circulation breakers, as a complement to traditional property insurance. When a water leakage 
is detected by the sensor, it sends the information to the breaker, that then shuts off water 
circulation in order to prevent larger water damages in the property (Länsförsäkringar, n.d.). 
The second project is Ride Safe, a solution that is meant to reduce phone usage whilst driving 
and is currently in a pilot phase to measure the demand in private and business segments. 
According to Ride safe (n.d.), around 30 % of traffic accidents happen due to a lack of 
attention related to phone usage. Ride Safe consists of a sensor that is attached to the vehicle 
and connected to the policyholders’ phone, so that whenever the vehicle is in movement, all 
incoming notifications are blocked, including calls, unless the user has a hands-free solution.  
Additionally, the policyholder can also choose the level of protection, by allowing maps or 
automatic replies to incoming messages. The phone usage per trip is then used to calculate a 
driving score, that allows Länsförsäkringar to advise customers on how to become better 
drivers (Ridesafe, n.d.).  
Lastly, Aifloo is the name of a MedTech company that Länsförsäkringar has invested in. 
Aifloo has developed a bracelet device that, together with in-home sensors and Artificial 
Intelligence, studies movement patterns and detects if there are any anomalies. For example, 
if the user has not moved in a relatively long time, the system sends an SOS alarm, which can 
also be activated manually by the user. According to Arnström, there were two major reasons 
that Länsförsäkringar chose to invest into this technology. The first one was that 
Länsförsäkringar noticed a growing elderly customer segment, with a growing percentage 
living alone in private homes, even though they might suffer from an illness. The second one 
was to create a service that gives both elderly and their relatives ease of mind and a sense of 
security.  
4.1.5. IoT Expert – Talkpool 
Finally, the author found it relevant to interview an IoT provider that is currently working in 
the Swedish insurance market, in order to add an outside perspective to the empirical 
findings. Stefan Lindgren works for Talkpool, a global provider of IoT products and services 
to businesses that want to create smarter buildings, industries and cities. One of the customer 
segments targeted by Talkpool is the Swedish Insurance industry, where they are currently 
providing water leakage sensors capable of detecting high humidity levels.  
33 
 
4.2. Impact of the IoT on the Business Model of Insurance Companies 
Below, the reader will find a BMC summarizing the empirical findings on the impact of the 
IoT on the BM of insurance companies: 
 
Figure 5: BMC summary of empirical findings (own elaboration) 
4.2.1. Financial Aspects 
4.2.1.1. Revenue Streams 
In general, most respondents agree that there is a trend towards more dynamic pricing 
mechanisms that are decided on individual usage and behaviour, rather than the traditional 
historical and socio-demographic indicators that insurers have relied on. This is the case of 
such usage-based car insurance solutions like Enerfy, Paydrive and Köra Säkert. 
Both Paydrive and Enerfy, for example, rely on the mileage and driving behaviour of the 
individual policyholder, which is collected by the OBD sensor. Specifically, the policyholder 
pays for a base premium depending on the traditional indicators, which covers the cost of the 
sensor hardware and software, and a variable usage-based premium, that depends on their 
mileage and driving behaviour. Folksam and their solution Köra Säkert make use of a similar 
mechanism to incentivise positive driving behaviour, except that the customer receives a 
premium discount of maximum 20 % on the normal car insurance premium, if they 
constantly drive under the speed limit 
Nyman, Alholt and Groth, however, argue that it is hard to measure the current demand for 
UBI. For example, Nyman mentions that most of Länsförsäkringar’s customers want a 
consequent sense of security throughout their lives and have not sensed any real demand for 
insurance that only covers the insured object for a certain amount of time or when it is being 
Key Partners Key Activities Customer Relationships Customer Segments
•Strategic alliances 
with IoT-providers
•Improved risk and 
pricing models
•Shift to more frequent 
interactions
•Shift to more profitable 
segments
•Outsource technical 
product development
•Shift to advice-led interactions •Shift to personalized 
offerings
•Efforts to improve 
cooperation
•Speedier claim management
•Gamification/loss aversion
Key Resources Channels
•Human labour with 
digital competence
•Reliance on partner channels
•Increased supplier costs •Shift from fixed menu to dynamic pricing
•Claim frequency and severity reduction •Up-sale opportunities
•Cost reduction
Co t Structure Revenue Streams
•Shift from cost- to value-driven BM •Shift from subscription to usage-based fees
•Risk reduction
•Increased 
personalization
•Increased convenience
•Reduced negative 
impact on environment
•Improved pricing 
•Premium discounts
•New offerings/revenue streams
Value Proposition
34 
 
used. Furthermore, the three respondents argue that insurance premiums have always been 
adjusted according to usage in one way or another, meaning that usage-based insurance does 
not save that much money for customers in the long-run. For example, when a customer 
applies for car insurance, they are required to fill in an estimation of their monthly mileage, 
which is then included into the risk and pricing calculation. Lastly, Lindmark sees a hard time 
moving to more dynamic pricing in home insurance, a context in which an UBI solution 
would not make a large difference in pricing models. 
Another area of interest is the creation of new offerings and revenue streams. The IoT’s 
applications in the insurance industry are not limited to risk insurance but extend to integrated 
solutions that constitute potential sources of revenue for insurance companies. To help 
illustrate, the water leakage sensors implemented by Folksam are able to detect water 
leakages, as well as measure water consumption and turn off water circulation. Lindmark 
mentions how these applications could help insurers develop a dashboard where the customer 
can view their water usage, recommendations on how to reduce it or even set a consumption 
level at which the water circulation is shut off. Additionally, Lindmark provides another 
example where insurers guarantee that a repairman is sent to fix the problem identified by the 
water leakage sensor. The cost of the service could be covered by the reduced risk provided 
by the IoT, embedded into the premium or paid as an extra fee. According to Lindmark, this 
is mainly dependent on how the offering is packaged, the customer willingness to share their 
sensor signals, and what customers are interested in and willing to pay. 
This trend can also be seen in the PAYD market. Services such as Paydrive and Enerfy 
provide policyholders with several non-risk related functions such as a driving journal, car 
tracker and parking assistance. Frid mentions that the data collected from the car can also be 
used for car repair, inspection, maintenance and financing, in ways that would otherwise not 
be possible. These functions are currently embedded into the premium, which is paid by the 
policyholder through a monthly subscription. 
Moreover, some of the respondents mention that with access to more structured data on the 
individual policyholder and their situation, insurers have the possibility to become more 
relevant in their communication and offerings, thus presenting insurers with an opportunity to 
up-sell to the customer. This would probably not be experienced as negative by customers 
since insurers will have a better basis for such targeted recommendations, instead of 
35 
 
providing generalized recommendations that have a negative impact on the customer 
experience. 
Lastly, insurers have to address the issue with increasing customers’ willingness to share their 
data and adopt IoT-centred insurance solutions. In the discussions Lindgren has had with 
insurance companies, the idea has been to primarily reduce the risk that insurance companies 
take on when they insure their customers. Currently, insurers subsidize fire and smoke alarms 
when policyholders purchase a home insurance, so they should be willing to do the same for 
IoT sensors in order to reduce their risks in the long-term. In this case, the policyholder 
receives a lower premium if they decide to install an IoT sensor and share their data with the 
insurance company, since there is a lower risk for damages going by unnoticed. Accordingly, 
this lower premium is one of the main value propositions of PAYD solutions. According to 
Wärn, this premium discount constitutes a loss of revenue. However, insurers expect to make 
up for it on the cost side, in the form of reduced claim frequency and severity. 
4.2.1.2. Cost Structure 
Most respondents agree that the IoT will provide insurers with a better opportunity to detect 
damages in their early stages, or even before they arise, allowing for a reduction in the 
number of submitted claims, and thus their following costs. Furthermore, with claim pay-outs 
being the biggest expense for insurance companies and the IoT being a relatively cheap and 
accessible technology, Nyman expects it to create a net cost reduction for insurance 
companies. 
As an example, Lindmark and Lindgren mention how water leakage sensors are able to 
reduce the number and cost of each individual claim. According to Lindgren, a bathroom that 
has suffered a water damage can easily cost insurance companies about 100.000 SEK, as well 
as another sum for the property owner in the form of a deductible and value depreciation. By 
detecting water leakage in its early stages, the IoT helps insurers reduce the frequency and 
severity of claim pay-outs. Additionally, Lindmark also points out that by reducing the 
number of claims, the IoT can indirectly reduce the amount of labour necessary to manage 
them. This can also be done by automating certain parts of the claim process that are captured 
by the IoT sensors, a priori. 
However, both Frid and Wärn hesitate to confirm that their sensors have had a positive effect 
on cost reduction, since they have not yet seen any objective evidence that their offering has 
actively reduced the number of claims. This is because it takes time to gather enough data to 
36 
 
compare the results, and that there are several other factors that influence the numbers, 
making it hard to isolate the cause and effect relationship. Still, both respondents see a logic 
in how a driver that is rewarded for good behaviour will most likely incur fewer damages. 
Most of the respondents explain these cost reductions as a result of an increased focus on the 
prevention of damages before they happen. However, some respondents describe this effort 
as more of an extension of the preventive work the insurance industry is already doing.  As an 
example, both Lindgren and Arnström mention how the building regulation for bathrooms in 
Sweden was largely developed by insurance companies because poorly built bathrooms have 
a higher risk for water damages. The same can be seen when insurers sell fire alarms in 
conjunction with home insurance. Therefore, preventing damages, is, according to Arnström 
and Wärn, built into the “DNA” of insurance companies. 
Regarding the costs in the implementation of the IoT, Lindmark mentions how the close 
cooperation with IoT providers increases supplier procurement and monitoring costs. Another 
cost that is created, depending on the relationship of the insurer and external partner, are 
distribution costs. Frid mentions how insurance policies differ from IoT sensors, in that the 
former are easily and cheaply distributed, whilst the latter brings logistical challenges. 
Folksam, in their turn, took some of the light indicator development in-house, but did 
however, outsource its production and distribution to an external partner. 
4.2.2. Customer Interface 
4.2.2.1. Customer Relationships 
According to Lindmark and Frid, one of the objectives of insurance companies is to improve 
on customer relationships in a way that creates customer loyalty. Currently, a customer will 
only interact with an insurance company when they are looking for an insurance solution, 
paying their premium or making a claim. Since the aim of insurance is to share risks among a 
large group of people, most customers do not suffer any damages, which considerably 
decreases the number of interactions.  
Additionally, Lindmark explains that the premium is currently a big factor in the customer 
purchase decision. Folksam, as well as other insurance companies, must set competitive 
premiums in order to maintain their market share. This approach is, however, not sustainable 
in the long run. By improving on customer relationship management, customers might start 
looking at more than premium pricing when comparing insurance solutions. As will be 
discussed below, this is an area which the IoT can potentially improve on. 
37 
 
The empirical findings suggest that the IoT allows insurers to expand on this customer 
interaction and provide almost a daily service to policyholders, by sharing advice and 
recommendations on how to reduce the risk for losses. For example, both Enerfy and 
Paydrive’s applications provide policyholders with driving recommendations as often as they 
wish2. In this way, the IoT allows for a more supporting and interactive relationship with 
customers. Alholt and Groth also mention that the IoT can help inform the policyholder that 
they are about to enter an area with an increased risk for traffic accidents at a specific time of 
the day.  
Another example is provided by Lindmark, who mentions that when a water leakage is 
detected, the insurer can contact the policyholder to discuss different measures they can take 
to alleviate or eliminate the problem. Furthermore, Lindgren mentions in his response how 
customers most likely prefer to receive an offer to purchase a de-humidifier due to a higher 
risk for dry rot, than simply letting it go unnoticed until the damage is irreversible. No one 
wants to suffer damages, and if insurers are able to nudge customer in the right direction, 
Lindgren does not believe they will view it as a “Big Brother community” where they are 
constantly monitored, but instead, that they benefit from a better insurance protection. 
This advice-led interaction is not entirely new for insurance companies. Lindmark explains 
how Folksam has worked with high-risk customers by informing them of different ways to 
reduce their risk for losses. According to Lindmark, these preventive recommendations have 
a positive effect on customer retention and behaviour. Likewise, Arnström states that a 
customer will most likely prefer the insurance provider that not only provides accident 
insurance, but that also helps the customer avoid injury in the first place. So, by improving 
the customer experience and insurance protection, customers might find themselves looking 
for more than the lowest insurance premium. 
Lastly, by providing policyholders with a driving score and active feedback, Wärn and Jändel 
both discuss how their service adds a gamification effect, whereby the policyholder is more 
prone to change their behaviour due to loss aversion. 
4.2.2.2. Customer Segments 
Insurance is a collective product where many are supposed to share few risks. At the same 
time as the customer experience becomes more individual, insurers want to diversify their 
                                                 
2 The application provided to Paydrive users is developed by Telia Sense, but is included into the premium 
payment, and therefore considered as a part of the insurance offering. 
38 
 
risks among a larger group of individuals. Therefore, most of the respondents did not mention 
any major changes in the way insurers segment their customers. 
According to Lindgren, the added individualization brought by the IoT has both advantages 
and disadvantages. For example, is it reasonable that a healthy person that does not drink and 
exercise should have a different premium from someone who does not? Or someone that 
suffers from a certain kind of illness? The respondent did not have a clear answer to these 
questions but believes that somewhere there is a need for some solidarity. On the other hand, 
the added individualization can also remove some of the structural unfairness that there is 
today. Should someone who lives in a dangerous neighbourhood but is very careful with their 
belongings, be charged more simply because they happen to live there? 
Similarly, Alholt and Groth state that a balance is necessary, and that insurers need to make it 
possible for anyone to insure themselves. The insurance industry is regulated by law and 
must therefore conduct business in a certain way regarding premium pricing and excluding 
people from insuring themselves. This includes more risk-prone customers and policyholders 
that suffer relatively more damages. Insurers can work with such groups, with the help of the 
IoT, to reduce their risk level.  
Furthermore, Lindmark adds that it is dangerous to create a situation where some cannot 
insure themselves because they are not technically interested. The three PAYD solutions, for 
example, have attracted a more technology-friendly customer segment that is more willing to 
engage in digital solutions. Additionally, their service is more advantageous to low-risk 
customers that benefit from a more dynamic pricing that is decided on individual driving 
behaviour. By that same logic, those that have a riskier behaviour will not want their data 
collected. For this reason, insurance companies will face some difficult ethical questions 
about how to segment their customers. 
4.2.2.3. Channels 
Most IoT-centred insurance solutions mentioned in this paper include some sort of user 
interface where the customer can view relevant information regarding their insured objects., 
allowing insurers to increase the number of customer interactions. For example, the customer 
might receive a push notification when they leave their house informing them that everything 
is in order, so that they focus on the day ahead of them.  
The Telia Sense and Enerfy application, for example, allow insurers to more easily interact 
with policyholders by offering them driving recommendations as often as they wish. On 
39 
 
Folksam’s homepage, for example, the policyholder can log into their account to see 
pertinent information and other details relevant to them. However, Lindmark and Wärn both 
mention the importance of ensuring that only the policyholder can view that information, and 
that they are not spammed with irrelevant information. 
4.2.3. Value Proposition 
Insurers will not be able to experience the positive effects from the IoT if customers do not 
feel as if they benefit from adopting an IoT-centred insurance solution. The empirical 
findings suggest that the value proposition is the BMC block most affected by the IoT.  
Below, the author presents the different ways in which the respondents believed the IoT 
influenced their value proposition. 
Most often, there is an indication that something is awry, where an incremental deterioration 
takes place long before the damage is noticeable. The IoT provides the customer with an 
active monitoring that detects such indicators in their early stages before the damages takes 
place. Lindmark provides an example related to water leakage sensors. If the sensor detects 
that there is mould in the roof, the insurer can very cheaply either inform the customer of a 
humidity problem or offer to fix it themselves. On the other hand, if the customer lets it go 
unchecked until infrastructure beams start deteriorating, then it can become extremely 
expensive for the insurance company, but even the policyholders themselves, in the form of 
deductibles and value depreciation. Additionally, Alholt and Groth comment on how 
customers that have experienced a water leakage damage in the past will most likely not want 
it to happen again, even if their insurance policy covers it.  
Frid and Jändel, on the other hand, point to more accurate pricing models as the primary 
reason for the development of their usage-based car insurance solutions. With PAYD 
solutions, policyholders receive a better premium that is decided on their individual driving 
behaviour rather than other traditional external factors insurers rely on. On the other hand, 
Frid argues that such dynamic pricing has its drawbacks. One of them is that some customers 
prefer a predictable and steady premium rather than one that fluctuates from month to month. 
Nevertheless, there is a close relationship between risk and premiums, as Alholt and Groth 
discuss. Primarily, insurers get paid for risk and if there is a potential to reduce that risk with 
the help of the IoT, then the price follows with. For example, if an individual is older, there is 
a higher risk that they die, and therefore a higher life insurance premium. 
40 
 
Besides reducing the premium and the risk of paying costly deductibles, policyholders’ costs 
are also reduced through the non-risk related functions mentioned in 4.2.1.1. For example, the 
water leakage sensors can measure the water consumption in the household and turn off the 
circulation after a certain point. Besides this, Folksam has calculated that the reparation of 
one water leakage damage can emit an average of 200 kilograms of carbon dioxide. By 
reducing the risk and therefore the number of water leakage damages, the IoT reduces the 
negative impact of such damages on the environment. 
Similarly, the PAYD solutions have observed a positive effect on fuel consumption by 
incentivising policyholders to drive in a moderate speed and avoiding harsh accelerations or 
brakes. Additionally, in their respective applications, the drivers can also view their carbon 
dioxide emissions and receive recommendations on how to reduce them. Therefore, the IoT-
centred insurance solutions allow policyholders to live more sustainably, both economically 
and environmentally. Furthermore, the PAYD solutions also offer other non-risk related 
services that improve the convenience for the customer, by having different functions 
integrated together with their insurance. This includes car inspections, parking assistance, 
GPS, driving journal, and so on. 
Finally, these shifts towards non-risk related services, individualized pricing and 
recommendations, provide customers with an added level of personalization and 
convenience. The customer is also able to receive a better insurance protection, since 
insurance companies can take on new risks and insure more things. For example, Lindmark 
mentions that it is possible that if the policyholder takes good care of their home, then 
perhaps insurers will be able to take on new risks such as the façade or the roof. 
4.2.4. Infrastructure Management 
4.2.4.1. Key Activities 
According to the respondents, understanding and valuing risks is the core activity of 
insurance companies. Today, insurers look mostly at age, geographical location and other 
external factors for their risk and premium calculations. Hence the big difference in the car 
premiums for a car that spends most of its time in a garage in the countryside and a car that is 
often driven around in Central Stockholm. 
Most respondents agree on the fact that the IoT has the potential to provide insurers with 
more detailed and relevant data from the physical world, thus improving risk understanding, 
which in its turn creates more accurate pricing models. At the same time, both Nyman and 
41 
 
Lindgren explain how insurance companies are currently working with Big Data analysis, 
meaning that it is already one of their key activities and will most likely not suffer big 
changes. Instead, the main difference is in the quantity and complexity of the data, which is 
why insurers need to develop the internal assets to integrate it into their current risk models. 
Currently, Länsförsäkringar and Folksam are only using their water leakage sensors to 
prevent damages, rather than to collect data and develop their understanding of risk. Even 
though this might grant opportunities to learn patterns and trends in signals, that might later 
lead to forecasting accidents, managing the data collected from the sensors is challenging for 
insurers, a topic that will be further developed in 4.3. So, in the short term, the IoT serves 
more of a preventive role in tracking and monitoring insured objects.  
Another key activity that insurers chose to keep in-house involves customer relationship 
management and business development. It is only when the customer has accepted the 
offering, that the IoT provider takes over and manages the activities in which they are skilled. 
The activities of key partners are further developed in 4.2.4.2., but usually include the 
activities related to the sensors themselves, their development, production, distribution, 
maintenance and data collection. 
Similarly, Arnström mentions that Länsförsäkringar usually does not partake in the base 
R&D, instead they manage the commercial side, which according to Nyman encompasses 
customer relationship management and business development. For this reason, 
Länsförsäkringar enters the project in a phase when the product is ready for pilot testing. 
Arnström clarifies that this does not mean that the product has to be perfect by the time the 
insurer enters the picture, but it does need to be functional. 
4.2.4.2. Key Resources 
It seems that insurers will not suffer radical changes in terms of key resources, but a minor 
shift towards more digitally competent human capital. Even though insurers already analyse 
data to understand risks, Nyman mentions that there is a need for human capital with skills in 
data management and modelling. This is because the IoT provides more sophisticated data 
that makes it more challenging to integrate it with the risk models insurers are currently 
working with.  
For example, Folksam has recently recruited several IT specialists in order to obtain more in-
house IT and digital competence, instead of relying on external consultants. Lindmark 
mentions that the same must be done in regard to business development and customer 
42 
 
relationship management, to develop the types of competencies that a company needs to be 
successful in an ever-changing landscape. According to Nyman, this will also require a good 
understand of customers’ behaviour, wants and needs in order to develop relevant offerings. 
4.2.4.3. Key Partners 
Among the cases chosen for this study, all insurance companies relied on external partners 
for the development of the sensors and their data collection. According to Lindmark, insurers 
will most likely never be leaders in technology, but by joining forces with external partners, 
insurers are able to collect real-time information on insured objects and develop better risk 
and pricing models. 
For example, Talkpool provides insurers with a package that includes the hardware itself, 
data collection and sometimes, connectivity, if the insurer does not prefer to use an own 
network. Similarly, Moderna partnered up with Greater Than for the development of the 
application and the AI algorithm used to calculate the risk score for the individual 
policyholder. 
Both Paydrive and Köra Säkert are made available through Telia, who is responsible for the 
development and distribution of the sensors, as well as the mobile application. Wärn 
mentions that one of the reasons for this collaboration, is to allow customers to change 
insurers without having to purchase another OBD sensor from another insurer. Instead Telia 
provides the platform for insurers to adopt a standardized technology easily installed by the 
customers themselves. 
According to Lindmark, there is a need to discuss who has the right to manage data and profit 
from it. If this information is shared with third parties for free, then insurers might 
inadvertently create new competitors. For example, the IoT provider can have a distributor 
role where they provide the data to insurers, without being able to use the data in any other 
way than to distribute it.  
Furthermore, part of this relationship with key partners is ensuring that they deliver on what 
was agreed upon. Lindmark mentions that with the water leakage breakers, Folksam has 
chosen to procure the IoT providers themselves and set up mutual KPI, to create a closer 
relationship where the partners learn and develop together. Similarly, Länsförsäkringar and 
Aifloo have had a couple of mutual workshops to define the content and objective of the test 
pilot, where Länsförsäkringar noticed some elements were lacking. 
43 
 
4.3. Managing the Risks with the Implementation of the IoT 
In this subsection, the author presents the empirical findings from the interviews, regarding 
the risks and challenges related to the implementation of the IoT, and the potential solutions 
presented by the respondents. 
4.3.1. External Resistance 
Most respondents believe that data management and privacy concerns are currently the 
biggest challenge for insurance companies. Lindmark, for example, discusses how there is a 
regulatory and ethical question regarding the collection and management of data, and how 
insurers cannot allow policyholders to feel as if they are living in George Orwell’s 1984, 
being constantly monitored.  
Lindgren touches upon the same subject by asking the following questions: Who owns the 
data? Typically, insurance companies are only able to write one-year contracts for home 
insurance, so how can they manage data and its transfer from one insurer to the next? How 
can insurers ensure that end-customers feel they have control over the data and give 
acceptance to share their data, without experiencing that they are being watched over? Most 
insurance companies are reinsured, so where does the data trail end? Does it end at the 
service provider, insurance provider or reinsurance provider? 
According to Lindmark, managing this issue in a clever way is a major challenge for 
insurance companies and needs to be managed in parallel with technology advancements. 
Insurers must understand how to make customers recognize that they benefit from sharing 
their data with insurance companies. Most respondents agree that for this to happen, insurers 
will need to provide premium discounts or a value-added service that is enticing enough to 
warrant data-sharing. Specifically, Lindgren predicts that this will happen mostly through 
premium discounts, a little bit varying on the kind of data. For example, the data collected 
from water leakage sensors in the bathroom floor is not that sensitive. On the other hand, 
there is the kind of data that shows where the user is and how fast they are driving. By 
discounting premiums, customers have a bigger incentive to share their data with their 
insurers. 
In the end, according to Lindgren, it is up to the wallet of the policyholder. But more 
importantly, Lindmark states that the customer needs to trust the insurer and the way in which 
they manage data. When a company loses trust, customers become sceptic to future offerings, 
which negatively impacts the business in the long run. For this reason, Jändel and Lindgren 
44 
 
believe it is very important for insurers to clearly communicate on how the data will be used, 
analysed and shared. 
Similarly, Arnström, Alholt and Groth believe that the key to increasing customers’ 
willingness is communicating to them how they benefit from the offering, otherwise they 
have no real incentive to share their data with insurers. In summary, insurers need to find a 
good balance, without making the customer feel like they or their property is being controlled 
by guaranteeing that the data will not be used in an inappropriate or controlling way, but that 
it adds a sense of security.  
Technology can also play a role in solving this issue. For example, Arnström mentions that in 
many cases, it is not a person that is actively monitoring an individual customer’s 
movements, but an AI engine. Additionally, Wärn mentions that Köra Säkert only collects 
data on an aggregate level, so Folksam cannot pinpoint where and when exactly the driver 
was over the speed limit. Regarding the customer data transfer from one insurer to another, 
Talkpool is currently looking at a blockchain solution to ensure that the policyholder owns 
the data and can transfer it when they change insurance companies.  
Another external challenge mentioned by Lindmark is getting other parts of society to be a 
part of the transformation. One of the groups that must be willing to accept these changes are 
the craftsmen that insurers currently rely on to fix damages. The increased focus on 
prevention might reduce the demand for their services. Instead of going in to rip open a floor 
and repair a damage that has already occurred, repairmen will then go in to make sure that the 
damage never happens in the first place. That in its turn might create indirect resistance due 
to a decrease in labour hours and subsequent revenue streams. Additionally, Lindmark thinks 
of Government authorities as a part of the supply chain, since they can heavily influence the 
adoption of IoT solutions. As an example, in the future, all buildings might be required to 
have smart water sensors installed, since it has such an impact on damage reduction.  
4.3.2. Organisational resistance 
The conservative nature of insurance companies is another of the major challenges this 
industry must tackle. Lindmark compares insurance companies to a hangar ship, where one 
does not simply turn the wheel and change the navigation curse. Instead, it takes quite a few 
sea miles before the course is changed, and with that, time.  
Currently, Folksam has an older IT infrastructure that has shaped the way in which they work 
and organise themselves. Changing to a newer infrastructure with a new way of working 
45 
 
might influence different groups of employees, which might create some internal 
organisational resistance. Some will have to change the way they work, whilst some tasks are 
created, and others eliminated. For this reason, both Nyman and Lindmark point out the 
importance of finding the right kind of competence and working with change management, in 
order to address digitalization and improve the customer experience. So, it not just the IT 
infrastructure and algorithms that need to adapt, but the intellectual resources as well. 
According to Lindmark, insurance companies have faced similar problems in the past, but 
perhaps in different contexts. So, by being aware of these hinders, insurers will be able to 
think of new solutions. Moreover, the respondents stress the need to understand customer 
insights and demand before jumping onto a business proposition. 
4.3.3. Disruption threats 
Lindgren believes that insurers can be easily disrupted from different directions. Such 
disruption threats include servitization, where more people are choosing to lease their 
products rather than owning them; new types of insurance, that come from the development 
of new technologies such as blockchain; and lastly, manufacturers taking a stronger position 
and weakening that of their distributors, which can also be witnessed in the insurance 
industry. For these reasons, Lindgren means that insurers should feel the pressure to change. 
Furthermore, with the increased focus on prevention, the risk is continuously disappearing 
from the market. For this reason, Frid recommends that insurers shift their focus towards 
businesses, rather than private consumers. In a business landscape, the IoT has more to offer 
such as product monitoring and tracking. Frid also sees the increasing disintermediation of 
insurers as a challenge, whereby insurance solutions are increasingly being offered directly 
by car manufacturers. In order to address this and maintain their market share, insurers must 
search and build new alliances. 
According to Nyman, there is also a question of how much data product manufacturers are 
willing to share with insurers. There is a trend where more customers choose a service over 
owning a product, which creates a demand for a new kind of insurance that is not necessarily 
targeted at the subject, but the object. When an insured object is shared among different 
subjects, it makes more challenging to calculate the risk. 
Ultimately, there is not much insurers can do to prevent technology advancement. The only 
thing companies can do is to try to adapt to the forces that rule the market. One part of the 
solution includes getting in direct contact with product manufacturers and distributors, whilst 
46 
 
another part is excelling within advisory, prevention and customer experience. Ultimately, 
insurance companies need to change their role and be more frequently involved with both 
their private and business customers, rather than coming in contact once a year through a 
yearly renewal letter.  
47 
 
5. Analysis and Discussion 
This section aims to provide the reader with a structured comparison of the theoretical and 
empirical findings, ultimately leading to generation of new theory on the impact of the IoT on 
the business model of insurance companies, and its challenges related to its implementation. 
5.1. Impact of the IoT on the Business Model of Insurance Companies 
Below, the reader can find a structured comparison of the subthemes mentioned in the 
theoretical and empirical findings, which was used to aid in the analysis. See 3.4. for more 
information on how the following table was constructed. 
The Impact of the IoT on the Business Model of Insurance 
Companies 
Theoretical 
Findings 
Empirical 
Findings Impact 
BMC Block Factor Mentioned Mentioned 
Fi
n
an
ci
al
 A
sp
ec
ts
 Revenue 
Streams 
Shift from subscription to usage-based fees   Medium 
Shift from fixed menu to dynamic pricing   Medium 
New offerings/revenue streams   Low 
Reduced revenues from premium discounts   High 
Up-sale opportunities   Medium 
Cost 
Structure 
Claim frequency and cost reduction   High 
Shift from cost- to value-driven BM   Low 
Increased supplier costs   Low 
Shift to more profitable segments   Low 
C
u
st
o
m
e
r 
In
te
rf
ac
e 
Customer 
Relationship 
Shift to more frequent interactions   High 
Shift to advice-led interactions   High 
Gamification/loss aversion   Low 
Speedier claim management   Low 
Customer 
Segments 
Shift to personalized offerings   Medium 
Shift to more profitable segments   Medium 
Channels Reliance on partner channels   High 
 
  
48 
 
 
The Impact of the IoT on the Business Model of Insurance 
Companies 
Theoretical 
Findings 
Empirical 
Findings Impact 
BMC Block Factor Mentioned Mentioned 
Value 
Proposition 
Risk reduction   High 
Cost reduction   High 
Increased personalization   Low 
Increased convenience   Medium 
Improved pricing models   Medium 
Improved brand/status   Low 
Reduced negative impact on environment   Low 
In
fr
as
tr
u
ct
u
re
 M
an
ag
em
e
n
t 
Key 
Activities 
Improved risk and pricing models   Medium 
Improved fraud detection   Low 
Move towards non-risk related offerings   Low 
Key 
Resources 
Shift to intellectual & data-driven capabilities   Low 
Human labour with digital competence   Low 
Key 
Partners 
Strategic alliances with IoT-providers   High 
Outsource technical product development   High 
Efforts to improve cooperation   Low 
Figure 6: Comparison of theoretical and empirical findings (own elaboration) 
5.1.1. Financial Aspects 
5.1.1.1. Revenue Streams 
Both the empirical and theoretical findings (EY, 2016; Canaan et al., 2016; Ammiot, 2015) 
confirm a general shift in the insurance industry from fixed menu towards more dynamic 
pricing mechanisms based on usage, rather than the traditional indicators insurers have relied 
on in the past. UBI solutions, such as those applied by Paydrive, Köra Säkert and Enerfy 
allow policyholders to pay an individual premium that is based on their driving behaviour, 
rather than solely on age, geographical location, car model, among others. Yet, these factors 
still play a role in risk and premium calculations, even for UBI solutions. 
However, UBI solutions seem to be more relevant in some insurance lines, rather than others. 
For example, Lindmark mentions that UBI models are not that as relevant for in-home 
insurance, since customers would most likely not experience any noticeable difference in 
their premiums. This might explain why most UBI offerings are found in car insurance, 
49 
 
whereas in-home sensors are mostly implemented in order to prevent damages from 
escalating or taking place. 
During the interviews, some of the respondents questioned the demand for usage-based 
insurance. One of the arguments was that insurance companies already ask policyholders to 
provide an estimation of the usage of their insured objects in order to set an appropriate 
premium. In this sense, insurance premiums have always been partly usage-based. Another 
argument is that the dynamic nature of such solutions creates fluctuations in the premium 
payments from month to month. This might be undesirable for customers that prefer a stable 
premium that covers their insured objects no matter what. For this reason, it is important, as 
Jändel, Alholt and Groth mention, to understand customer insights and do small experiments 
in the real world, to ascertain the demand for a specific insurance solution, before 
implementing it in full-scale. 
Another area in which the empirical and theoretical findings agree is in the creation of new 
offerings. Among others, the IoT provides insurers with more detailed and relevant data that 
allows them to send targeted recommendations to policyholders on how to take better care of 
their insured objects and prevent avoidable damages (Ammiot, 2015; Reifel et al., 2014). 
Moreover, Reifel et al. (2014) state that insurance companies will increasingly move towards 
services not directly related to risk insurance, such as assisting customers with purchase 
decisions. In the PAYD cases, for example, the insurers provide different extra services such 
as a driving journal, a car tracker, parking assistance, and so on, as an added convenience to 
policyholders. Similarly, Lindmark also mentions that the water leakage breakers could be 
used to regulate water consumption in the household, as well as send a repairman at the first 
sign of trouble.  
The way in which customers pay for these value-added services is, according to Lindmark, 
dependent on how the offering is packaged, their willingness to share data and willingness to 
pay. In the case of the PAYD solutions, Paydrive and Enerfy cover the cost of the mobile 
application and sensor lease into the premium, whilst with Köra Säkert, the customer must 
first get a separate subscription from Telia Sense, that provides them with the mobile 
application and car sensor. 
Still, most of the insurance solutions covered in this paper, on top of these value-added 
services, provide some sort of premium discount in exchange for the customer reducing their 
risk, by installing an IoT sensor and sharing their data with the insurer. This premium 
50 
 
discount constitutes a revenue stream decrease, but insurers expect to make up for it by 
reducing their claim frequency. 
Finally, by having access to more relevant and detailed data, the IoT also facilitates up-sale 
opportunities, whereby the insurer can make targeted recommendations to the customer, 
rather than relying on generalized ones. Lindgren and Lindmark both mention that customers 
also want to avoid damages, and if insurers are able to help in this front with appropriate 
recommendations, this will not be experienced negatively. 
In summary, the IoT allows for a shift towards more dynamic pricing mechanisms and usage-
based insurance, rather than relying solely on the static indicators insurers have used for so 
long. There is however a need to understand the demand for such solutions depending on the 
business line and customer base, before implementing it in full-scale. Moreover, new 
offerings are created by moving towards advisory services and other non-risk related 
integrated services. Still, most companies can expect to have to provide a premium discount 
in order to entice customers to share their sensor signals with insurers, thereby reducing the 
risk for damages. 
5.1.1.2. Cost Structure 
With the IoT becoming increasingly cheaper and accessible, insurance companies are now 
able to offer them to customers in conjunction with their insurance policies. This allows for a 
move from restitution to prevention, whereby the IoT sensors are able to detect early 
indicators of future damages and thus reduce the biggest expense for insurance companies: 
claim pay-outs (Reifel et al., 2014; Ammiot, 2015; EY, 2016). 
However, the respondents describe this move towards preventive services as an evolution of 
their prior work, rather than a major innovation, with Wärn and Arnström describing it as part 
of the “DNA” of insurance companies. For example, insurance companies already provide 
smoke alarms together with home insurance and work intensively with Government 
authorities in order to improve building regulations, among others.  
Moreover, Frid and Wärn hesitate to confirm that their IoT solutions have had a positive 
effect on claim frequency and severity, since it hard to isolate and measure the cause and 
effect relationship with other factors influencing the statistics. There is however a logic in 
that customers rewarded for responsible driving behaviour would most likely suffer less 
damages in the long run. At the same time, PAYD solutions inherently attract risk-averse 
customers that are aware they will benefit from individualized pricing. This by itself might 
51 
 
reduce claim frequency and severity in comparison to other traditional car insurances. Still, 
there is an uncertainty whether the added individualization brought by the IoT has a positive 
effect on cost reductions. This topic is further discussed in 5.1.2.2. 
Most of the respondents point to the cost reduction as the major reason for their IoT adoption. 
However, this seems to be highly dependent on the purpose of the IoT solution. Even though 
UBI solutions indirectly lead to cost reduction by rewarding good behaviour, the main value 
proposition is to give the customer a better premium that is based on their individual 
behaviour. Water leakage breakers, on the other hand, were developed in order to detect 
humidity levels and turn off water circulation before damages become irreversible. For this 
reason, the author has chosen to divide the solutions studied in this paper into two major 
groups, UBI solutions and preventive solutions, where the main goal of the former is to 
provide the policyholder with an individualized insurance protection, and the latter is to 
reduce risk for damages. 
An inherent consequence of the reduction of the number of claims is the decreased labour 
costs used to manage them. Moreover, both the empirical and theoretical findings mention 
that the IoT sensors can improve the claim process by having collected data prior to the 
incident, thus making the process speedier and more convenient to the customer as well. 
Even though the IoT is becoming increasingly inexpensive, it will still not come free of cost 
to insurance companies. According to the literature (Ammiot, 2015; EY, 2016; Reifel et al., 
2014), the implementation of the IoT will require short-term investments to reduce internal 
and external resistance, among others. Such efforts include change management, premium 
discounts, among others. Other costs included by the respondents include increased supplier 
procurement and monitoring costs, as well as added distribution and product development 
costs, depending on the relationship with key partners. 
Ultimately, a move towards advisory and more preventive insurance solutions allows insurers 
to move away from a cost-driven BM and their consequent “premium wars”, towards a 
value-driven BM. According to the respondents, the premium is currently a big factor in the 
customer purchase decision, and with customer interaction rarely taking place, insurers have 
a hard time building customer loyalty. Lindmark and Arnström state that this approach is not 
sustainable in the long run and that customers will most likely prefer an insurer that helps 
them in the prevention of damages over the insurer that only provides traditional risk 
insurance. 
52 
 
5.1.2. Customer Interface 
5.1.2.1. Customer Relationships 
Insurance companies aim to improve on customer relationships and loyalty by increasing the 
number and quality of customer interactions. Other than the yearly renewal contract, 
policyholders will get in contact with an insurer when they initiate a claim process. Since the 
aim of insurance is that many share few risks, the average customer makes a claim every four 
to five years. At the same time, claim management is a critical moment in customer 
relationship management, seen as it is the point in which customers test the benefits of their 
subscription to the insurance provider. The IoT offers insurers the opportunity to make the 
claim process speedier and more convenient by collecting the sensor signals prior to the 
incident. 
Both the empirical and theoretical (Ammiot, 2015; Reifel et al., 2014) findings point towards 
a shift towards more advice-led interaction with policyholders, which in its turn allows for a 
more frequent and supportive relationship. The data collected by the IoT provides insurers 
with a basis for targeted recommendations, as well as up-sales, that help the customer prevent 
avoidable damages and their consequences. On this front, UBI solutions seem to have the 
edge, seen as they add a gamification effect on their customers. Since users are able to 
influence their own premiums, they are more prone to adapt their behaviour to avoid losing 
the premium discount. 
The respondents add however that this shift towards more advice-led interaction constitutes 
an extension of their work with prevention, rather than a major innovation. Länsförsäkringar, 
for example, already works with risk-prone customers by giving them advice on how to 
properly take care of their insured objects. According to Lindmark, this effort has had a 
positive effect on customer retention. 
In summary, the IoT allows insurers to increase the number of interactions with their 
customers through increasingly personalized recommendations that influence customer 
behaviour in a way that benefits both parties. Ultimately, this allows for the shift from 
infrequent personal assistance to a more frequent personalized automated service, thus 
improving on customer loyalty and retention. 
5.1.2.2. Customer Segments 
According to the literature, a better understanding of individual customer behaviour allows 
insurers to move from generalized offerings to increasingly personalized ones. Additionally, 
53 
 
the IoT allows insurers to focus on the most profitable insurance groups, whilst leaving less 
profitable ones to competitors (Ammiot, 2015; EY, 2016; Reifel et al., 2014). 
Even though customer interaction will become increasingly individual, the respondents did 
not agree on how the IoT will affect the customer segmentation. As previously mentioned, 
the aim of insurance is to spread the risk among a large group of individuals and at the same 
time, the insurance industry is heavily regulated to ensure that almost anyone can insure 
themselves. For this reason, there is a regulatory and even ethical issue with choosing to 
focus insurance solutions on particular customer segments. 
On the other hand, UBI solutions are tailored for customer segments that benefit from 
individualized premium pricing. So, risk-prone customers are more likely to opt for 
traditional insurance solutions that do not require them to share their data. For this reason, 
UBI solutions inherently attract low-risk customers. 
However, the IoT can also help to improve the behaviour of more risk-prone customers. 
Alholt, Groth and Lindmark mention that insurance companies already work with more risk-
prone policyholders that suffer relatively more damages and mention that this usually has a 
positive effect on customer behaviour, as well as retention. Additionally, Lindmark mentions 
the IoT might remove some of the structural unfairness by allowing insurers to take on new 
risks. 
In conclusion, insurers must pay attention in applying insurance solutions and ensure that 
customer segments that are known to be particularly risk-prone can still find an appropriate 
insurance solution at a reasonable price. 
5.1.2.3. Channels 
Both the empirical and theoretical findings show little change on this BMC block. Insurance 
companies mostly rely on partner channels in order to distribute the IoT sensors. Whilst 
insurance policies are easily and cheaply distributed, the distribution and installation of 
physical objects such as IoT sensors are not the expertise of insurers. 
Moreover, most IoT-centred insurance solutions also rely on partner channels to reach their 
customers with relevant information on their insured objects. Examples include Telia Sense 
or Enerfy’s mobile application. Furthermore, Lindmark emphasizes the importance of 
avoiding spamming the user with generalized recommendations, and the importance of 
confirming that it is indeed the user viewing the sensitive information. 
54 
 
In summary, the distribution, communication and marketing channels will not suffer major 
innovations. Since the distribution of physical objects and development of digital user 
interfaces is not the core competence or focus of the insurance industry, insurers will initially 
have to rely on partner channels to provide the IoT sensors to customers, as well as relevant 
information on their insured objects. 
5.1.3. Value Proposition 
Below, the author discusses the different ways in which the IoT allows insurers to improve on 
their value proposition, thus enhancing the customer experience and standing out in an 
increasingly commoditized market. 
By detecting potential or occurring damages in their earlier stages, the IoT allows for an 
increased focus on the real-time prevention of avoidable damages, thus reducing their size 
and extent. This in turn reduces the risk that policyholders suffer any harm or inconveniences 
as a result of something that could have been detected by say, a water leakage sensor.  
Due to the nature of insurance, a risk reduction is closely related to the cost and premium. By 
reducing the risk of a damage taking place, policyholders avoid costly deductibles and other 
similar value depreciations. Additionally, premiums are calculated based on risk, so the 
higher the risk, the higher the premium, and the other way around. For this reason, a 
reduction of risk calls for a lower premium for the policyholder, which can be seen with the 
premium discounts offered by most of the cases in this study. 
In this area, there is also a difference between the main unique selling points of UBI and 
preventive insurance solutions. On the one hand, PAYD offerings such as Paydrive, Enerfy 
and Köra Säkert focus their communication around an improved and fairer pricing model 
based on usage, rather than conventional static indicators. On the other hand, preventive IoT 
solutions such as water leakage sensors, focus on risk reduction. Still, since risk and premium 
are tightly related, both present the policyholders with the opportunity to reduce their risk and 
subsequent insurance premium. 
Additionally, by shifting towards more frequent advice-led interactions, insurers can 
influence customers’ behaviour in a way that benefits both parties, thus reducing risk even 
further. This move from generalized to increasingly individualized interactions are also an 
improvement in personalization, whereby the customer receives information tailored to their 
particular situation. 
55 
 
Furthermore, as mentioned in the empirical findings, IoT applications are not limited to risk 
insurance. The IoT makes it easier for insurers to develop complimentary and integrated 
value-added services that improve convenience and overall customer experience. Other than 
the addition of individualized advisory services, the sensors can measure resource 
consumption, such as water or fuel, which in its turn allows insurers to provide 
recommendations to the policyholder on how to reduce their consumption and live in a more 
sustainable fashion. Another example of an integrated service found in the PAYD solutions 
include driving journals, car tracker, parking assistance, among others. 
This reduction of the negative impact of damages on the environment is not found in the 
theoretical findings. As Lindmark mentions, the average water damage reparation emits about 
200 kilograms of carbon dioxide. By detecting humidity levels and stopping water 
circulation, IoT sensors can considerably reduce the risk of severe water damages, thus 
reducing their negative impact on the environment. 
In summary, the IoT allows for an increased prevention of avoidable damages. Due to the 
nature of the insurance market, this risk reduction calls for a premium and cost reduction for 
policyholders. At the same time, policyholders experience an increased personalization and 
convenience through more frequent individualized interactions and integrated non-risk 
related services. Although not explicitly mentioned during the interviews, these effects on the 
value proposition, allow first-movers to differentiate themselves in a commoditized market. 
This allows for the shift towards a more value-driven mentioned in 5.1.1.2., and an improved 
brand/status effect on their offerings, by positioning themselves as innovative and responsive 
to customer needs. 
5.1.4. Infrastructure Management 
5.1.4.1. Key Activities 
Analysing and understanding risk is the core competence of insurance companies. According 
to the literature (Ammiot, 2015; Reifel et al., 2014; Canaan et al., 2016), the IoT provides 
insurers with an increased understanding of risk by integrating vast amounts of structured and 
detailed data from the physical world into the segmentation and pricing models that insurers 
use today. 
However, according to the interviewees, insurance companies are currently using the IoT for 
improved pricing models and prevention efforts in the short-term, rather than using it to 
develop risk models that can learn data trends and patterns that allow to forecast damages 
56 
 
before they actually arise. Additionally, another area that respondents did not mention was 
the improvement of fraud detection. This is either because the respondents did not see that the 
IoT allowed for this, or that it was not one of the key concerns among Swedish insurance 
companies. However, the author finds UBI solutions to allow insurers to rely on real mileage, 
rather than trusting on the estimation provided by the customer when they fill out their 
insurance application. 
Other than that, no major changes are observed in the key activities that insurance companies 
perform today. The major difference, according to the respondents, is in the quantity and 
complexity of the data, that needs to be integrated into their current risk models. This is most 
likely because in order to provide IoT-centred insurance solutions, insurers rely to a great 
extent on external IoT providers that manage the hardware and software development part of 
the offering, including the non-risk related services mentioned previously. Meanwhile, 
insurers continue to focus on their competencies: customer relationship management, 
business development and data analysis. The impact of the IoT on the Key Partners will be 
further developed in 5.1.4.3. 
In short, the key activities insurance companies perform will not suffer big changes. 
Currently, the IoT is mostly being used by insurers to improve on their preventive efforts and 
develop more dynamic pricing models. In the short term, insurance companies will most 
likely continue to rely on external IoT providers to develop the technical activities necessary 
to provide these IoT-centred solutions. Accordingly, insurers will continue to focus their 
efforts on their core competencies: risk analysis, pricing models, business development, 
customer relationship management and so on. 
5.1.4.2. Key Resources 
In order to create a sustainable competitive advantage, Ammiot (2015) and EY (2016) both 
recommend that first-movers develop more data-driven capabilities and intellectual property 
that are difficult for imitators to replicate.  
Insurance companies are already working with big amounts of data. Still, technology such as 
the IoT creates extensive and rich data that needs to be integrated into the risk and pricing 
models insurers use today. For this reason, Nyman and Lindmark mention the importance of 
acquiring the right competence in the form of human labour. These competences are mostly 
centred around digitalization but include customer relationship management and business 
development as well. 
57 
 
To sum up, by relying mostly on external partners for the development of the technical 
processes of the supply chain, insurers will not suffer major changes in the key resources 
either. Instead, insurance companies will continue to focus on their key competences. In the 
long run, however, by relying less on external consultants, insurance companies will develop 
the skills needed to adapt to an ever-changing landscape. 
5.1.4.3. Key Partners 
Developing hardware and software is not the core competence of insurance companies. For 
this reason, insurers are largely dependent on third party IoT providers to supply their 
customers with their IoT-centred insurance solutions. Accordingly, all of the companies in 
this study have chosen to partner up with an external company to develop the more technical 
components for the IoT sensors.  
Insurance companies usually enter the picture when the IoT sensor is ready for pilot testing, 
and do not partake in the base R&D. Instead, as previously mentioned, insurers choose to 
focus on analysis of risk and the commercial side, which includes business development, 
customer relationship management, among others. This form of cooperation resembles what 
Osterwalder and Pigneur (2010) define as a strategic alliance, where the two parties work 
together towards a common goal, but without joining equity.  
According to the literature (Ammiot, 2015; Reifel et al., 2014; EY, 2016), in an increasingly 
digitalized world, it is of critical importance to find the right partners able to collect and 
maintain high-value data. At the same time, there is a need to discuss what role IoT providers 
play in the data collection and distribution. According to Lindmark, if customer data is shared 
with third parties for free, this might inadvertently create new competitors.  One solution is 
that IoT providers have more of a distributor role where they do not have the right to manage 
or use the data in any other way than to distribute it to the respective insurance company. 
Another area of interest is how to increase partner commitment and collaboration. One of the 
reasons to enter a joint venture, where the two parties join equity to form a new entity, is that 
both parties have a stake in the business that keeps them committed to making it successful. 
For this reason, it is important that insurers utilize mutual workshops and Key Performance 
Indicators in order to improve communication and create common goals and targets. 
In summary, insurance companies will have to rely on external IoT providers for the 
development and distribution of the IoT sensors. Moreover, in order to boost the productivity 
58 
 
of this strategic alliance, insurers will have to find ways to communicate their expectations by 
using mutual targets and workshops, among others. 
5.2. Managing the Risks with the Implementation of the IoT 
Below, the author summarizes and analyses the major challenges with the implementation of 
the IoT, and the potential solutions to mitigate them provided by the theoretical and 
empirical findings. 
5.2.1. External resistance 
One of the biggest challenges in implementing the IoT is related to data privacy and integrity 
concerns. There are some ethical and regulatory questions that need to be addressed: Who 
owns the data? How can insurers ensure that policyholders feel that they have control over 
their data? How can insurers manage transferring data from one insurer to the next? And, 
where does the data trail end? 
There is not a clear answer to all these questions, but there are steps insurers can take to 
reduce the impact of data privacy and integrity concerns on the adoption of IoT-centred 
insurance solutions. Both the empirical and theoretical findings show that insurers have to 
increase customer willingness to share their data through such incentives as discounted 
pricing, value-added services, customer loyalty points, and so on (Canaan et al., 2016; EY, 
2016; Reifel et al. 2014). In this study, most cases were found to utilize premium discounts to 
reward the customer for reducing the risk in insuring them. 
As mentioned in the empirical findings, insurers cannot allow policyholders to feel like they 
are part of a “Big Brother” community, where they are constantly monitored and controlled. 
Another measure insurers can take to avoid this, is communicating clearly what insurers 
intend to do with the customers’ data, and how they can manage it in a secure manner.  This 
becomes increasingly important depending on the nature of the insurance protection, seen as 
data from water leakage sensors is not as sensitive as the driving behaviour of an individual 
policyholder.  
This is an area in which technology advancements might be of help. For example, Lindgren 
mentions how Talkpool is currently looking into a blockchain solution to allow the safe 
transfer of policyholder data from one insurer to the next. Additionally, Wärn mentions that 
policyholders are required to identify themselves in order to view more sensitive information 
59 
 
on their application. Wärn also mentions that a solution to this is gathering data on an 
aggregate level, so that policyholders do not feel like insurers can track their every move. 
Furthermore, one area not addressed by the literature is the external resistance from other 
parties different from the customers. One such party includes the craftsmen that insurers rely 
on to repair the damages their policyholders incur. By preventing damages from happening, 
the IoT reduces the demand for such work, thus negatively impacting the revenue streams of 
said companies. For this reason, it is important to build strong alliances and work alongside 
companies that see this change as inevitable. 
Another party includes Government authorities, that can truly influence the adoption of IoT-
centred insurance solutions. By working together with other insurance companies, the 
insurance industry might be able to influence regulation in their favour, as well as increase 
the protection provided to the end-customer. An example includes requiring all bathrooms to 
have a water leakage sensor installed, since it can have such positive effects on damage 
reduction. 
In summary, in order to benefit from the risk reduction the IoT brings, insurers have to 
increase customers’ willingness to share their data by ensuring that they benefit from their 
insurance solution. Currently, this is mostly done through premium discounts. Additionally, 
insurers must communicate clearly on how the data is going to be used, shared and managed, 
and how that in its turn benefits the customer. Finally, insurers must also monitor and work 
alongside Government authorities and other affected industries, to make sure that there are no 
unexpected external barriers to the adoption of IoT-centred insurance solutions. 
5.2.2. Organisational resistance 
Digitalization brings along opportunities such as the IoT, as well as threats, including 
servitization, disintermediation of insurers and new types of insurance. And although 
incumbents have an advantage in customer awareness and trust, insurers can still be disrupted 
by new entrants with innovative BMs. The conservative nature of the insurance industry also 
brings along a major challenge in the implementation of the IoT, especially in the case of 
bigger and older companies that might face incumbent inertia. Nonetheless, BMI is 
challenging in of itself for most companies, since it challenges the traditional configuration of 
assets (Chesbrough, 2010). For this reason, insurance companies need to become more 
adaptable to their ever-changing landscape. This section discusses different measures insurers 
can adopt in order to successfully implement the IoT in their BMs. 
60 
 
The IoT has the potential to alter the way in which insurers organise themselves. By reducing 
the number of claims, and potentially automating some its components, some tasks might be 
eliminated or altered.  To avoid creating organisational resistance, insurers must set the right 
corporate culture in place to embrace BMI. As mentioned in the theoretical findings (Canaan 
et al., 2016; Reifel et al., 2014), such efforts include leadership commitment, enterprise-wide 
training and change management. Additionally, the respondents point out the importance of 
finding the right kind of competence to address digitalization and customer experience issues. 
Secondly, insurers need to perform low cost and small-scale experimentation in real life 
customer segments. Mapping tools, such as the BMC, allow insurers to visualize the 
implications and impact of innovations on the different blocks that constitute their BM 
(Chesbrough, 2010). This in its turn allows for first-movers to learn ahead of competitors, as 
well as develop a better understanding of customers’ wants and needs. Seen as not all IoT 
applications will be relevant, experimentation allows insurers to avoid getting caught in 
trends, and instead focus on a problem-solving approach. 
Ultimately, insurance companies must accept that the rate of digitalization is not slowing 
down and that the BMs they have relied on for so long, are not sustainable in the long run. 
For this reason, insurers will need to develop a corporate culture that embraces change and 
BMI, in order to survive in a changing environment. Such initiatives include change 
management, enterprise-wide training, leadership commitment, low-scale and low-cost 
experimentation, BM mapping tools and working intensively with understanding customer 
insights.  
5.2.3. Disruption threats 
By positioning themselves as innovative and responsive to customer needs, first-movers will 
be able to stand out in a commoditized market. However, if first-movers are unable to set 
mechanisms in place that hinder competitors from imitating their BM, this differentiation will 
be short-lived (Teece, 2010). 
In the past, large amounts of risk-related data used to constitute an entry barrier for new 
entrants. Now, the IoT provides insurers as well as other companies, like GAFA (Google, 
Amazon, Facebook, Apple), the ability to gather relevant data at a faster rate and create on-
demand value propositions (Ammiot, 2015). Furthermore, insurers are currently highly 
dependent on external partners for the development of these IoT-centred solutions. Therefore, 
in order to create a sustainable competitive advantage, insurance companies must develop 
61 
 
their own data-driven capabilities and intellectual property that is difficult to replicate 
(Ammiot, 2015; EY, 2016). An example could be moving towards value-adding integrated 
services that are not necessarily risk-related. 
Lastly, a shift from restitution to prevention invokes a sort of self-disruption to certain 
insurance policies. With other words, if water leakage and driving sensors are able to 
effectively reduce the risk of a damage taking place, then this will in turn reduce premium 
prices and negatively affect insurers’ revenue streams. However, by adopting a more flexible 
and innovative corporate culture, insurance companies will be better equipped to handle these 
issues. Moreover, Canaan et al. (2016) recommend that insurance companies shift from 
human error to product liability, or low-frequency, high-severity events that are harder to 
price. Alternatively, Frid recommends that insurers focus on business-to-business insurance 
solutions and start building alliances with manufacturers, in order to decrease the 
disintermediation of insurers.  
In summary, the IoT brings opportunities to insurance companies and new entrants alike. In 
order to create a sustainable competitive advantage, insurance companies need to develop 
their own data-driven and intellectual competencies that hinders imitation and allows them to 
stand out from a commoditized market. Moreover, by reducing the risk and severity of 
individual damages, insurers can expect decreasing revenue streams. Some different 
recommendations to lessen these threats include moving towards business settings, product 
liability, low-risk high-severity events and non-risk related services. 
  
62 
 
6. Conclusion 
The aim of this final section is to summarize the results of the analysis and discussion, thus 
answering the research questions this paper set out to answer. Additionally, the author ends 
with the implications of the research for different stakeholders and provides suggestions for 
future research. 
6.1. Answering the research questions 
This paper set out to study the impact of the IoT on the business model of insurance 
companies, as well as the challenges that its implementation brings, and how these might be 
eliminated. Previous literature has mostly focused on the technical or technological aspects of 
the IoT, or its application on the manufacturing industry. For this reason, this paper provides 
to the academic literature with a novel managerial perspective by looking at the impact of the 
IoT on the business model of a conservative service-provider. Below, the author presents the 
main findings related to the following two research questions: 
 How will the implementation of the IoT affect the Business Model of Insurance 
Companies? 
 How can insurance companies manage the challenges with the implementation of the 
IoT? 
In order to answer these research questions, a multiple case design was chosen where semi-
structured interviews were held with four different insurance companies currently working 
with different IoT-centred insurance solutions. In total, eight interviews were performed, with 
one of them being an IoT-provider, for an added external perspective. The business model 
canvas is the framework used throughout this report to study the effect of different IoT 
applications on the different blocks that constitute a business model. Moreover, business 
model innovation theory is used to study the challenges related to IoT adoption as well as 
their possible solutions.  
6.1.1. Impact of the IoT on the Business Model of Insurance Companies 
In summary, the findings suggest that the IoT has the biggest impact on the value proposition, 
where customers will experience an improved insurance protection, primarily as a result from 
a risk and subsequent cost and premium reduction. By reducing the risk for losses, the IoT 
also has a great effect on the financial aspects. Even though insurers can expect to have to 
provide premium discounts to entice customers to reduce their risks, insurers can also expect 
a reduction in the frequency and severity of claims, which currently stands for about 69 % of 
63 
 
revenues. In the customer interface, the IoT seems able to greatly improve the personalization 
and frequency of customer interactions by providing policyholders with individualized 
recommendations on how to reduce their risk for losses and properly take care of their 
insured objects. In order to achieve this, insurers will have to rely to a great extent on partner 
distribution, communication and marketing channels. Lastly, insurers will also have to rely to 
a great extent on external IoT providers and other key partners for the development and 
production of the sensors and its software, which is why insurers will not suffer bigger 
changes in their key resources or activities. 
6.1.1.1. Financial Aspects 
Starting in the revenue streams, there is a noticeable shift towards dynamic pricing models 
that rely on usage-based data, rather than the static factors insurers have relied on for so long. 
This is, however, not apparent in all business lines such as housing, where there is no real 
demand for such UBI solutions. For this reason, it is important that insurance companies 
implement low-scale and low-cost experiments in real customer segments to assess customer 
demand for any IoT-centred solution and their willingness to share data. Moreover, insurers 
can expect to have to implement some sort of premium discount so that customer are more 
willing to share their data with insurers. Lastly, by moving towards advisory and other non-
risk related services, the IoT allows for the creation of new offerings. The way in which 
insurers capture value from these extra services is dependent on how the offering is 
packaged, by either reducing the severity and frequency of claims on the cost side or offering 
it as an add-on or complementary service with an extra fee on the revenue side. 
In the cost structure, the IoT has the potential to reduce claim frequency and severity on 
several fronts. First though, it is important to differentiate between the purpose of UBI 
solutions, such as PAYD, and preventive solutions, such as water leakage sensors. By 
detecting potential hazards in their earlier stages, preventive solutions can reduce the 
frequency and severity of claims for insurers, whilst UBI solutions mostly focus on providing 
customers with more accurate and dynamic pricing models. However, by rewarding 
responsible behaviour with premium discounts, UBI solutions also have the potential to 
reduce claim severity and frequency. Furthermore, by reducing the number of claims and 
even automating certain parts of the claims process, the IoT also indirectly reduces the labour 
costs related to their management. Other than that, insurers have not taken on many new costs 
other than supplier procurement and monitoring costs, since they mostly rely on IoT 
providers for the development and distribution of the sensors. 
64 
 
Ultimately, the IoT will allow insurers to stand out in a commoditized market and shift from 
a cost-driven BM, where the insurer with the lowest premium wins, to a more value-driven 
BM, where the customer perceives that they receive an improved sense of security from the 
insurance company that positions themselves as innovative and responsive to customer needs. 
6.1.1.2. Customer Interface 
With the help of the IoT, insurers can also view and provide policyholders with real-time 
information on their insured objects, as well as recommendations on how to take better care 
of them and avoid costly damages. This gives insurers the opportunity to improve on 
customer loyalty and retention by increasing the frequency and individualization of customer 
interactions. Ultimately, this allows for a move from rare occasions of personal assistance to 
more frequent dedicated and automated interactions by using, for example, automated push 
notifications with recommendations tailored to the policyholder.  
At the same time, insurers must make sure that even more risk-prone customer segments can 
find appropriate insurance protection at reasonable prices. However, it was found that UBI 
solutions inherently attract low-risk customers that know they will benefit from dynamic 
pricing models. At the same time, the IoT also allows insurers to take on bigger risks and 
actively work with more risk-prone customers to reduce the risk of insuring them, rather than 
solely focusing on more profitable customer segments. 
Lastly, in order to provide customers with the IoT sensors and frequent and dedicated 
automated assistance, insurers will have to rely to a great extent on external partners. This is 
because the distribution of physical objects and development of user interfaces is not the core 
competence of insurance companies. 
6.1.1.3. Value Proposition 
By detecting potential hazards in real-time and influencing customer behaviour through 
rewards and recommendations, the IoT reduces the risk of the customer suffering any harm or 
inconveniences as a result of avoidable damages. This also means a reduction in the risk of a 
customer incurring costly deductibles or similar value depreciations. Moreover, costs are 
further reduced through improved pricing models, that are based on actual usage rather than 
estimations, as well as premium discounts, since the risk of insuring an individual customer is 
decreased.  
  
65 
 
Additionally, the shift towards advisory and other non-risk related services increase the 
convenience and personalization customers experience. For example, resource consumption 
measurement allows customers to live in a more sustainable fashion, both economically and 
environmentally. Ultimately, this will allow for the shift towards a value-driven BM 
previously mentioned, which in its turn will improve the brand/status effects on first-movers’ 
insurance solutions. 
6.1.1.4. Infrastructure Management 
In the short-term, insurers have built strategic alliances with external partners for the 
development of the technical parts of the IoT sensors, whilst continuing to focus on what they 
do best: developing risk and pricing models, business development, customer relationship 
management and so on. Of these key activities, the risk and pricing models have suffered the 
most change by relying on vast amounts of real-time usage-based data, rather than historic 
and socio-demographic risk indicators. Hence, the biggest difference is in the quantity and 
complexity of the data being integrated into their risk and pricing models. In the long term, it 
will be interesting to see if insurers are able to use data to learn patterns and forecast future 
damages before they arise. Currently however, this is hard to implement in a heavily 
regulated industry.  
The key resources are the area in which the author found the least change. In the long term 
however, finding the right competence to address digitalization and customer experience 
issues will be of critical importance to adapt to an ever-changing landscape. 
Lastly, insurance companies have for a long time used strategic alliances in order to help 
policyholders purchase fire extinguishers, smoke and security alarms, and other objects that 
reduce the risk for losses. The same can be observed with IoT-centred solutions, which is 
why there are no major changes in this area either. However, for these alliances to be 
successful, it is important that insurers develop mutual goals and targets, as well as boost 
communication through mutual workshops or cross-functional teams. 
6.1.2. Managing the Risks with the Implementation of the IoT 
This paper identified three major challenges with the implementation of the IoT: external 
resistance from partners and customers, internal organisational resistance and disruption 
threats within and outside of the insurance industry. 
  
66 
 
According to the research, the biggest challenge is increasing customers’ willingness to share 
their data with insurers. For this to happen, customers need to feel that they benefit from IoT-
centred insurance solutions. Other than through marketing the value propositions mentioned 
previously, insurance companies need to reward customers for reducing their risk through 
premium discounts, value-added services, loyalty points, among others. It is also important 
that customers trust the way in which insurers manage and utilize data. For this reason, 
insurers need to clearly communicate how data is collected, transferred and utilized, and how 
that ultimately benefits the end-customer. 
Additionally, insurers should monitor regulation and work alongside Government authorities 
in a way that improves insurance protection for policyholders and decreases risks for insurers. 
Insurers also need to monitor and work alongside other external partners whose businesses 
are affected by the implementation of the IoT, such as the case of craftsmen and repairmen. 
Another major challenge relates to the conservative nature of the insurance industry. In order 
to keep up with an ever-changing landscape, firms will have to develop a corporate culture 
that is more embracing of change and BMI by making use of BM mapping tools, low-cost 
and low-scale real life experimentation, leadership commitment, enterprise-wide training, 
change management and customer insights. 
Eventually, with the help of the IoT, first-movers will be able to stand out in a commoditized 
market and move away from premium wars and towards a more value-driven BM. However, 
the insurance industry is not the only one to benefit from the IoT. New entrants will have an 
easier time to collect vast amounts of risk-related data and create de-modularized and niche 
value propositions. For this reason, insurance companies must develop their own data-driven 
and intellectual property capabilities that are hard for imitators to replicate.  
Lastly, by preventing damages from escalating or happening in the first place, the IoT will 
increasingly remove the risk from the market, thus reducing premium and subsequent 
revenue streams from certain business lines. For this reason, it is again important that 
insurance companies develop a culture that is adaptable, at the same time as they shift their 
attention towards non-risk related services. Moreover, insurance companies may want to 
move towards product liability and other low-frequency, high-severity events that are harder 
to price.  
67 
 
6.2. Implications 
In conclusion, the rate of change is not slowing down, and digitalization is only set to 
increase the long-term growth and profitability issues the insurance industry is facing. By 
allowing for individualized pricing and recommendations, integrated non-risk related services 
and preventing damages before they happen, the IoT provides the opportunity for insurers to 
improve on the customer experience as well as their economics. For first movers, this means 
an improved brand image by positioning themselves as a responsive and innovative company, 
allowing them to stand out in a commoditized market and move away from destructive 
premium wars. 
One important thing to mention is that many of the findings in this paper, related to the 
impact of the IoT on the BM of insurers, are a result of the increased prevention of avoidable 
damages, which could be achieved through other methods than the IoT. However, the 
prevention of avoidable damages is a consequence of the IoT, and therefore included in the 
scope of this paper. Moreover, many of the IoT applications in this study, are dependent on 
parallel technology advancements in other areas such as connectivity, big data management, 
cloud and artificial intelligence, among others. Nevertheless, it is the IoT sensor that allows 
for the collection of data from the physical world by connecting the insured object to the 
internet. 
With most academic literature today being focused on the technical aspects of the IoT or its 
applications on the manufacturing industry, this paper provides a novel managerial 
perspective on the impact of the IoT on the established BM of a conservative service-
provider, such as the insurance industry. Moreover, this study provides insurers and managers 
alike with an improved understanding of the mechanisms used to create and capture value 
with the help of the IoT, as well as potential solutions to mitigate the major challenges in the 
implementation of this digital transformation. 
For IoT providers, this paper provides a business case with which to study the impact of their 
offering on the BM of insurance companies, as well as the implications of its implementation. 
With this knowledge, IoT providers can potentially develop solutions for some of the barriers 
to IoT adoption that insurers and end-customers are experiencing. 
  
68 
 
6.3. Suggestions for Future Research 
This paper investigates the impact of the IoT on the BM of insurance companies. For this 
reason, the paper does not consider the effect of external factors on the implementation of the 
IoT to a great extent. For future research, it would be interesting to see how macroeconomic, 
market and industry factors and trends play a role in the development and implementation of 
the IoT in the insurance industry. As a suggestion, this could be done through a longer 
longitudinal research where interviews are held with Government authorities, customers, 
suppliers, partners, competitors and other parties that play a role in the adoption of the IoT or 
are affected by its implementation.  
Moreover, this paper chose to not consider the technical or technological aspects of the 
implementation of the IoT. However, the author observed that many of the IoT-centred 
solutions studied in this paper, were developed and interplayed with other technologies. For 
this reason, it would be interesting to understand how the development of complementary 
technologies such as 5G, AI, blockchain and the Cloud, among others, have facilitated the 
adoption of the IoT. Conceivably, this could fit into the scope of the first research 
recommendation. 
Lastly, it was found that insurance companies are highly dependent on strategic alliances with 
IoT providers for the development of their digitalized insurance offerings. For this reason, it 
could be helpful to investigate the nature of these alliances, and the common steps and pitfalls 
that dictate the success or failure of such projects. As a suggestion, this could also follow a 
longitudinal design where interviews are held with employees at insurance and partner 
companies, capturing the nature and management of the relationship through time. 
Alternatively, due to limited resources, a case study can also potentially provide valuable 
insights on how insurers can build successful strategic alliances with external IoT providers. 
  
69 
 
7. Bibliography 
Ammiot, E. (2015). The Internet of Things: Disrupting Traditional Business Models. Oliver 
Wyman. 
Baecke, P. & Bocca, L. (2017). The value of vehicle telematics data in insurance risk 
selection processes. Decision Support Systems, 98, 69-79. 
Braun, V. & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research 
in Psychology, 3(2); 77-101 
Bryman, A., & Bell, E. (2015). Business research methods. Oxford: Oxford University Press. 
Canaan, M., Lucker, J. & Spector, B. (2016). Opting in: Using IoT connectivity to drive 
differentiation: The Internet of Things in insurance. Deloitte University Press. 
Chesbrough, H. (2010). Business Model Innovation: Opportunities and Barriers. Long Range 
Planning, 43(2), 354-363. 
Derikx, S., De Reuver, M., & Kroesen, M. (2016). Can privacy concerns for insurance of 
connected cars be compensated? Electronic Markets, 26(1), 73-81. 
Desyllas, P. & Sako, M. (2013). Profiting from business model innovation: Evidence from 
Pay-As-You-Drive auto insurance. Research Policy, 42(1), 101-116. 
Dijkman, R., Sprenkels, B., Peeters, T. & Janssen, A. (2015). Business models for the 
Internet of Things. International Journal of Information Management, 35, 672-678  
EY (2016). The Internet of Things in insurance: Shaping the right strategy, managing the 
biggest risks. EY. 
Folksam (n.d. a). Förebygg vattenskada. Accessed [03/04/2019] at 
https://www.folksam.se/forsakringar/rabatter-och-formaner/forebygg-vattenskada 
Folksam (n.d. b). Köra Säkert – Betala som du kör. Accessed [03/04/2019] at 
https://www.folksam.se/forsakringar/rabatter-och-formaner/kora-sakert 
Greater Than (n.d.). Teknik. Accessed [22/04/2019] at https://greaterthan.eu/teknik/ 
Lambert, F. (2018, 19th May). Tesla is building up in-house insurance as Model S tops list of 
most expensive cars to insure. Elektrek. Accessed [13/03/2019] at 
https://electrek.co/2018/05/29/tesla-insuremytesla-insurance-model-s-most-expensive-car/ 
Lincoln, Y. & Guba, E. (1985) Natural Inquiry. Beverly Hills, CA: Sage. 
Lueth, K. L. (2018, 8th August). State of the IoT 2018: Number of IoT devices now at 7B – 
Market accelerating. IoT Analytics. Accessed [14/03/2019] at https://iot-analytics.com/state-
of-the-iot-update-q1-q2-2018-number-of-iot-devices-now-7b/ 
Markides, C. (2006) Disruptive innovation: in the need for better theory. Journal of Product 
Innovation Management, 23, 1, 19–25. 
70 
 
Metallo, C., Agrifoglio, R., Schiavone, F., & Mueller, J. (2018). Understanding business 
model in the Internet of Things industry. Technological Forecasting and Social Change, 136, 
298-306. 
Nickolas, S. (2019, 28th February). Moral Hazard vs. Adverse Selection: What's the 
Difference? Investopedia. Accessed [13/03/2019] at 
https://www.investopedia.com/ask/answers/042415/what-difference-between-moral-hazard-
and-adverse-selection.asp 
Osterwalder, A. & Pigneur, Y. (2010). Business Model Generation. New Jersey: John Wiley 
& Sons, Inc. 
Paydrive (n.d.). Paydrive – Smart digital bilförsäkring. Accessed [03/04/2019] at 
https://www.paydrive.se/ 
Reifel, J., Pei, A., Bhardwaj, N., Hales, M. & Lala, S. (2014). The Internet of Things: 
Opportunities for Insurers. AT Kearney. 
Ride Safe (n.d.). RideSafe – En smart tjänst mot rattsurfning. Accessed [03/04/2019] at 
https://www.ridesafe.se/ 
Spieth, P., Schneckenberg, D., & Ricart, J. (2014). Business model innovation – state of the 
art and future challenges for the field. R&D Management, 44(3), 237-247. 
Teece, D. (2010). Business Models, Business Strategy and Innovation. Long Range Planning, 
43(2), 172-194. 
Teicher, J. (2018, 7th February). The little-known story of the first IoT device. IBM Blogs. 
Accessed [09/03/2019] at https://www.ibm.com/blogs/industries/little-known-story-first-iot-
device/ 
Yin, R. K. (2009). Case Study Research: Design and Methods. London: Sage Publications 
  
71 
 
8. Appendixes 
Appendix A – Contacting Respondents 
The following message was sent to relevant respondents through E-mail and LinkedIn 
InMail. If the respondent did not reply within a week, the message would be re-sent once 
more. 
Hej XX! 
Jag heter Simon Emanuelsson och studerar nu mitt sista år av min "Masters in Innovation and 
Industrial Management" i Handelshögskolan i Göteborg. Jag skriver för närvarande min 
uppsats om” The Impact of the Internet of Things on the Business Model of Insurance 
Companies”. Jag söker mig till dig eftersom du arbetar med innovation och digitalisering 
inom försäkringsbranschen. 
Jag vill gärna intervjua dig angående vad du upplever är de största utmaningarna som 
försäkringsbranschen står inför som IoT skulle kunna lösa, genom att titta närmare på hur ni 
arbetar med dess implementationen. I korthet, vill jag skapa en bättre förståelse för hur IoT 
påverkar de olika pelarna som utgör en affärsmodell, samt vad du upplever är de största 
utmaningarna med sådana digitala transformationer. 
Skulle du ha tid för en intervju någon gång under de nästkommande veckorna? Skulle du 
trotts allt känna att du inte har tid eller att jag borde prata med någon annan, skulle jag vara 
väldig tacksam om du kunde hänvisa mig vidare. 
Bifogat finner du intervjuguiden med frågorna jag tänkte ställa dig. 
Med vänlig hälsning, 
Simon Emanuelsson 
 
Appendix B – Interview Guide 
Every interview would begin with a presentation of the author and the topic of the paper, 
followed by a visualization of the Business Model Canvas and finally, the interview questions. 
Introduction 
Thank you again for your time. As we discussed earlier, I am currently studying the impact of 
the Internet of Things on the Business Model of insurance companies. So, I would like to ask 
a few questions how the IoT will transform the different blocks that constitute your Business 
Model, with the help of the Business Model Canvas. You can opt to remain anonymous if 
you wish to. Would it be alright with you if I recorded this interview? 
72 
 
 
Background 
1. Name, role in the company? 
2. What is your experience in working with IoT? 
Impact of the IoT on current BM 
1. How will the IoT transform the: 
a. Customer segment 
b. Value proposition 
c. Cost structure 
d. Customer Relationships 
e. Revenue streams 
f. Key activities 
g. Key resources 
h. Key partners 
i. Channels 
Managing the risks with IoT 
1. What are the biggest challenges with implementing the IoT in the insurance industry? 
2. How can insurance companies mitigate these risks/challenges? 
  
73 
 
Appendix C – Empirical Findings 
Below, the reader will find two tables summarizing the themes and subthemes found 
throughout the interviews, as well as the number of respondents that mentioned each 
subtheme. The first table relates to the first research question, the impact of the IoT on the 
BM of insurance companies, whilst the second table relates to the second research question, 
managing the risks with the implementation of the IoT. 
  Times 
mentioned  Themes Subthemes 
Fi
n
an
ci
al
 A
sp
ec
ts
 
Revenue 
Streams 
Shift from subscription to usage-based fees 3 
Shift from fixed menu to dynamic pricing 5 
New offerings/revenue streams 1 
Reduced revenues from premium discounts 6 
Up-sale opportunities 3 
Cost Structure 
Claim frequency and cost reduction 8 
Shift from cost- to value-driven BM 1 
Increased supplier costs 3 
C
u
st
o
m
e
r 
In
te
rf
ac
e 
Customer 
Relationship 
Shift to more frequent interactions 8 
Shift to advice-led interactions 8 
Gamification/loss aversion 2 
Speedier claim management 1 
Customer 
Segments 
Shift to personalized offerings 3 
Shift to more profitable segments 3 
Channels Reliance on partner channels 8 
Value Proposition 
Risk reduction 7 
Cost reduction 7 
Increased personalization 2 
Increased convenience 4 
Improved pricing models 3 
Reduced negative impact on environment 5 
In
fr
as
tr
u
ct
u
re
 
M
an
ag
em
e
n
t Key Activities Improved risk and pricing models 4 
Key Resources Human labour with digital competence 2 
Key Partners 
Strategic alliances with IoT-providers 8 
Outsource technical product development 8 
Efforts to improve cooperation 3 
 
  
74 
 
   Times 
mentioned  Themes Subthemes 
External 
resistance 
Clear communication on data management 4 
Use tech to solve data integrity issues 2 
Monitor and influence regulation 1 
Monitor and influence external partners 1 
Create incentives to share data 6 
Organisational 
resistance 
Setting right corporate culture 3 
Understand customer insights 4 
Low-scale, low-cost experimentation 1 
Finding right competence 2 
Disruption 
Improve customer relationship management 4 
Build new alliances with manufacturers 2 
 
  
75