INOM EXAMENSARBETE INDUSTRIELL EKONOMI, AVANCERAD NIVÅ, 30 HP , STOCKHOLM SVERIGE 2020 Exploring Idea Selection in Innovative Early-Stage Startups A Framework for Analyzing Ideas in a Lean Development Context EMIL FRÖBERG KTH SKOLAN FÖR INDUSTRIELL TEKNIK OCH MANAGEMENT
83
Embed
Exploring Idea Selection in Innovative Early-Stage Startups
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
INOM EXAMENSARBETE INDUSTRIELL EKONOMI,AVANCERAD NIVÅ, 30 HP
, STOCKHOLM SVERIGE 2020
Exploring Idea Selection in Innovative Early-Stage StartupsA Framework for Analyzing Ideas in a Lean Development Context
EMIL FRÖBERG
KTHSKOLAN FÖR INDUSTRIELL TEKNIK OCH MANAGEMENT
Exploring Idea Selection in Innovative Early-Stage Startups
A Framework for Analyzing Ideas in a Lean Development
Context
by
Emil Fröberg
Master of Science Thesis TRITA-ITM-EX 2020:141 KTH Industrial Engineering and Management
Industrial Management SE-100 44 STOCKHOLM
Granskning av Idéurval i Innovativa Startups
Ett Ramverk för att Analysera Idéer i en Lean Development Kontext
Emil Fröberg
Master of Science Thesis TRITA-ITM-EX 2020:141 KTH Industrial Engineering and Management
Industrial Management SE-100 44 STOCKHOLM
Master of Science Thesis TRITA-ITM-EX 2020:141
Exploring Idea Selection in Innovative Early-Stage Startups
A Framework for Analyzing Ideas in a Lean Development Context
Emil Fröberg
Approved
2020-06-02 Examiner
Kristina Nyström Supervisor
Ebba Laurin Commissioner
Hedvig Contact person
Alexander Segerby
Abstract Innovation is a key factor for companies in order to survive in the long-term and to stay competitive. However, it can be challenging to know what ideas (or innovations) to pursue. Large incumbent companies can be slow to innovate, but startups on the other hand, are a great source of innovation. However, many startups fail. Starting a new company is risky, but a countervailing force has emerged called Lean Development. There is a strong need for research and studies on the subject of product development in a Lean Development context and the process of Idea Selection has been somewhat ignored in research. Idea Selection is a part of the Front End of Innovation and is often mentioned as an important factor for success in the innovation process. This study was aimed at creating a framework for Idea Selection, building on the intersection between the Front End of Innovation and the New Product and Process Development in a Lean Development context. The findings are based on current theory on Innovation, Idea Selection and Lean Development as well as a case study conducted at a Swedish insurance startup. The main conclusion of this study is a framework containing four themes: Reach, Impact, Confidence and Effort. By using this framework, startups can more easily decide on what ideas to pursue and thereby move quicker and achieve more success in the innovation process. Key-words Front End of Innovation, Idea Selection, New Product Development, Lean Development, startup
Examensarbete TRITA-ITM-EX 2020:141
Granskning av Idéurval i Innovativa Startups
Ett Ramverk för att Analysera Idéer i en Lean Development Kontext
Emil Fröberg
Godkänt
2020-06-02
Examinator
Kristina Nyström
Handledare
Ebba Laurin Uppdragsgivare
Hedvig Kontaktperson
Alexander Segerby
Sammanfattning För ett företag är innovation en nyckelfaktor för långsiktig överlevnad och för att vara konkurrenskraftig. Det kan emellertid vara utmanande att veta vilka idéer (eller innovationer) som bör utforskas vidare och implementeras. Stora etablerade företag kan vara långsamma att innovera, men nystartade företag, startups, kännetecknas å andra sidan ofta av att ha stor innovationsförmåga. Men många startups misslyckas. Att starta ett nytt företag är riskabelt. För att minska den risken har något som kallas Lean Development växt fram. Det finns ett starkt behov av forskning och studier om produktutveckling i Lean Development-sammanhang och processen att välja ut idéer har ignorerats i forskning. Idéval är en del av Front End of Innovation och nämns ofta som en viktig faktor för framgång i innovationsprocessen. Denna studie syftar till att skapa ett ramverk för Idéval, som bygger på skärningspunkten mellan Front End of Innovation och New Product and Process Development ett Lean Development-sammanhang. Resultat är baserade på aktuell teori om Innovation, Idéval och Lean Development samt en fallstudie genomförd vid en svensk startup inom försäkringsbranschen. Den huvudsakliga slutsatsen i denna studie är ett ramverk som innehåller fyra teman: Reach, Impact, Confidence and Effort. Genom att använda detta ramverk kan nystartade företag lättare bestämma vilka idéer de ska gå vidare med och därmed bli mer snabbrörliga och uppnå mer framgång i innovationsprocessen. Nyckelord Front End of Innovation, Idea Selection, New Product Development, Lean Development, startup
2. Case Company and Industry 11 2.1 Industry for Home Insurance 11 2.2 Insurance Offering 11 2.3 Insurance and Travel-Related Compensation 12
3. Theoretical Framework 13 3.1 Front End of Innovation (FEI) 14
3.1.1 Idea Selection 17 3.1.2 Analysis & Evaluation of Selection Criteria 21
3.2 New Product and Process Development (NPPD) 22 3.2.1 Lean Development 23 3.2.2 The Three Key Principles of Lean Development 26
4. Conceptual Framework 30
5. Research Design and Methodological Choices 34 5.1 Research Design 34
5.1.1 Part 1: Pre-study 35 5.1.2 Part 2: Case study and Expert Interviews 35
5.2 Data Collection 37 5.2.1 Interview data 38 5.2.2 Observations and Secondary Sources 40
5.3 Research Quality 42 5.4 Research Ethics 42
6. Findings & Analysis 44
1
6.1 Components of the Framework for Idea Selection 44 6.1.1 Findings from Expert Interviews 45 6.1.2 The Components of the Final Framework for Idea Selection 50 6.1.3 Implementing the Framework 51
6.2 Applying the Framework for Idea Selection 54 6.2.1 Idea Description 54 6.2.2 Evaluating the Idea using the Framework 55
7. Discussion and Conclusions 60 7.1 The Framework for Idea Selection 60 7.2 Answering the Research Questions 61 7.3 Limitations 62 7.4 Managerial Implications 63 7.5 Suggestions for Future Research 63
8. References 65
Appendix 1 i
Appendix 2 ii
Appendix 3 iii
Appendix 4 iv
Appendix 5 v
Appendix 6 vi
Appendix 7 vii
2
List of Figures
Figure 1: The innovation process in three parts 14
Figure 2: The stages in the FEI 15
Figure 3: The New Concept Development model (NCD) 16
Figure 4: The seven-step new product process 23
Figure 5: Three parts of Lean Development 24
Figure 6: Lean Development in the NPPD 25
Figure 7: The Lean Canvas 27
Figure 8: The Customer Development process 28
Figure 9: Relevant areas for the conceptual framework 30
Figure 10: The context of the conceptual framework for Lean Development
Idea Selection 31
Figure 11: Research design 34
Figure 12: Steps of the case study 36
Figure 13: Community and functionality are important parameters for
creating successful products 48
Figure 14: Framework for Idea Selection 53
Figure 15: Illustration of the Idea Selection Process 54
Figure 16: The Framework for Idea Selection applied to an idea at Hedvig 57
3
List of Tables
Table 1: The key differences of FFE and the product development phase 16
Table 2: A framework for selecting ideas 20
Table 3: Proposed selection criteria 32
Table 4: Weights and importance - the value of the weight signifies the
importance of the criteria 33
Table 5: Empirical material used in this study 38
Table 6: Interview participants 40
Table 7: Categorization of ideas 41
Table 8: Proposed selection criteria, conceptual framework for Idea Selection 44
Table 10: Components of the final framework for Idea Selection 50
Table 11: Weights and importance 54
4
Acknowledgements
Conducting this Master’s thesis has been a challenging, but inspiring journey. This research is the final and concluding step of my Master of Science degree in Industrial Engineering and Management, and it would not have been possible without the help and support from several people. I would like to sincerely thank my supervisor at the Royal Institute of Technology, Ph.D. Ebba Laurin. You have contributed with continuous support and valuable insights to guide me in all steps of the research process. I would also like to express my gratitude to my supervisor at the investigated company; Alexander Segerby. Thank you for offering support, knowledge and guidance. Moreover, I would like to thank my girlfriend, Linnea, for helping me to stay motivated. Lastly, I also want to thank all the interviewees who participated in this study. Without you, this study would not have been possible.
Emil Fröberg Stockholm, June 2020
5
1. Introduction
In the following chapter a background is given, which includes a problematization of the
process for startups to select the best ideas to pursue in a Lean Development context. After
this, the purpose and the main research questions that the report aims to answer are
presented. Lastly, the scientific contribution, sustainability and the delimitations of the study
is introduced.
1.1 Background
A key factor for the survival of a company is innovation (Tidd et al. 1997) in order to stay
competitive and adapt to new market conditions. In today's fast evolving environment
resulting from globalisation and changing technologies, this has become even more essential
and harder (Laurin, Kaulio & Nuur 2017). Startups are a great source of innovation (Weiblen,
2015) and due to the emergence of new technologies such as the internet, opportunities for
new innovative companies have arisen. The emergence of these technologies have also led to
faster product development processes, since software products can be developed and
distributed quickly. Tech startups have changed the dynamics of many industries in later
years. For example, consider what Spotify has done for the music industry, Netflix for the
film industry, AirBnb for the travel industry, Uber for the taxi industry or Amazon for the
retail industry. Tech startups bring new technology to incumbent industries and are changing
the world. Putting this aside, most startups unfortunately fail. It is estimated that 20-30% of
venture backed startups in the US fail, and more recent reports estimates that up to 75% of
venture backed startups fail in the sense that they liquidate their assets (Gage, 2012). Adding
to this, another 95% of venture backed startups do not deliver the projected return on
investment (Blank, 2013). Being that so many startups fail, starting a new company is risky.
However, a countervailing force has emerged - the Lean startup - which can make the process
of starting a company less risky. This product development process and the increasingly
popular entrepreneurial approach has its origin in software development and is known as the
Lean Startup process or Lean Development (Blank, 2013; Eisemann et al., 2012; Mansoori,
2017; Maurya, 2012; Ries, 2011). A cornerstone of the Lean Methodology involves an
iterative approach to testing ideas or hypotheses to learn what works and what does not.
6
This thesis explores innovation processes through studying the case of Hedvig, a swedish
insurance startup. The industry in Sweden for home insurance has not changed significantly
in decades and the market for homeowners insurance is dominated by four incumbent
companies: Folksam, Länsförsäkringar, if and TryggHansa (see Appendix 1). What is more,
many customers in Sweden are dissatisfied with their homeowners insurance. Capgemini
shows that less than 40% of customers are satisfied (Capgemini, 2015). The low satisfaction
in the industry has given opportunity for companies like Hedvig to emerge. Hedvig is a
startup within homeowners insurance. Hedvig aims to change the insurance industry just like
Spotify, Netflix, AirBnb, Uber and Amazon have done for their respective industries.
Accordingly, there are many ideas at Hedvig, regarding for instance new features to
implement, new service offerings and other projects. Hedvig follows a Lean Development
methodology where customer feedback and iterative development is in focus. In general, as
the case for Hedvig, startups are characterized by having limited resources in the form of
capital, time and human resources. Therefore, the importance of prioritizing the right projects
and tasks and to select the best ideas to execute becomes even more crucial, as the resources
need to be used as efficiently as possible.
The process of driving innovation is well explored and many examples of different
approaches used by organizations can be found. For example, the E-commerce giant
Amazon, drives innovation by hiring entrepreneurial and innovative employees (Stone,
2013). To tackle the problem of selecting the best ideas (or innovations) suggested by the
employees, Amazon developed an approach called the “PR/FAQ”. This is a six-page
document consisting of a press release and an FAQ relating to the imagined finished product
(Denning, 2019). The PR/FAQ can be seen as a prototype of a product announcement as if it
was ready to launch. The purpose of this approach is to kill bad, unfinished or unstructured
ideas as quickly as possible, to enable a platform for launching new ideas and to provide a
structured and unified way of communicating ideas. While Lean Development might provide
an answer to how to go from product idea to launch, the methodology does not provide
insights on what to do if too many ideas exist. The PR/FAQ involves the Front End
Innovation (FEI), which is a process that is defined as the period between when an
opportunity is first considered and when an idea is judged to be ready for development. The
7
importance of the FEI in the innovation process has been emphasized by several authors
(Cooper, 1988; Kim & Wilemon, 2002). A central step in FEI is a process called Idea
Selection, which is conducted with the aim of identifying the best idea to pursue. The process
of Idea Selection involves evaluating a set of selection criteria (Rochford, 1991; Bitman et
al., 2008; Ebner, 2009) and is therefore a typical multi-criteria decision-making (MCDM)
problem. This evaluation process was designed with a linear innovation process in mind, as
described by Cooper (1988). After the introduction of Lean Development, the process of Idea
Selection has become less important - as the original idea is now subject to change. It is
unclear whether the methods for Idea Selection as suggested by Rochford (1988), Bitman et
al. (2008) and Ebner (2009) apply to Lean Development. Furthermore, Collis (2016) argues
that an initial strategic screening of ideas can save a startup from going down the wrong path:
one that might be validated in a Lean Development process but is unlikely to support a long
term business.
1.2 Problematization
Until now, launching a new startup has been a hit-or-miss proposition (Blank, 2013). Most
likely, the entrepreneur starting a company will suffer a fatal setback. Again, most startups
fail for reasons such as the idea that make the foundation for starting the company is not good
enough, there are too many ideas and the most promising ideas are not always pursued. Lean
Development may be a countermeasure for startup failure, but Lean Development assumes a
starting point in the form of an idea that already exists. The methodology only works if a
starting idea has been selected and Lean Development does not give any insight on how to
select this idea.
What is more, there is a gap in the current body of research on how startups that implement
Lean Development can benefit from adopting a more formal Idea Selection process. For
example, Collis (2016) states that the usefulness of an initial strategic screening in a Lean
Development process at a startup has not been verified in any studies. Furthermore, Pernstål
et al. (2013) argue that there is a strong need for research and studies on the subject of
product development in a Lean context, and Girotra et al. (2010) suggests that the Idea
Selection process has been ignored in research. Exploratory interviews at the case company
8
Hedvig have shown that while a lot of ideas can be a good thing, it can also be distracting.
Hence, there is a need for a method to select which ideas to be pursued.
1.3 Thesis Purpose
The purpose of this thesis is to explore how startups can make better decisions regarding
Idea Selection, with focus in a Lean Development context. Through an in-depth case study at
the startup Hedvig, a framework for Idea Selection will be constructed and implemented.
Furthermore, this thesis will identify what implications Lean Development have on the Idea
Selection process in a startup.
1.4 Research Question
In order to fulfill the purpose of the thesis, the study vill answer the following research
questions:
Main Research Question (MRQ): How can the Idea Selection process in an early-stage
startup be conducted in a Lean Development context?
A sub-research question (SRQ) has been defined to assist when answering the MRQ:
SRQ: What implications does Lean Development have on the Idea Selection process at an
early-stage startup?
1.5 Contributions
The main contribution from this paper will be an insight on how Idea Selection can be
conducted at a startup in a Lean Development context. By doing this, the report will also
contribute by filling research gaps on this subject. First of all, this report will serve to
increase the research on Lean Development and shed light on the Idea Selection process.
Secondly, Lean Development is a relatively new phenomena and this report is an empirical
case study of this phenomena.
9
1.6 Delimitations
The delimitations of this study are presented below:
● This study will focus on the Idea Selection process and the prioritization of the most
promising ideas
● This study involves only the FEI and NPPD, not commercialization. Evaluation
criteria relating to the long-term financials of the idea will not be included
● The industry is insurance and the in-depth study of one disrupting firm
● This study will only focus on early-stage startups (an early-stage startup is defined as
a startup with less than 100 employees)
● Geography is Sweden (Stockholm)
These delimitations and the effects they have on the thesis will be discussed further in section
7.
1.7 Sustainability Aspects
The United Nations have defined 17 goals for sustainable development. These goals address
the global challenges we face and aim to contribute to achieving a better future for all.
Among other, these goals include Quality Education, Affordable and Clean Energy, as well
as Sustainable Cities and Communities (United Nations, 2020). Even though this study does
not directly relate to sustainable development, the conclusions of this study can be used to
enhance innovation within the field of sustainable development. Innovations are central in
order to tackle the global challenges. Innovation and the development of new technologies
are related to many aspects of sustainability (Weaver, 2000). Accordingly, finding ways to
improve the Idea Selection process can help in pursuing the most promising ideas within
sustainable innovations, hence increasing the chance of success.
10
2. Case Company and Industry
In this section, the industry setting and the case company is presented to provide a
background to the context in which the research took place.. This is aimed at proving a basic
understanding of the home insurance industry, as well as introducing the case company,
Hedvig.
2.1 Industry for Home Insurance
A definition of insurance is given by The Cambridge English Dictionary: “An insurance is an
agreement in which you pay a company money and they pay your costs if you have an
accident, injury, etc.”. In other words, insurance is a protection from financial loss and is
used to hedge the risk of uncertain loss. There are many types of insurances: auto insurance,
health insurance, life insurance and home insurance, among more. Historically, the industry
for home insurance in Sweden has been dominated by four companies. Together, these four
companies have a 96% market share and Folksam, the largest of the four, has a 49% market
share alone. These four insurance companies have their roots in companies all formed around
the year 1900 (see Appendix 2). The market share of the companies dominating the industry
has barely changed in 25 years, nor has the services or products they provide.
2.2 Insurance Offering
Hedvig was founded in late 2017. Since then, the company has acquired 15 000 customers.
The company states that the insurance industry is broken and wants to challenge the
incumbents of the industry. In a speech at the Brilliant Minds conference in 2019, the CEO of
Hedvig argued that the incentive model of the insurance incumbents is structured in a way
that is disadvantageous to the end-customer. The profits of the incumbents directly depend on
how many claims they pay out to the customer, therefore the incumbents have an incentive to
not pay out claims. Hedvig provides insurance through a mobile app that they claim is
“Insurance that is designed to be used”. There are relatively few startups within the insurance
space, this is partly due to the high barriers to entry. To start an insurance company, there are
complex legal regulations that have to be abided by and a partnership with a so-called
reinsurer (i.e. insurer of insurance companies) has to be formed.
11
2.3 Insurance and Travel-Related Compensation
In Sweden, a basic form of travel insurance is included in all home insurances
(Konsumenternas, 2020). There is an additional layer of coverage for people traveling in
Sweden by train and in the EU by airplane. SJ has a clear compensation scheme for how
travellers should be compensated if their train is delayed. Also, in the EU there are certain
rules for compensation for delayed/cancelled flights. The compensation scheme is outlined in
Appendix 3.
Travel compensation management is a large market. Companies in this market are called
claims management companies (CMC) and help travellers make claims for compensation
related to travel, often to airlines. Every year travellers are missing out on 6B USD in
unclaimed compensation and 85% of travellers in the EU are not familiar with their rights
when it comes to compensation for air travel (AirHelp, 2018). There is an opportunity for
Hedvig within this field. Hedvig can provide a valuable service to their customers if they
assist their customers in taking part of the compensation guaranteed by SJ and EU. This way,
Hedvig would act similar to a CMC. Because of COVID-19, travel has declined. This is
because of restrictions and the risk of infection. In March 2020, the number of flights
declined 40% compared to the same period last year (Flightradar, 2020). This affects this
study because travel compensation may become very relevant now that many trips are
cancelled.
12
3. Theoretical Framework
The Theoretical Framework consists of two main building blocks and serves as a foundation
for the theories and models which are used in this study. First, the Front end of Innovation is
introduced and methods for Idea Screening and Selection are presented. These methods are
used to evaluate ideas and prioritize between ideas to find the best idea to pursue. Secondly,
the theory behind New Product and Process Development, as well as Lean Development are
presented. These two fields of research will be connected in the Conceptual Framework in
the next chapter.
The theoretical framework used in this study consists of two parts: (1) Front End of
Innovation (FEI) (Kim & Wilemon, 2002; Cooper, 1988; Koen et al., 2001; Kahn 2007) and
(2) New Product and Process Development (NPPD) (Cooper, 1990; Kotler, 2006). However,
the entire innovation process can be described to consist of three parts: FEI, NPPD and
Commercialization, as illustrated in Figure 1 (Koen et al., 2001). Nonetheless, the last part,
Commercialization, falls outside the scope of this report. It is important to study the FEI to
understand the dynamics between when an idea is first born to when an idea is ready to enter
the NPPD stage. The second part of the theoretical framework covers NPPD and focuses on
Lean Development (Blank, 2013; Ries, 2013; Maurya, 2013; Girgenti, 2016); and more
specifically the initial stages of Lean Development. These two main areas (FEI and NPPD)
are included to build a theoretical background of the Idea Selection process, but also to
understand how this process works in a Lean Development context. In order to answer the
research question, the area of NPPD needs to be considered, as Lean Development involves
an iterative process that can redefine the idea. Therefore, the Idea Selection process is
dependent on NPPD.
13
Figure 1: The innovation process in three parts (Koen et al. 2001, pg. 51).
3.1 Front End of Innovation (FEI)
Innovation is a key factor in the survival of a company, as previously mentioned (Tidd et al.,
1997). The FEI, also known as the Fuzzy Front-End (FFE) of innovation, is a process that has
been defined as: “The period between when an opportunity is first considered and when an
idea is judged ready for development” (Kim & Wilemon, 2002, pg. 269). An idea in the
context of this report is a new product idea. This is defined as an idea for a new product or
service that is new to the firm developing it. A product is defined by Kotler (2006) as: “An
object or system made available for consumer use; it is anything that can be offered to a
market to satisfy the desire or need of a customer”. This definition can therefore include
improvements to an existing product or features since this can lead to the product satisfying
another need, desire or customer.
The importance of the FEI is emphasized in literature (Cooper, 1988). The process presents
the greatest opportunities for improving the overall innovation process (Koen et al., 2001).
Kahn (2007) argues that the FEI consists of three tasks: Strategic Planning, Idea Generation,
and Idea Selection (or evaluation as Kahn (2007) refers to it). In this report, focus lies on the
last step in the FEI: evaluation, i.e. the Idea Selection process. The other steps in the FEI are
not within the scope of this report.
14
Figure 2: The stages in the FEI. Adapted from Kahn (2007).
The FEI starts with Idea Genesis and Opportunity Identification as illustrated by the inwards
pointing arrows in Figure 2 above. The purpose of Idea Selection is to filter out the most
promising idea to pursue. The New Concept Development model (NCD) was introduced by
Koen et al. (2001) and explains the influencing factors in the FEI as illustrated in Figure 3.
The end goal of the FEI is a blueprint or a plan. This blueprint is then used to enter NPPD.
While consensus for the definition of FEI has not been reached and a dominant framework
has not been developed so far (Hüsig & Kohn, 2003), the NCD serves to build an
understanding of the FEI. Koen et al. (2001, pg. 47) describes the NCD in the following way:
“provides a common language and definition of the key components of the Front End of
Innovation”. The NCD model consists of five elements: Idea Genesis, Idea Selection,
Concept & Technology Development, Opportunity Identification and Opportunity Analysis.
The “engine” is driven by these five components and it represents the management of the
firm. The outer edge of the model represents the influencing factors (for example
organizational capabilities or business strategy) that affect the decision of the inner parts.
15
Figure 3: The New Concept Development model (NCD). Koen et al. (2001, p. 47)
Managing the FEI is important to create new product success (ibid). Moreover, the FEI is a
place to find low cost opportunities to achieve large improvements in time to market
(Reinertsen, 1998). Kim & Wilemon (2002) define a couple of significant differences
between the FEI and the product development phase (i.e. NPPD). These differences are
highlighted in Table 1 below. It is important to note that the end result of the FEI is a
blueprint or a plan for a product while the end result of the NPPD is a finished product.
Factor Characteristics of FEI Characteristics of the Product Development phase
State of an idea Can easily change, fuzzy Specific, defined in detail
Outcome / end result A blueprint A product
Ease of rejecting the idea Low High
Budget Small/None Large, designated
Table 1: The key differences of FFE and the product development phase. Adapted from Kim
& Wilemon (2002).
16
3.1.1 Idea Selection
Idea selection is the process of finding the best idea to pursue from a set of ideas. Hamilton
(1974, pg. 16) states that the purpose of Idea Selection (also known as Idea Screening) is to:
“Select from a large list of ideas the few that warrant extensive and expensive analysis”
The area of Idea Selection has been studied by multiple researchers (Hamilton, 1974;
Rochford, 1991; Bitman & Sharif, 2008; Ebner, 2009), but the application of current
approach has not been studied in a Lean Development context. This stage in the FEI that is
Idea Selection is conducted after Idea Generation and is the third and final stage in the FEI
(Kahn, 2007). The Idea Selection process is part of the Opportunity Identification process.
This process consists of two basic steps: the identification of new product ideas as well as the
screening and evaluation of these ideas (Rochford, 1991). The selection process can be
conducted by considering a set of criteria. Several authors, such as Rochford (1991), Bitman
& Sharif (2008) and Ebner (2009) have suggested criteria that can be used to evaluate ideas
for new products and R&D projects. These criteria are presented in their entirety in this
section and will be further narrowed down in the conceptual framework (the next chapter)
where these selection criteria are used to create a conceptual framework.
Bitman & Sharif (2008) conducted a study to develop a framework for R&D project ranking
with the purpose of helping firms improve their competitive advantage. The authors
conducted a literature review comparing previous multiple perspective systems for R&D
project ranking. The study is concluded by presenting a project evaluation form that can be
used to evaluate projects (see Appendix 4). The form contains five perspectives with two to
five criteria in each perspective:
17
(1) Reasonableness (a) Tools needed to perform this project (b) Skills needed for the tools needed for this project (c) Facts needed to perform this project (d) Methods to perform and manage this project (e) Subcontracting needed to perform this project
(2) Attractiveness (a) Strategic fit of this project within the firm (b) Influencing actors (c) Track record of submitter of this project
(3) Responsiveness (a) Ethics of this project (b) Morality (c) Ecological implications of performing this project (d) Regulatory constraints
(4) Competitiveness (a) Capability (b) Competency
(5) Innovativeness (a) This projects improvements to technological dimensions (b) Novelty of this project (c) Research lifecycle phase (d) Idea source
Some critique has been directed towards studies of opportunity identification. Rochford
(1991) suggests that opportunity identification, as described in previous literature may be
incomplete. She suggests a multi-stage screening process in two stages. The first step
involves the initial screening criteria and is a simple, preliminary screening of ideas, which
can be achieved by asking the questions about the project, stated under Stage 1 below:
Stage 1: Initial screening criteria (1) Consistent with company objectives? (2) Do-able?
The second step is a more extensive process, where more aspects are taken into consideration.
The different parts of the secondary screening is presented below:
18
Stage 2: Secondary screening criteria
(3) Market (a) Size (current and potential) (b) Growth (current and potential) (c) Appeal (d) Role for the company
(9) Other (a) Gut feel (b) Is it realistic? (c) Probability of success
Rochford (1991) further states that the most important reason for having a multi-stage
screening process is to minimize the cost (in terms of time and information required) of
evaluating the ideas. By having a more strict pre-screening (i.e. initial screening) process,
many ideas can be filtered out at a very low cost and the ideas that pass the pre-screen will be
19
subjected to the more rigorous next parts of the screening process. Furthermore, the author
suggests that criteria is to be weighted on need to have and nice to have. The former refers to
the requirements the project must meet whereas the latter refers to “want”-objectives, i.e.
requirements that the projects necessarily do not need to meet, but would be beneficial.
Looking further on frameworks for Idea Selection, Ebner et al. (2009) conducted a study with
the aim of using crowdsourcing to develop a framework for community based innovation
development. The framework was developed to generate innovations, process and product
ideas by utilizing the wisdom of the crowd. An important part of the framework is detailed in
Table 2 below.
Table 2: A framework for selecting ideas. Adapted from Ebner et al. (2009).
Although the wisdom of the crowd can be powerful, Girotra et al. (2010) finds that this may
not be the case when it comes to Idea Selection. There are four factors underlying the
performance of the idea generation and Idea Selection process (Girotra et al., 2010). These
four factors are: (i) the number of ideas generated, (ii) the average quality of the ideas
generated, (iii) the variance in quality of the ideas generated, and (iv) the ability of the
selection process to discern the best idea. The effectiveness of two group structures that
require creative idea generation followed by selection were compared. A team work group
structure, where the participants work as a team in the same time and space was compared to
a hybrid structure where the participants work individually for some fraction of the time and
20
then work together. The authors found that the hybrid structure was superior with regards to
all four factors except the ability of the selection process to discern the best idea. The hybrid
approach proved to be advantageous when it comes to identifying the best ideas, but when it
comes to assessing the quality of ideas, both structures proved to be weak.
The three Idea Selection methods presented by Rochford (1991), Bitman & Sharif (2008) and
Ebner et al. (2009) contain some differences, but also similarities. These three methods will
be synthesized and further analysed in the conceptual framework in chapter 4.
3.1.2 Analysis & Evaluation of Selection Criteria
The screening criteria can be used to analyse ideas from different perspectives. However,
some perspectives may be more relevant than others. For example, in a startup, it might be
more relevant to consider the time needed to develop the idea, as well as the customer benefit
and less relevant to initially consider the profitability. In other words, there is a need to
prioritize between the selection criteria. This chapter will elaborate on the subject of how to
prioritize selection criteria and present quantitative methods for how ideas can be selected
using selection criteria.
Multiple-Criteria Decision Analysis (MCDA) is a sub-discipline of operations research. The
purpose of MCDA is to evaluate multiple conflicting criteria in a decision process. In the case
of this study, models that utilize MCDA will be studied to gain an understanding of how
screening criteria can be utilized to evaluate a project. Screening models are used to evaluate
projects and find the best project to pursue. Poh et al. (2001) states that the quality of the
result is not only dependent on the project itself, but also on the method of evaluation.
Furthermore, the authors have conducted a comparative analysis of project evaluation
methods. They find that the evaluation models can be classified into two broad categories:
weighting and ranking models as well as benefit-contribution models. The most common
types of weighting and ranking methods are the Weighted Scoring Model (WSM) and the
Analytic Hierarchy Process (AHP). Common methods of benefit-contribution are economic
analysis as well as cost-benefit analysis. When comparing models in the two different
categories, Poh et al. (2001) find that the WSM and AHP are the best with regards to three
categories: simplicity, availability of data needed and adaptivity. Being that this study focuses
21
on idea evaluation at an early stage, all three categories are highly important. Therefore, only
weighting and ranking method WSM. Furthermore, it has been shown that focusing too much
of the financial aspects of a project causes the project not to perform well over time (Cooper,
1999) which is a reason not to focus on benefit-contribution models.
Weighted Scoring Model (WSM)
The WSM builds on the idea that the evaluator rates the quality of an idea by assigning it a
score. The score should reflect how well the project meets the defined objectives on some
screening criteria. The WSM functions as a mathematical formula where each screening
criteria is assessed by weight. The purpose of the weight is to reflect the relative importance
of each screening criteria.
Consider n alternatives for ideas {A1, A2, ... ,An}. These ideas have m number of selection
criteria {C1, C2,...,Cm} and weights {W1, W2,...,Wm}. The performance, P, of each idea in
terms of each selection criteria has to be evaluated. This forms a decision matrix {Pij} where
Pij is a score that measures how well idea Ai performs on selection criteria Cj. Jadhav (2009)
presents a mathematical formula to calculate the score for idea Ai:
S(Ai) = ∑Wj*Pij
Where the sum is over j=1,2,..,n. The ideas can be evaluated by ranking S in descending
order.
A similar method to the WSM is the ranking model. In the ranking model, the evaluators
simply order all ideas according to their perceived qualities. A recent study by Cui et al.
(2019) suggests that the WSM strictly outperforms the ranking model in terms of the
likelihood of selecting the highest-quality ideas.
3.2 New Product and Process Development (NPPD)
When the most promising idea is selected, the next step is to turn that idea into reality. The
process of turning ideas into products is called Product Development and is a part of NPPD.
22
In 1988, Cooper defined a linear process for Product Development as shown in Figure 4.
Cooper (1988) called this Product Development process the seven-step new product process.
The steps in the process are pivotal stages in which GO/KILL decisions are taken in the first
three stages of the Product Development process, (i.e. at the idea stage, preliminary
assessment and concept stage). This idea is similar to the stage-gate model that Cooper
developed a couple of years later (Cooper, 1990). During the first three steps of the new
product process, Cooper (1988) considers two main perspectives: the market and the
technical perspective. In other words, the first steps evolve around considerations about if the
project or idea is technically possible to execute and if there is a market need for the product.
Figure 4: The seven-step new product process. Adapted from Cooper (1988).
3.2.1 Lean Development
Lean Development has roots in lean production, which originates from a study of Toyota’s
production system (Womack, 1997). The purpose of Lean Development is to incorporate
continuous customer feedback into the NPPD. The first step in Product Development
originates in an idea, but Lean Development does not include Idea Selection methods or any
other methods for how the ideas should be chosen, this lies within the scope of FEI. The
methods suggested by Rochford (1991), Bitman et al. (2008) and Ebner et al. (2009) were
developed with linear NPPD in mind such as the seven-step new product process. It is unclear
whether the methods of evaluating the ideas by using a set of criteria as suggested by
Rochford (1991), Bitman et al. (2008) and Ebner et al. (2009) apply to Lean Development.
Lean Development is also known as Lean Startup because the methodology has its roots in a
startup context. Ries (2011) defines a startup as “a human institution designed to deliver a
23
new product or service under conditions of extreme uncertainty”. In this thesis, a startup has
been defined as a company working on a product or service that is new in some aspect. This
newness can for example be in the form of the product itself or in the distribution of the
product. A startup is not the same thing as an incumbent company in small-scale, usually a
startup doing something different than the incumbents of the market.
The concept of Lean Development has been defined by taking inspiration from York et al.
(2014, pg 21) as:
“An approach to entrepreneurial and innovative activities that emphasizes placing resources
into the creation of customer value, viewing all other activity as waste until a fit is found
between the product and the intended market“
In contrast to the linear seven-step new product process, iterative Product Development
Processes like Lean Development can be utilized to avoid failure in innovation processes. As
research shows, 75% of all startups fail (Blank, 2013) This method can make the process of
starting a company (and launching new products) less risky (Blank, 2013). Lean
Development can also be applied to ideas and projects within companies. Lean Development
can be illustrated as an iterative process with three parts: (1) building a product, (2)
measuring and collecting feedback and (3) learning from the feedback (Ries, 2013). This
process is illustrated in Figure 5 below.
Figure 5: Three parts of Lean Development. Adapted from Ries (2013).
It is impossible to know beforehand whether a new product idea will be successful or not.
Therefore, it can be argued that it is unnecessary to spend too many resources on analysing if
24
the idea will be successful, given that the cost of testing a minimal version of the idea is low
waste. If this is the case, instead of conducting rigorous analysis, a simple version of the
product can be built and tested to get real feedback from the market. Collis (2016) argues for
conducting a relatively simple analysis to understand which ideas should be tested by
building a minimal version of the product according to the principles of Lean Development.
The Lean Development approach aims to reduce risk in new business development by
replacing the traditional business plan with a list of hypotheses to be verified and swapping
the entrepreneurs assumptions with customer feedback (Girgenti, 2016). Girgenti (2016)
articulates the value of customer satisfaction and value generation. The authors propose a
novel approach to systematically build a customer development model with the purpose of
verifying that what is offered meets the customer need. The model builds on generating an
idea, developing a Minimum Viable Product (MVP) and then gathering feedback analysis
and pivoting. A MVP is a very simple version of the product that can satisfy a customer's
needs. This process can be applied within the NPPD as illustrated in Figure 6 beneath.
Figure 6: Lean Development in the NPPD. Adapted from Girgenti (2016).
Clement (2019) discusses the need for a MVP and the $100-million mistake
Excelsior-Henderson motorcycles made by not applying a Lean Development process and
thereby allowing a small mistake to become a mistake of catastrophic proportions. The
company was founded based on the premise that a need exists for Harley-Davidson-like
motorcycles. This assumption of the demand was based largely on the success of market
leader Harley-Davidson and their inability to meet the high customer demand.
Excelsior-Henderson never developed a MVP to verify the high customer demand. Blank
(2013) explains this way of starting a business as doing it in stealth-mode. This refers to the
process of developing a business plan, raising capital based on the assumption that the
business model will be executable and then try to execute the business model without
25
customer or market feedback. If Excelsior-Henderson had developed a MVP to test the
hypothesis of high market demand for Harley-Davidson-like motorcycles, they might have
discovered that there was no market demand for Harley-Davidson-like motorcycles but
instead a high demand for Harley-Davidson motorcycles.
One reason why so many startups fail can be that they operate in stealth-mode. During the
dot-com boom, many companies operated in stealth-mode. This meant that the product
development process was very secretive to avoid alerting potential competitors to a market
opportunity. Prototypes would only be tested during carefully orchestrated “beta” tests. The
Lean Development approach makes the concept of stealth-mode obsolete because in most
industries, customer feedback matters more than secrecy and constant feedback yields better
results than keeping the product unknown to competitors (Blank, 2013). The Lean
Development approach is to test hypotheses iteratively and fail quickly.
3.2.2 The Three Key Principles of Lean Development
Blank (2013) states that there are three key principles of the Lean Development method: (1)
The Business Model Canvas (and the Lean Canvas), (2) Customer Development and (3)
Agile Development. These principles are presented below.
The Business Model Canvas and the Lean Canvas
Rather than spending months of planning and researching, i.e. starting a business in
stealth-mode, entrepreneurs should embrace the fact that all they have when starting out is a
set of untested hypotheses about customers, a market and products. Instead of spending an
excessive amount of time on writing a detailed business plan, Blank (2013) suggests that the
entrepreneur should summarize their thoughts in a framework called the business model
canvas (BMC). Osterwalder & Pigneur (2010) introduced the concept and it outlines nine
building blocks; customer segments, value propositions, channels, customer relationships,
revenue streams, key resources, key activities, key partnerships and cost structure. The
authors state that a business model can be described through these nine basic building blocks
and these can also show how a company intends to make money. Moreover, Osterwalder &
Pigneur (2010) define a business model as a model that describes the rationale of how an
organization creates, delivers, and captures value.
26
Blank (2013) argues that business plans rarely survive first contact with the customer.
Therefore spending too much time on building a detailed business plan and planning for
events that will never occur is a waste and one purpose of Lean Development is to eliminate
waste. In the case of the Lean Startup, this means not wasting time on making a detailed
business plan, because the business plan will likely change before any of the planned-for
events occur. As the boxer Mike Tyson once said about his opponents’ pre fight strategies:
“Everybody has a plan until they get punched in the mouth.”
-Mike Tyson
From the Business Model Canvas, a similar model called the Lean Canvas has emerged.
Maurya (2012) has adapted the BMC to visualize an hypothesis in a startup context. The
Lean Canvas is illustrated in Figure 7 beneath.The differences between the BMC and the
Lean Canvas is the focus on the Problem instead of Key Partners, Solution rather than Key
Activities, Key Metrics instead of Key Resources and the perspective of an Unfair Advantage
rather than Customer relationships. The reasons for the shift in focus on some perspectives is
to adopt the BMC to a startup setting. The Lean Canvas main purpose is to evaluate business
ideas, not necessarily product ideas.
Figure 7: The Lean Canvas. Adopted from Maurya (2012).
27
Customer Development
The search of a business model that works is called Customer Development (Blank, 2013).
Customer discovery is related to this and refers to the process of capturing the vision of the
entrepreneur and turning this into a hypothesis that can then be proven, or disproven - in
which case a new hypothesis if formed and tested. The process of customer validation tests if
the business model is repeatable and scalable. If not, the startup pivots and goes back to the
process of customer discovery. Once a business model is validated, the startup can exit the
search phase and continue building and developing the idea by moving on to the execution
phase. Customer creation is the process of building end-user demand and scaling the
business. Lastly, company building is the process of transitioning the business from a startup
to a company focused on executing the validated model (Ries, 2008). This process is
illustrated in Figure 8.
Figure 8: The Customer Development process. Adapted from Blank (2013).
The process of hypothesis valuation can be defined by the key concepts of validated learning,
minimum viable product (MVP) and the strategy of “getting out of the building” (Batova et
al., 2016). Validated learning is the process of learning what the end customer really wants.
The purpose of this is to avoid wasting time and money by planning, developing and
marketing a product that the end customer does not actually care about. This process is
similar to what Blank (2013) describes as customer discovery. Another key concept of
hypothesis valuation is the MVP. Batova et al. (2016) defines a MVP as a product or service
to be developed that has the smallest number of features for which the customer would be
willing to pay. A MVP is created by minimum effort and is rarely perfect. The purpose of
28
doing this is to test whether a customer wants or needs a service. Lastly, a cornerstone of a
lean startup is to “get out of the building”. The purpose of this is to counteract the
confirmation bias, i.e. the tendency to search for, interpret, favor, and recall information in a
way that confirms or strengthens one's prior personal beliefs or hypotheses (Plous, 1993). The
confirmation bias can be counteracted by talking to potential customers; getting out of the
building and talking to real people. The goal is to get feedback in the initial stages of the
process to avoid spending resources on creating a product that no one wants (Batova et al.,
2016).
Agile Development
According to Qumer and Henderson-Sellers (2006), agility is defined as the ability to
accommodate and adapt to changes in a dynamic environment. Jyothi and Rao (2012) states
that being agile is to apply previous knowledge while learning from current experience in
order to deliver high-quality products, under budget constraints and in short time frames.
Agile development has its roots in the software industry. The Agile Manifesto is a document
authored by Fowler et al. (2001) that outlines the four core values of agile development:
● Individuals and interactions over processes and tools
● Working software over comprehensive documentation
● Customer collaboration over contract negotiation
● Responding to change over following a plan
These four values can be adapted to work with the theory of the lean startup and works
hand-in-hand with customer development (Blank, 2013). This process is illustrated in
Appendix 5. Even though the four principles were developed to improve the software
development process, the principles can be translated to a more general sense. For example,
working software over comprehensive documentation can mean that it is more important to
have a working product than an in detail instruction manual. By developing the product
iteratively and incrementally, agile development eliminates wasted time and resources instead
of having long product development cycles that are more reminiscent of a company operating
in stealth-mode.
29
4. Conceptual Framework
For the sake of answering the research question, a conceptual framework is derived by
combining theories of Idea Selection in the FEI and the theory of Lean Development from the
NPPD. The end result of the conceptual framework is a model in the form of a framework
that is applicable for Idea Selection. Arguments for why the components of the framework
are relevant are also given.
To answer the MRQ, a method to evaluate ideas in a Lean Development context is needed.
As presented in the previous section, several authors have suggested methods to enhance the
Idea Selection process (Rochford, 1991; Bitman et al, 2008; Ebner, 2009) - but these methods
are not adapted to a Lean Development context. Following this, the purpose of this
conceptual framework is to adapt these methods to a Lean Development context. An
illustration of the scope of the study is given in Figure 9 beneath.
Figure 9: Relevant areas for the conceptual framework
The Idea Selection process is different in a linear context compared to in a Lean Development
context; mainly because in a linear context, the process of Idea Selection and evaluation is a
one time, non-recurring event. On the other hand, in a Lean Development context the process
of Idea Selection and evaluation is a recurring event. The context of the conceptual
framework that is proposed by the researcher is presented in Figure 9 below. This framework
consists of two main parts: (1) a pre-screening of ideas to reduce the size of the initial pool of
ideas and (2) an Idea Selection process adapted for a Lean Development context.
30
Figure 10: The context of the conceptual framework for Lean Development Idea Selection
Step (1): As illustrated in Figure 10, the conceptual framework starts with a Large Pool of
Ideas. These ideas are then subjected to a pre-screening. In this step of the process, a
multistage selection process as proposed by Rochford (1991) can be utilized to make the
selection more efficient. This process serves to quickly sort out ideas that should not be
pursued. The pre-screening consist of asking three basic questions:
(1) Is this idea consistent with company objectives?
(2) Will this idea serve to increase revenue?
(3) Is the idea do-able?
Question (1) have been formulated with inspiration from Rochford (1991) and Bitman &
Sharif (2008), who both argue that an important factor to consider is the strategic fit to the
firm of the proposed project. The second question was inspired by the research of Ebner et al.
(2009). The author argues that an important dimension for evaluation is market potential. The
dimension of market potential can be captured in the potential revenues. Lastly, the third
question was inspired by Rochford (1991).
Step (2): The Idea Selection process is conducted by evaluating a set of criteria. These criteria
are presented in Table 3 below and should be evaluated on a scale from 1-5, where 1
corresponds to completely false and 5 corresponds to completely true. The selection criteria
presented by Rochford (1991), Bitman et al., (2008) and Ebner (2009) display four common
themes: Market potential, Impact, Probability of success and Effort. These themes are similar
to the RICE-framework proposed by (McBride, 2016).
31
Theme Criteria Source
Reach Market size
Rochford (1991), Ebner et al. (2009)
Potential for market growth Rochford (1991)
Impact Customer benefit Ebner et al. (2009)
Competitiveness Bitman & Sharif (2008)
Confidence Profitability of success
Rochford (1991), Ebner et al. (2009)
Personal experience Bitman & Sharif (2008)
Effort Time and cost of execution Adapted from Rochford (1991) Table 3: Proposed selection criteria
Reach: This theme describes the number of customers included in the potential users for the
new product. The potential future growth of the market is also included. Rochford (1991) and
Ebner et al. (2009) bring attention to the criteria of a large market size and Rochford (1991)
also includes the criteria of an attractive future market.
Impact: The second theme describes what impact the new product will have on the intended
customer. The Impact can be measured in the perceived customer value and the degree of
how advantageous the new product is compared to alternatives (competitiveness). Bitman &
Sharif (2008) argues that an important criteria for Idea Selection is the benefit of increased
competitiveness. In other words, the question regarding whether the new product will have an
impact on the customer that is bigger than the impact that the competitors products have
should be considered. Moreover, the criteria of high customer benefit as suggested by Ebner
et al. (2009) has been included because it reflects the impact that the product will have on the
customer.
Confidence: Thirdly, the Confidence perspective serves to reflect the probability of success
if the product is developed. An iterative Lean Development process allows for continuous
learning and therefore the Confidence in the idea can be increased when assumptions about
Reach, Impact and Effort are verified. The confidence is reflected by the probability of
success, which is a selection criteria proposed by Rochford (1991) and Ebner et al. (2009).
32
Also, Bitman & Sharif (2008) suggest that a selection criteria should take into account the
track record of the submitter of the project. This criteria also serves to reflect the confidence.
Effort: Lastly, the Effort perspective describes the amount of work that has to be done to
develop the product. The effort is measured in both the cost of execution and the time needed
for execution as suggested by Rochford (1991).
To evaluate these criteria in a quantitative manner, a study conducted by Poh et al. (2001)
suggests that the WSM is the best when it comes to three categories: simplicity, availability
of data needed and adaptivity. Therefore, this model has been chosen to prioritize the ideas.
By letting a weight represent the importance of each criteria, the WSM can be used. The
weight should be a value ranging between 0.5 and 1.5 as illustrated in Table 4. The value of
0.5 means that the criteria has the lowest importance and 1.5 means that the criteria has the
highest importance. These weights will be different depending on the organisation.
Value of weight Importance of criteria
0.5 Low
1 Medium
1.5 High
Table 4: Weights and importance - the value of the weight signifies the importance of the
criteria
Step (3): The next and third step in the conceptual framework for Idea Selection is to test an
hypothesis. The purpose of this is to prove or disprove assumptions about the idea. These
assumptions relate to one of the main themes. When the test has been conducted, a decision
has to be made whether to continue developing the project or not. To decide this, the Idea
Selection process is conducted again. If there is another opportunity that is more promising,
that opportunity should be pursued. This is an agile iterative process to test a MVP.
33
5. Research Design and Methodological Choices
In the following chapter the methodology of the study is presented, with a starting point in
describing the selection of research design. Secondly, the details of the research design are
addressed. Thereafter, the quality of the research and the ethical considerations are
discussed.
5.1 Research Design
In order to answer the research questions, an exploratory case study has been conducted. This
was considered suitable as the thesis is conducted in a fairly unexplored research area. Even
though research in the fields of innovation, FEI and Idea Selection are not unexplored, Lean
Development methodology is relatively unexplored. Due to this, the research area is not
subject to a large amount of research. An exploratory approach is appropriate in cases when
studying an area that has been subject to little research (Blomkvist and Hallin 2014; Yin
1994). Adding to this, an exploratory approach was appropriate as it expands the
understanding of the phenomenon and contributes to building theory (Eisenhardt, 1989).
The study consists of two main parts. The first part consists of a pre-study that includes an
explorative study as well as a literature study. The second part consists of a case study and
expert interviews. To refine the conceptual framework for Idea Selection in a Lean Context,
expert interviews with employees at Hedvig were conducted. The refined framework was
then applied to an idea at Hedvig selected from a set of ideas collected at Hedvig as part of
the case study. Figure 11 outlines the research design.
Figure 11: Research design
34
5.1.1 Part 1: Pre-study
The purpose of the pre-study was to form a better understanding of the research environment
and find a relevant research question. Moreover, the pre-study is effective in helping
outlining the direction of the research (Blomkvist and Hallin. 2014).
During the pre-study, the researcher acted as an insider of the company where the case was
conducted. This was done in part by initiating unstructured conversations with employees,
managers and supervisors but also by taking part in regular meetings. The advantage of
acting as an researcher was that the researcher was granted access to the main communication
channel of the company and internal documents. This was used to form a better
understanding of the relevant problems at Hedvig that could be translated into a research
question. Also, this contributed to a better understanding of the relevant scope. The
disadvantage of acting as an insider was that it made it more difficult to stay objective in the
research process. Furthermore, a literature study was conducted in order to create the
theoretical and conceptual framework and to understand the current state of the research field.
Recent articles in the field of research were studied to build a broader understanding of a
relevant scope for the study.
5.1.2 Part 2: Case study and Expert Interviews
A total of 12 interviews were conducted and seven of these were conducted with experts in
the field. The purpose of the latter was to gather information needed to refine the framework
for Idea Selection. The conceptual framework was applied to an idea that came from a set of
ideas collected at Hedvig. The reason for doing this was to better understand how ideas can
be evaluated and to validate the framework in a real-world setting. This allowed the
researcher to validate and test the framework in a real-world setting. Furthermore, by
conducting the case study, the researcher gained exposure to the NPPD and learned more
about how Lean Development affects the Idea Selection process. One idea from the set of
ideas was selected to explore further. This was done in accordance with the desire of the
management team at Hedvig and because this idea was most interesting to explore using the
suggested framework. This single idea was evaluated by using the framework and user
research for this idea was conducted to learn more about the Idea. The user research consists
35
of five interviews used to analyse the opportunity. This process is illustrated in Figure 12.
Figure 12: Steps of the case study
The main reason for conducting a case study is that it is well-suited for reaching answers to
“how” and “why” research questions. Moreover, case studies are useful when focusing on
contemporary events (Yin, 1994), which holds true for this thesis. Yin (1994) also states that
it is very important to state the propositions of the study. If an exploratory case study is
conducted, Yin (1994) argues that a clear purpose of the case study should be stated. The case
study conducted in this thesis is of exploratory nature and the purpose of it is to understand
what methodologies in Lean Development and Idea Selection can be applied and is applied in
a real-world setting. The result of this thesis will therefore become more meaningful, since it
can be applied at other startups.
Yin (1994) acknowledges that a single case study is suitable when researchers are
investigating an extreme example, wants to gain unusual research access or have the
opportunity to investigate a particular phenomena under rare circumstances. The lastly
mentioned is something that Eisenhardt and Graebner (2007) acknowledge as well. The case
study is a single case study focusing only on Hedvig and investigating the phenomena of Idea
Selection in the Product Development process of an early-stage startup. Since this
phenomena has not been studied in a real-world context to a large extent, a single case study
will make a relevant research contribution. A single case study can not make an equally
strong foundation for theory building as a multiple case study (Eisenhardt and Graebner,
2007). The same authors also state that a multiple case study can provide broader exploration
of the posed research question and also generate a more vigorous theory. On the other hand,
36
Yin (1994) argues that a single case study is generalisable to theoretical proposition and
therefore a single case study can provide valuable contributions to the field of research.
A linear workflow with planned phases does not highlight all the potential advantages from
case research (Dubois et al., 2002), but the process of going back and forth between empirical
and theoretical activities does. This process is called systematic combining. This approach
can create a greater responsiveness to the observed phenomena (Blomkvist et al., 2014). The
empirical findings from the pre-study, as well as the observations made from acting as an
insider and the literature study have served to continuously form and develop the evolving
case study. By alternating between theory and practice, a deeper insight of the studied subject
area was developed.
5.2 Data Collection
The primary data, i.e. the data collected specifically for the investigation at hand, was
collected by semi-structured interviews. This enabled a flexible approach and allowed for
new ideas and questions to be brought up during the interviews.When the studied phenomena
is episodic and infrequent, interviews are a highly efficient way to gather important empirical
information (Eisenhardt and Graebner, 2007). This is in line with new product development
and innovation in a highly innovative, early startup environment. Thus, the primary method
for data collection has been through interviews, but to broaden the empirical material,
relevant observations at the case-company have been gathered. The empirical material is
displayed in Table 5. The researcher spent about two days per week at the offices of the case
company for the first two months of the study. Because of the COVID-19 epidemic, the
offices of the case company were closed and less observations could be collected.
Nonetheless, the researcher continued to participate in meetings online and over video-chat,
but a degree of natural observations was lost.
37
Component Description Scope Documentation Purpose
Observations Natural observations that have been noted in a Journal
Six observations in Journal and notes from meetings
Notes in text Forming relevant RQ, learning about the subject and Hedvig
Internal documents
Presentations and product demos
Three internal documents of relevance
Original format (power point)
Forming relevant RQ, learning about the subject and Hedvig
Expert interviews
Interviews with employees at Hedvig involved in Product Development or Management
Two interviews (20-30 min each)
Notes from interviews
Develop framework for Idea selection, Pre-study
Expert interviews
Interviews with employees at Hedvig involved in Product Development or Management
Five semi-structured interviews (approx. 40 min each)
Recordings and notes
Refine framework for Idea Selection
Idea survey A survey made at Hedvig with the purpose of collecting new ideas
Approx. 46 ideas Responses in text format
Applying the framework for Idea Selection
User interviews Short interviews with users
Four semi-structured interviews. Six questions
Answers in text Applying the framework for Idea Selection
Table 5: Empirical material used in this study
Eisenhardt (1989) argues that it is advantageous to simultaneously collect data when
performing theory building from a case study. This allows the researcher to adjust and adapt
the case study as the study progressed. Accordingly, interview data was analysed after it was
collected and thereby giving the researcher the opportunity to adjust or investigate further
bases on the findings.
5.2.1 Interview data
Interviews are a suitable method to develop in-depth understanding of the studied area, find
new dimensions as well as ambiguity (Blomkvist and Hallin, 2014). Thus, interviews with
stakeholders in different positions were interviewed. Hedvig can be divided into three main
functions: technology and development, product as well as business and strategy.
Stakeholders from each of these functions were interviewed.
For the purpose of the pre-study, an unstructured method was used to form a better
understanding about Hedvig as a company and finding a relevant scope for the thesis.
Unstructured interviews are suitable for exploring an unknown subject in the early phase of
38
the empirical work (Blomkvist and Hallin, 2014). Ritchie et al. (2013) suggests that an
unstructured interview method gives the interviewer room to open up for questions and
turning the interview into more of a free dialogue. This was the case for the pre-study
interviews and allowed the researcher to adapt and adjust the focus of the questions as the
interview was taking place. For the interviews that focused on understanding the process of
Idea Selection and Product Development, semi-structured interviews were conducted. The
reason why a semi-structured approach was chosen is because this allowed the researcher to
adapt the interviews to focus on topics related to the relevant scientific theory of this thesis,
but still remaining flexible (Yin, 1994). Furthermore, in conjunction with the case study,
semi-structured interviews were conducted with potential users from Hedvig’s target
customer group. In total, 12 interviews were conducted. Moreover, in accordance with
Blomkvist and Hallin (2014), the semi-structured interviews conducted with the purpose of
understanding the process of Idea Selection and NPPD were audio recorded and notes were
taken. The user interviews were conducted in multiple ways, for example, face-to-face, over
the phone and by using Slack. This is Hedvigs main tool of communication and functions like
a chat-board.
Eisenhardt and Graebner (2007) point out that there is a risk of interview data being biased
because the interviewer might try to retrospectively reinterpret the data to fit certain notions.
The authors suggest interviewing several people with different perspectives on the studied
area. This might for example mean interviewing people from different levels of the
organisation or from different functions. In purpose of mitigating potential biases, people
who work in different functions were interviewed. Also, people in leading positions with
CXO titles were interviewed as well as interns and lower level employees. The participants in
the expert and user interviews are outlined in Table 6. The format Framework for Idea
Selection follows a semi-structures approach with three main areas of questions:
(1) About the person, their role and about Hedvig
(2) Ideation and screening of ideas
(3) Product development process
39
Focus was on the second question and the format is further outlined in Appendix 6. The user
interviews also followed a semi-structured format as outlined in Appendix 7. The users were
chosen either because they had experience in the field from working at Hedvig, or because
they were a person in Hedvigs target group.
Interviewee Role Number of interviews
Format
Manager 1 Executive 1
Framework for Idea Selection
Manager 2 Executive 2
Pre-study, Framework for Idea Selection
Manager 3 Executive 1
Framework for Idea Selection
Manager 4 Technology & development 2
Pre-study, Framework for Idea Selection
Manager 5 Technology & development 1
Framework for Idea Selection
User 1 Employee 1 User feedback
User 2 Employee 1 User feedback
User 3 Employee 1 User feedback
User 4 Potential user 1 User feedback
User 5 Potential user 1 User feedback
Table 6: Interview participants
5.2.2 Observations and Secondary Sources
The data gathered in these interviews was also completed with data gathered from smaller
interactions with employees at Hedvig, as well as the researchers partaking in meetings.
Relevant insights were documented in a journal that the researcher has had handy throughout
the study. Furthermore, during the case study, the researcher had the chance to learn how to
use tools that Hedvig use in the product development process. For example, one of these tools
is Figma, which is a prototyping software. While learning how to use this, the researcher was
instructed by employees at Hedvig and the researcher was able to gain insights in NPPD from
learning this tool.
40
During the project, the researcher gained access to internal documents. These documents
included user research, product development processes and strategic business analysis. By
taking part of these documents, the researcher gained insight in how Hedvig has worked with
NPPD historically. The set of ideas was collected from an internal study made by the product
team at Hedvig. These ideas were collected by asking Hedvig employees the following four
questions:
1. What features do you think our members really appreciate?
2. What do you think are the members' biggest pain-point of using Hedvig today?
3. What features would our members be really excited about?
4. What features are you really excited about?
The survey yielded answers with a total of 46 ideas The ideas were categorized into six
categories as illustrated in Table 7 below.
Category Definition Example
New Insurance Offering (NIO)
Involves the type of the underlying insurance product
Weather insurance
General Improvements (GI)
Incremental improvements to the core product or offering
Coverage of more expensive items
Claims
Relates to how the customer makes claims
The ability to make a claim on the web
Onboarding
Relates to the onboarding process
More discount codes
Information accessibility
Involves improvements in how customers access information
Visualisation of insurance terms
Other
Not covered by any of the other categories
Offer a personal butler
Table 7: Categorization of ideas
41
5.3 Research Quality
Collis and Hussey (2013) state that for quantitative research, reliability refers to accuracy in
measurements and likelihood of identical results if the research is performed by another
researcher. Being that the thesis is conducted as a case study, an exact replication of the
findings is hard to achieve. Thus, reliability is not reached in a similar way as for
quantitative research (i.e. being replicable). Transparency is achieved in the research as the
different parts and steps of the methodology is described in detail, providing the reader a
good understanding of how the study has been conducted. Also, the process of recording and
transcribing the interviews further strengthens the reliability of the results.
During the research, Validity was intended to be attained by the combination of external
interviews and the empirical material, as this enables for triangulating the data and gathering
information from different sources. Moreover, the continuous dialogue and close relationship
with Hedvig enabled that the findings were validated further. Due to the qualitative nature of
the study, it is problematic to generalize the results with high certainty. Nonetheless, the level
of generalizability was increased by including several, independent respondents in the case
study (Blomkvist and Hallin, 2014). Validity can be defined as the extent of tests measuring
what they are intended to measure and to what extent the results of the test reflect the studied
phenomena (Collis and Hussey, 2013). Moreover, the degree of truthfulness of the results can
be considered in addition to this as suggested by Golafshani (2003). Thus, the research results
can not be subject to manipulation in order to maintain the integrity of the research.
5.4 Research Ethics
Ethics and integrity have been taken into consideration during the research process.
According to Blomkvist and Hallin (2014), the most common ethical codes within social
science in Sweden are the principles of the Swedish Research Council. The research has been
conducted in agreement with the formal guidelines of the Swedish Research Council’s
“Principles of Research Ethics” as follows:
(1) Information Requirements
The researcher should inform the interviewees about the purpose of the study.
42
(2) Consent Requirements
Interviewees in the study should decide on their own participation.
(3) Confidentiality Requirements
Information about all interviewees shall be given the greatest possible confidentiality
and personal data must be stored in such a way that unauthorised persons cannot
access it.
(4) Utilisation Requirements
Data collected shall only be used for the research purpose.
43
6. Findings & Analysis
In this section, the findings and analysis are presented based on the collected empirical
material. Firstly, the components of the Framework for Idea Selection as well as how to
implement it is described. Then, the framework is applied to an idea at Hedvig to validate the
framework in a real-world setting.
6.1 Components of the Framework for Idea Selection
In this section, the findings from the expert interviews are presented and connected to the
theoretical material presented in the conceptual framework. First, the components of the
conceptual Framework for Idea Selection are presented again. Secondly, the findings from
the expert interviews are presented structured according to the four themes: Reach, Impact,
Confidence and Effort. Lastly, the implementation of the full final framework is described.
The aim of this section is to connect the expert interviews with the conceptual framework.
The conceptual framework for Idea Selection (first presented in chapter 4) is illustrated in
Table 8 below and contains four main themes: Reach, Impact, Confidence and Effort.
Theme Criteria Source
Reach Market size
Rochford (1991), Ebner et al. (2009)
Potential for market growth Rochford (1991)
Impact Customer benefit Ebner et al. (2009)
Competitiveness Bitman & Sharif (2008)
Confidence Profitability of success
Rochford (1991), Ebner et al. (2009)
Personal experience Bitman & Sharif (2008)
Effort Time and cost of execution Adapted from Rochford (1991) Table 8: Proposed selection criteria, conceptual framework for Idea Selection
44
The components of the conceptual framework have been refined to fit in the Lean
Development context of a startup. To achieve this the empirical findings have been
incorporated into the framework.
6.1.1 Findings from Expert Interviews
In this part, the expert interviews conducted in the case study are introduced. The material is
presented by using the four common themes presented in the conceptual framework: Reach,
Impact, Confidence and Effort. For each theme, selection criteria formulated using the
interview data is presented. In Table 9, a summary of the criteria is described.
Theme Criteria Source
Reach
Potential number of users Manager 3
Number of touchpoints Manager 1
User acquisition Manager 1, Manager 2,
Manager 4
Impact Customer benefit Manager 3
Confidence Personal experience Manager 1
Effort Possibility to experiment Manager 2
Number of stakeholders involved Manager 5
Table 9: Findings - Selection criteria derived from expert interviews
Reach
Potential number of users
Support for this criteria is found in the empirical material as well as the theoretical. This
criteria is relevant to consider because it directly relates to one of the main KPIs at Hedvig:
gross written premium (GWP) (Manager 3). The GWP is the total premium written by an
insurer before deductions for reinsurance and ceding commissions. One can think of GWP as
sales and a large market, with many potential users, entails high sales.
45
Number of touchpoints
Another criteria to consider is the number of touchpoints that the new product has. A
customer touchpoint occurs when a customer interacts with the product. A very good product
has a large number of potential users and has the potential to be used frequently. When new
product ideas are considered, unrealistic use-case scenarios are imagined and it is important
to stop and ask how often the imagined use-case actually occurs (Manager 2). When making
this point, the interviewee suggests an idea for product that is very useful in a very specific
scenario and makes the point:
“[...] but how often do the stars align and that exact scenario happens?” (Manager 2)
User acquisition
An important aspect of creating reach is virality. Several interviewees reflect on the
importance of creating virality in connection to user growth. One interviewee states:
“We are mostly focused on creating viralty and growth.” (Manager 4)
Further supporting this is the fact that one of the main KPIs at Hedvig is how efficiently
customers can be acquired (Manager 3). This is measured in payback time:
“The single most important factor [to consider for future growth] is the equation
between how much we pay to acquire a customer and how long it takes for us to
recover that cost.” (Manager 1)
Creating virality is a way of connecting with the potential users included in the reach of a
new product. Just because a product has a reach of a certain number of users, does not mean
that these users will actually use this product. The company that has created the product
needs to acquire the potential users to turn them into actual users. This can be done for
example by paid marketing or very cheaply and efficiently by creating virality. On this
subject, in relation to evaluating ideas, one of the interviewees note:
46
“What should we do next, in terms of product development, that will set us apart and make us viral? How can we create a sort of news value in the product offering?” (Manager 2)
At Hedvig, creating virality is important because it contributes to organic user growth. This is an efficient, cheap and sustainable way of acquiring users. Hence, the criteria of user acquisition has been included.
Impact
Customer Benefit
One of the interviewees states that there are two ways of creating successful products. Either
by creating a high-functionality product or by creating a product surrounded with a strong
sense of community as illustrated in Figure 13. Apple is an example of a company that has
succeeded with the latter. One of the interviewees makes this point and says:
“The MacBook is an inferior computer to the ThinkPad under the shell and it is more
expensive, but it has a glowing Apple on the front of it.” (Manager 2)
Apple is known for their strong sense of community and that’s one of the reasons why the
MacBook is a more popular laptop than the ThinkPad. One of the ways Hedvig creates
community is by sharing their vision and building products to reflect that vision. On this
subject, one interviewee states:
“We have to support Hedvig’s long term identity” (Manager 2)
The other way of creating successful products is to create high-functionality products. A
high-functionality product is synonymous with a product with a high degree of usefulness.
An example of a product with high-functionality and a low sense of community is the
microwave oven. Few associate their microwave oven with a sense of community, but most
would probably agree that it is very useful.
47
Figure 13: Community and functionality are important parameters for creating successful
products
One of the four main KPIs at Hedvig is retention and this is measured in the number of
months that the customer is expected to stay (Manager 3). Hedvig is a monthly subscription
service and the customer can choose to cancel their subscription any time. Therefore, if
Hedvig does not provide high enough customer benefit, either in terms of functionality or
community, the customers will cancel their subscriptions. For this reason, retention is a
measurement of the customer benefit.
Confidence
Personal Experience
As described above, the purpose of the confidence theme is to estimate the certainty of the
assumptions made when considering the other selection criteria. One way to do this is by
considering the personal experience that the individual (or team) that suggested the idea have.
On the subject of evaluating ideas, one interviewee from the management team states:
“There must exist an inherent system for assigning more weight to [the opinion of] a
person who is more believable within the subject.” (Manager 1)
48
The interviewee further justifies this statement by explaining that someone who is believable
is someone with experience who can explain how something has been executed in the past. A
way of judging the confidence is by considering the track record of the individual (or team)
that suggested the idea.
Effort
Number of stakeholders involved
The effort it takes to execute a project is not a function solely dependent on the consumed
resources, but it also depends on the complexity of the project. A high number of
stakeholders involved in the project will raise the complexity of the project (Manager 5).
Therefore, it is relevant to consider the number of stakeholders involved in the project when
evaluating the effort. A stakeholder can for example be in the form of external partners,
internal teams or employees involved in the project.
Possibility to Experiment
When considering the effort it takes to execute the project it is important to define the scope.
The scope can for example be to fully finish the project and launch the final product, or the
scope can be to develop a MVP. Some products are better suited than others for
experimenting. When asked about how one of the interviewees think about cost and benefit
of new product development, the interviewee explained that they do not necessarily use the
ROI-perspective when considering new ideas, but instead consider:
“What is the minimum thing that we can build so that we learn more about this?”
(Manager 2)
Effort can be considered in terms of evaluating the effort of developing the finished product
and comparing this to the potential rewards. Another way is to consider the effort to develop
a MVP and compare this to the potential information gain or learnings.
49
6.1.2 The Components of the Final Framework for Idea Selection
The components of the final framework for Idea Selection is presented below in Table 10.
The framework contains the four themes (Reach, Impact, Confidence and Effort) derived
from the conceptual framework and the empirical material. The criteria used in the
conceptual framework have been merged with the criteria that has emerged from the expert
interviews. Moreover, some criteria have been merged to make the criteria in the framework
as close to mutually exclusive as possible and more clear.
Theme Purpose Criteria
Reach How often will this product be
used?
Potential number of users
Potential future users
Number of touchpoints
User acquisition
Impact How much value does this
product bring users? Customer benefit
Competitiveness
Confidence How confident are we in our
assumptions? Probability of success
Personal experience
Effort What will it take to realize this
product?
Possibility to experiment
Time and cost to execute
Number of stakeholders involved
Table 10: Components of the final framework for Idea Selection
Reach: The Criteria within this theme is aimed at estimating how often the product will be
used. This is a function of the number of potential users, how many of these users that can be
acquired and the number of touchpoints (i.e. how often the users interact with the product).
Furthermore, Rochford (1991) suggested the potential for market growth as a selection
criteria and therefore the potential future users are also considered.
Impact: The purpose of this theme is to understand how much value the product can bring to
the users. This value is measured by considering the customer benefit in terms of the
functionality that the product provides and the sense of community. Moreover, the
50
competitiveness of the product should be assessed. If there are many competing products or
substitutes that satisfy the same need or solve the same problem, the value of the product that
is being evaluated is far less.
Confidence: The probability of success and the personal experience of the individual with
the idea is evaluated to understand the level of confidence in the assumptions made in the
evaluation. When considering the probability of success, it should be considered if the skills
and tools needed to execute the project exist. The probability of success should be an estimate
of the likelihood of succeeding with executing the project in the scope of the effort as defined
in the next theme. The confidence level can be raised if an experiment has been conducted as
explained in the next theme.
Effort: To raise the confidence level, an experiment can be conducted by building a MVP. In
accordance with Lean Development, assumptions can be tested by building a MVP and
thereafter gathering feedback. This way, if certain assumptions are tested, the confidence
level can be raised. For this reason, the possibility to experiment should be evaluated. Some
product ideas allow for a large degree of experimentation while others do not. If assumptions
can be easily tested, this can be highly valuable. Furthermore, the time needed and the cost of
building the MVP should be considered. What is more, if many stakeholders are involved,
this will raise the complexity of the project and therefore the effort.
6.1.3 Implementing the Framework
The components of the framework have been established by using the conceptual framework
and the empirical material. This section aims to explain how the framework can be used in a
startup to evaluate ideas and how to select the most promising ideas to pursue.The first step
in using the framework is a pre-screening process. The pre-screening consist of three simple
questions:
(1) Is the idea consistent with company objectives?
(2) Is this product solving a problem?
(3) Do I have this problem or do I know anyone that has this problem?
51
The purpose of having a pre-screening is to filter out ideas at a very low cost. The first
question is inspired by Rochford (1991) and serves to answer if the idea is relevant to the
company in question or not. Hedvig set up strategic goals that they are working towards.
These goals are called stones and can be for example expansion to a new market or a
significant update of the user interface (Lucas). Some ideas can be for products relating to an
existing stone or can mean entirely new stones. It is important to consider how (and if) the
product fits in strategically with the stones. The second question is inspired by a statement
made by an interview conducted with Manager 1 who notes the importance of actually
solving a problem for the intended user. Lastly, the final question challenges potential
assumptions made when answering the second question. This question was inspired by an
interview with Manager 2, who explained that the first step in evaluating an idea is to
imagine the process would look like if you were to solve the imagined problem for yourself
or a friend. If the answer to this question is no, the first step is to find out who has the
imagined problem. If the answer to all questions in the pre-screening is yes, then the idea
passes the pre-screening and can move on to the next step in the evaluation by applying the
framework.
The second step in applying the framework is to evaluate the selection criteria. These criteria
are presented in the framework illustrated in Figure 14, below. This framework brings up
vital themes and criteria to consider when evaluating an idea for a new product. The end
product is a numerical score, the comparative score, between 5.5 and 82.5 that reflects the
potential of the idea. If several ideas are compared, they can be evaluated using this
framework and then sorted by the comparative score.
52
Theme Criteria Importance
(0.5-1.5)
Score (1-5)
Comparative
score
Reach
Potential number of users x =
Potential future users x =
Number of touchpoints x =
User acquisition x =
Impact Customer benefit x =
Competitiveness x =
Confidence Personal experience x =
Probability of success x =
Effort
Possibility to experiment x =
Time and cost to execute x =
Number of stakeholders involved x =
Σ
Figure 14: Framework for Idea Selection
The importance of each criteria should be rated. The value of the weight should be between
0.5 and 1.5 as illustrated in Table 11. The next step is to iterate through each criteria and set a
score from 1 to 5 (low to high) depending on the potential of the idea with regards to that
specific criteria. In other words, the performance of the idea in relation to each criteria
should be evaluated. For example, if the product has a very high number of potential users,
the score for the first criteria, “Potential number of users” should be 5. This method of
scoring 1-5 on criteria is also utilized by Bitman & Sharif (2008) to collect an overall
assessment of the idea. This framework implements a WSM because a recent study by Cui et
al. (2019) suggests that the WSM strictly outperforms the simpler ranking model in terms of
the likelihood of selecting the highest-quality ideas.
53
Value of weight Importance of criteria
0.5 Low
1 Medium
1.5 High
Table 11: Weights and importance
The next step in the innovation process is to find the idea with the highest potential by
selecting the one with the highest comparative score. When building a MVP, the assumptions
made when evaluating the idea can be tested. The purpose of the MVP should be to verify the
assumptions made about the Reach, Impact, Confidence and Effort. If this is done, the
Confidence, or the probability of success will increase because the assumptions made will
have a higher likelihood of being true. Testing and analysing a MVP is part of Lean
Development. After this, assumptions and estimates can be adjusted in the framework and a
comparative score with higher accuracy can be calculated. This process is illustrated in
Figure 15 below.
Figure 15: Illustration of the Idea Selection Process
6.2 Applying the Framework for Idea Selection
The framework for Idea Selection has been applied to an idea at Hedvig within travel
insurance. The purpose of this was also to validate the framework in a real-world setting.
6.2.1 Idea Description
One idea from a set of 46 ideas collected at Hedvig was selected for the framework to be
applied to. The idea that was selected relates to automatic travel claims. The idea was
54
selected for further analysis because of directions from the management team at Hedvig,
because it was one of the most common ideas suggested and because it is an interesting idea
apply the framework to. The employees who participated in the idea collection survey have
slightly different ways of describing the idea:
Selected answers to question (3): What features would our members be really excited about?
Weaver, P. (2000). Sustainable Technology Development. International Journal of
Sustainability in Higher Education,1(3), pp.305-308.
Weiblen, T., & Chesbrough, H. W. (2015). Engaging with startups to enhance corporate
innovation. California management review, 57(2), 66-90.
Womack, J. P., & Jones, D. T. (1997). Lean thinking—banish waste and create wealth in
your corporation. Journal of the Operational Research Society, 48(11), 1148-1148.
Yin, R. K. (2003). Case study research: design and methods (ed.). Thousand Oaks.
York, J. L., & Danes, J. E. (2014). Customer development, innovation, and decision-making
biases in the lean startup. Journal of Small Business Strategy, 24(2), 21-40.
70
Appendix 1
Insurance Industry, Source: Hedvig
i
Appendix 2
Founding date of Insurance companies
Company Year founded
Folksam 1908
If (Skandia) 1855
Trygg-Hansa (Städernas allmänna brandstodsbolag)
1828
Länsförsäkringar 1801
ii
Appendix 3
Compensation scheme for flights within EU
Length of delay Journey distance Compensation
>3 hours
< 1 500 km €250
> 1 500 km (within EU) €400
1 500 - 3 500 km (all EU departing flights)
€400
3-4 hours > 3 500 km (from an EU country to a non-EU country)
€400
>4 hours > 3 500 km (from an EU country to a non-EU country)
€600
>5 hours Any flight departing from EU €600
iii
Appendix 4
iv
Appendix 5
v
Appendix 6
About the person, their role and Hedvig -Please describe your role as XX -...biggest challenges in your role? -Tell me about Hedvig -How did Hedvig start?
-Technical challenges? -Partnerships?
-Greatest challenges for Hedvig in the coming years? Ideation and Idea Selection -How do Hedvig find new ideas? -How do you choose which ideas to pursue? -How do you choose what to work on? -Is Hedvig an innovative environment? Tell me more? -Is there a formal way to assess new projects? What do you think when you hear a suggestion for a new product or feature? -How do you reach growth? Most important parameters? -[Explain ROI perspective] How do you think about effort? How do you think around impact? -[Automatic travel claims] What do you think about it? -[Automatic travel claims] What do you think when you hear a suggestion for a new product or feature? -[Automatic travel claims] How do you think about effort? How do you think around impact? Product development process -What do you do if you hear a great idea in your role as XX? -What are the steps between idea and launch? -What is an MVP? Tell me more? -What is a prototype? Tell me more? -Why is user feedback important? -How do you incorporate user feedback?
vi
Appendix 7 Purpose What’s the purpose of the interviews?
● Find out about the interest and market for the idea of mitigating worries related to long distance travel and a system that automatically makes a claim if the plane is delayed/cancelled
What do we want to learn? ● Do people feel insecure about travelling? ● What do people worry about when travelling? ● Do people know about and use services like AirHelp? ● Would users like an AirHelp-like service integrated into Hedvig? ● Do people care about making faster claims? Automatic claims? The estimated value?
Any hypothesis that you’re looking to validate or falsify?
● We believe people want to feel more safe regarding travel, especially when it comes to the travel by airline and train
● We believe people want help making claims to airlines/SJ ● We believe people think it's cumbersome to report delayed travels