Department of Informatics Master thesis, 30 hp

IT Management SPM 2020.15

Enablers and Inhibitors of Digital Startup Evolution

A Multiple-Case Study of Swedish

Business Incubators Andrew Page

1

Abstract Global advances in digital technology are facilitating a corresponding rise in digital entrepreneurship and its startup manifestation. Many uncertainties exist upon the road to digital startup evolution; a number of which may be successfully navigated with the assistance of business incubators. While these organisations provide valuable guidance and support to the startup community, their efforts are, at least in some part, constrained by the lack of a consistent and coherent roadmap to guide both them and their incubatees. This work proposes a solution to that deficiency by addressing the question - What are the enabling and inhibiting factors in digital startup evolution within an incubator setting? - via a multiple-case study that examined digital startups under the umbrella of three business incubators in the city of Umeå, Sweden. This work builds on the existing literature both through its narrowed focus on incubators as well as through its presentation of the Ideation Dynamics Model as a proposed guide for both incubators and digital startups to follow. Keywords: Digital Entrepreneurship, Digital Startups, Disintegration, Failure, Incubators, Ideation Dynamics Model, Inertia, Pivoting, Scaling.

1. Introduction Digital technologies of today, such as social media, mobile technologies, business analytics, big data, and advanced manufacturing, are opening fascinating innovation opportunities for entrepreneurs (Cohen, Amorós & Lundy, 2017). Digital entrepreneurship – defined as the practice of pursuing “new venture opportunities presented by new media and internet technologies” (Davidson and Vaast, 2010) – is attracting worldwide attention (Nambisan, 2017) and (Fang, Henfridsson & Jarvenpaa, 2018). As such, while this form of entrepreneurship is similar to traditional entrepreneurship in many ways, one significant difference lies in the fact that within the former, some or all of the key activities take place digitally instead of in non-digital formats.

As digital entrepreneurs adopt new technologies to develop novel forms of entrepreneurial actions that accelerate the evolution of new ventures (Huang, Henfridsson, Liu and Newell, 2017), the complexity also increases. As such, there are many organisations and service providers that seek to act as an incubator in order to facilitate entry to the world of entrepreneurship. Specifically, incubators aim at reducing the barriers associated with entrepreneurship through coaching, office space provision, knowledge, and funding (Ratinho, Amezcua, Honig & Zeng, 2020). An incubator can be defined in many ways, but according to the branch organisation, Swedish Incubators and Science Parks (SISP), an incubator is “an entity that offers a dynamic process to developing businesses, people, and companies”. Thus, an incubator assists entrepreneurs with management, financial support and technical competence while also facilitating connections to both new environments as well as to a commercial network within which to prosper. It is important for prospective digital entrepreneurs to build, learn and evolve in such incubator contexts. Ries (2011)

2

defines one such learning method – pivoting - as “a structured course correction designed to test a new fundamental hypothesis about the product, strategy, and engine of growth”.

It has been suggested by the work of Huang et al. (2017) that the digital and therefore relatively flexible nature of today’s startups facilitates comparatively rapid implementation of these learnings compared to more traditional organisations; “bits are easier to change than atoms.” Such authors as Kirtley and O’Mahony (2020) and Bandera and Thomas (2019) agree on the desirability of adaptability as an entrepreneurial characteristic that will allow them to respond more rapidly – for example by scaling up their business – to their environmental stimuli. However, this position is at least partially contradicted by Eesley and Wu (2020). Their more nuanced view identifies both advantages and disadvantages related to flexibility depending upon, for instance, both the time scale through which the enterprise is viewed and the advisory networks that the organisation is able to access and utilise. Furthermore, the same authors, while endorsing the requirement for startups to react to environmental incitements, fail to offer any meaningful guidance regarding what form this may take.

In the context of considering a potential reaction to these indications, Ries (2011) writes in great detail on the topic of startup pivoting which - while of undoubted importance to the startup community - is not, generally speaking, adequately represented in well cited papers. The majority of work that does exist in the area has been concentrated on the software industry. The importance of pivoting is well summarised by Cusumano (2013) who explains that without being able to demonstrate the flexibility required to pivot, a startup may struggle to raise its life blood; investment capital. It is generally agreed by a number of other authors – including Bajwa, Wang, Duc & Abrahamsson (2016), Bajwa, Wang, Duc & Abrahamsson (2017), Ochoa Zambrano & Garbajosa (2017), Bohn and Kundisch (2018) and Terho, Suonsyrjä, Karisalo & Mikkonen (2015) – that at least twelve different types of pivot exist.

It can be imagined therefore, that when one attempts to subdivide the topic of pivoting further by examining the triggers of startup pivots, the literature field becomes even sparser; with Bajwa et al. (2016), Bajwa et al. (2017) and Comberg, Seith, German & Velamuri, (2014) contributing the bulk of the relevant research. Between them they distinguish fourteen separate pivot triggers that are both internal and external in nature.

The consequences associated with both successful, unsuccessful and even a complete absence of pivoting, are also considered in this piece; these comprise scaling, inertia and disintegration/failure. Scaling is defined by Sahay and Walsham (2006) as “the process through which a product or process is taken from one setting and expanded in size and scope within that same setting and/or also incorporated within other settings”. Therefore, and in the context of this thesis, scaling represents the increase in operational magnitude that digital startups may undertake in order to maximise the fiscal benefits of their strategic adjustment. An example of this could include a huge uptake in the sales of a pharmaceutical product that was previously approved for use in a niche disease area but, after a change in the disease area focus – and resultant strategy – by the patent holder, is now approved for usage in a far broader area. The marketing authorisation holder would then have the opportunity to execute a scaling exercise. The work of Huang et al. (2017) outlines a number

3

of key mechanisms – data-driven operation, instant release and swift transformation - by which the process of scaling may take place.

Startup disintegration/failure on the other hand can be viewed as both a positive and negative. The positive effects associated with failure are emphasised by the work of Gartner and Ingram. (2013) and Da Silva, Fabrìcio, Da Silva Pinto, Galegale & Akabane (2019). However, while failure may hold certain benefits, there are of course certain advantages inherent in not failing. Opinions of failure are greatly dependent upon the geographical – and associated cultural setting in which it takes place. Part of the rationale underpinning this socio-cultural divergence of perception between, for instance, EU citizens and those of the USA, can be traced to the significant gap (45% vs 67% respectively) regarding their proclivity towards being self-employed in the first place, (Bosma and Schjutens, 2011). Since self-employment is associated with a higher risk of failure, it seems reasonable to argue that at least some who pursue it are more comfortable with this level of jeopardy than those who do not. Brenner and Fornahl (2008) expand upon this point by citing further differences between the above two geographies; including relative scarcity of funding, consequences associated with bankruptcy, education & culture and fragmentation of member states. All these may have an impact upon failure perception. Even within Europe, the work of Bosma and Schjutens (2011) explains that fear of failure varies significantly between regions and countries. This is due to a number of factors including the potential negative consequences of such an outcome, underlying unemployment rates and population densities. Four strategies – focusing on the team over the product/service, business model innovation, business model development and business model scalability - to avoid failure, specifically with an incubator setting, are identified by the work of Nair and Blomquist (2019).

An additional but unexpected outcome that emerged from the empirical results is that of inertia, a dynamic state more associated with organisations that are both greater in magnitude and maturity than startups. On the basis of the literature search described below, entrepreneurial inertia does not appear to be a topic that has been widely considered within extant startup research. In fact, other than the work of Ruef (2006) – who defines entrepreneurial inertia as “the lag time exhibited by organizational founders or investors entering a market niche” – and Gong, Baker & Miner (2009), I was unable to locate any relevant research on the topic. In simpler terms, inertia can be considered as a company deciding to take no action regarding its strategic position. It may feel that its current course is the correct one and demonstrate no further interest in either testing any of its existing hypotheses or incorporating the results of any such analysis into its operational model. This unwillingness or inability to capitalise upon a pivot – for instance by scaling up a digital startup – is also a phenomenon that exists, however due to its very limited presence in published material, it is not addressed in the literature review below.

This work thus proposes to examine the subject of pivoting within digital startups in combination with the potential consequences of such an act – inertia, scaling and failure/disintegration - while considering the influence brought to bear upon the organisations by entrepreneurial focused incubators. While most of these topics are examined in extant literature, there appears to be no work that uniquely focuses on them

4

in combination with each other. This permutation of topics therefore led me to the following research question: What are the enabling and inhibiting factors in digital startup evolution within an incubator setting?

In an effort to answer this question I undertook a multiple-case study, examining a wide variety of digital startups within three incubators in the city of Umeå in Northern Sweden. Two of the three incubators focus specifically on the life science and creative industries while the third facilitates organisational development for startups that operate within a relatively more diverse set of industries. It seems to be plausible that, compared to digital startups operating outside of incubators, a different set of environmental forces may exist within these organisations. This may, in turn, result in the existence of enabling and inhibiting factors that are of an equivalent (or greater/lesser) diversity.

This paper is divided into six main sections. Section 2 attempts to examine the extant research around both pivoting and the triggers that can lead to it. Furthermore, it discusses two of the potential outcomes associated with pivoting; scaling in the case of a successful pivot and disintegration/failure in the case of either not pivoting or executing an unsuccessful pivot. Section 3 examines the methodology underpinning this work - as well as the publications and factors justifying the choice of the various methodological components over alternatives ; specifically focusing on a comprehensive explanation of the multiple-case study as well as such important matters as sampling, data collection, data analysis and ethical issues. Section 4 presents the key components of the thesis’s empirical findings. Section 5 discusses and reflects upon the essence of the relationship between the findings and the existing literature, while also considering the implications that this relationship may have for the practices of both incubators and digital startups. Section 6 summarises conclusions and thoughts regarding potential future research opportunities that could contribute to the richness and diversity of extant knowledge contained within this area. This work concludes with an acknowledgements section, a comprehensive list of references and a brief appendix.

2. Literature Review In order to gain a thorough understanding of previous work related to the research question, a comprehensive audit of literature relevant to enabling and inhibiting factors in digital startup evolution was undertaken. As mentioned above, there is little data on these factors within the more nuanced field of incubators. This work will therefore attempt to populate this gap in the extant literature. A comprehensive and focused search of a number of database sources was undertaken,

with specific focus being on Google Scholar, Scopus, The Association for Computer Machinery Digital Library and last, but not least, Umeå University Library’s proprietary search tool. The keyword combinations utilised are too numerous and tedious to recount in this work, however the many variations were populated with the following search terms: • Startup(s)/ Start-up(s) • Pivot(s)/Pivoting • Incubator(s)

5

• Digital Entrepreneurship • Scaling/Scaling-up • Disintegration/Failure • Inertia

Although in some papers, such as Comberg et al. (2014), the term ‘business model innovation’ is used as a proxy for pivoting, in most published sources the two terms exist as separate phenomena. Papers including this term were therefore jettisoned from the review in an effort to minimise confusion and enhance clarity of thought. Research Streams Definition Supporting Papers Pivoting Modifying an

organisation’s strategy in a manner that does not result in a corresponding change to its vision.

Bajwa et al. (2017) Bohn and Kundisch (2018) Comberg et al. (2014) Ochoa-Zambrano and Garbajosa (2017) Terho et al. (2015) Hirvikoski (2017)

Scaling Maximising the ability of an organisation to achieve rapid growth.

Huang et al. (2017) Picken (2017) Brynjolfsson and Saunders (2009) Henfridsson and Bygstad (2013) Sahay and Walsham (2006)

Disintegration/Failure The inability of an organisation to continue operations in its current form.

Bajwa et al. (2016) Bandera and Thomas (2019) Giardino, Wang & Abrahamsson (2014) Unterkalmsteiner, Abrahamsson, Wang & Nguyen-Duca (2016) Eisenmann, Ries & Dillard (2013)

Table 1: Research Streams, Definitions and Supporting Papers

Various combinations of the above search criteria resulted in a primary source catalogue numbering several hundred, comprising papers, online articles, conference proceedings, academic textbooks as well as theses authored by both PhD and masters level students. While not all of these had the chosen research topic as their core focus, those selected all make some contribution to the research foundation upon which this work is built. The decision to retain or discard the sources that were unearthed during the literature review was informed by a number of factors. While there are a number papers related to pivoting – the vast majority of which pay tribute to the ground-breaking work of Eric Reis; a truly original pioneer in this field - relatively few focused upon its execution within digital startups or explored the actual triggers of the phenomenon. Papers that did not examine pivoting in the context of a digital startup were mostly discarded from further

6

consideration. This latter criterion was also rigorously applied to sources related to scaling and disintegration/failure. Finally, where multiple papers covered very similar or identical topics, consideration was given to the standing of the journal in which they were published, the number of citations attributed to them, as well as the previous work and reputation of the authors. This process facilitated the streamlining of utilised references to a level appropriate for a work of this magnitude.

The literature review identified three potential outcomes for organisations considering whether or not to pivot; pivoting, disintegration/failure and scaling. These outcomes – defined by supporting literature and displayed above in Table 1 – therefore comprise the main components of this literature review. Their essence will now be discussed further.

2.1 Pivoting This component of the thesis focuses on both the act of pivoting, the various forms it exists in, as well as a myriad of factors that have been shown to trigger it.

Ries (2011) defines a pivot as “a structured course correction designed to test a new fundamental hypothesis about the product, strategy, and engine of growth”. The work of Kirtley and O’Mahony (2020) on the other hand describes it as “a change in a firm’s strategy that reorients the firm’s strategic direction through a reallocation or restructuring of activities, resources and attention”. In other words, and in the context of this thesis, pivoting exists when a digital startup develops a product or service, tests it with its market and proceeds in a direction dictated by the outcomes – and informed by the learnings - of that experiment.

The work of Bajwa et al. (2017) explains that pivots often result from the analysis of corroborated customer feedback related to a particular hypothesis or product/service. This feedback forms a central component of Ries’s (2011) Lean Startup approach which mandates the development and testing of a premise and subsequent incorporation of the results into a startup’s decision-making processes. Both Giardino et al. (2014) and Ochoa-Zambrano and Garbajosa (2017) explain that pivoting may come about both in an effort to facilitate the matching of a product to the needs of a market, as well as being a result of this objective not being achieved.

Divergent views exist concerning the magnitude of strategic change required for a pivot to exist. While Bandera and Thomas (2019) believe that the scale of change associated with pivots can be both significant and small, others such as Ries (2011) and Blank (2020) regard them as major adjustments in the journey of a startup.

One can consider the experimental work associated with pivoting as the antithesis of what Crilly (2017) refers to as fixation; the idea of remaining loyal to a group of beliefs, thereby reducing or removing the potential for learnings to be absorbed and pivoting to take place. Cusumano (2013) however suggests that successful pivots require a significant degree of organisational flexibility.

According to the work of Ries (2011), Terho et al. (2015), Hirvikoski (2014), Bajwa et al. (2016) and Bajwa et al. (2017), the following distinct pivot types exist:

7

• Zoom-in: A solo feature of a product becomes the product in its entirety. Alternatively, it can be a narrowing of market focus, possibly from an entire market to an individual segment.

• Zoom-out: What was considered to be a product does not address market needs and instead becomes a feature of a more complex and substantial offering.

• Customer Segment: A startup understands that its initial target audience does not possess an unmet need matching the product’s profile; however, a different group of customers does and so the product is repositioned accordingly.

• Customer Need: The product doesn’t meet the original customer need envisaged, however it has the potential to solve a previously unseen problem instead.

• Platform: An app evolves into a fully-fledged platform, although this term can also refer to the reverse process.

• Business Architecture: This pivot type concerns two divergent organisational structure designs; namely complex systems model – which represents a low volume, high margin philosophy - and volume operations model, which represents the diametric opposite. In a business architecture pivot therefore, the pivot is from one model to the other.

• Value Capture: A change in the means – of which there are clearly many - by which a startup derives value from its operations.

• Engine of Growth: Often accompanies the preceding category on this list and denotes a change in the growth strategy pursued by the startup.

• Channel: A simple change in the route by which a startup ensures that its product reaches its customers.

• Technology: A variation in the form of technology that a startup utilises to generate a solution to its customers’ issues.

• Complete: A pivot where, while the startup’s core team remains intact, seismic shifts occur in such business aspects as product offering, target market and financing.

• Side Project: A switch to a supplementary initiative that is running on a similar timeline to the primary endeavour of the startup and shows greater potential.

• Social Pivot: where dynamic alterations in social factors, for example, people and environments, alter the direction of an organisation.

2.1.1 External Pivot Trigger Factors In their studies, Bajwa et al. (2016) and Bajwa et al. (2017) identify a number of pivot triggers; i.e. factors that cause the execution of a pivot. These can be categorised as either external or internal. The external triggers comprise the following components: • Negative customer feedback: To operate effectively, a startup requires customers.

They must be acquired, retained and capable of generating useful volumes of insightful data; otherwise the organisation may have to pivot.

• Failing to compete with competitors: If a startup finds it cannot successfully contest a market as a result of its competitors being able to more effectively commercialise a similar concept, then a pivot may be required to remedy the situation.

• Technology challenge: Technology is often a startup’s unique differentiator. If it cannot access the technology required to solve its customers’ problems and/or more

8

effective options become available, then a change of strategic direction can be a required outcome.

• Influence of investment: It can be challenging to resist pressure and/or proposals from powerful stakeholders within a startup who feel a new course should be plotted to achieve what they see as the organisational vision. Submitting to and pivoting in the direction of their ideas may be the only option if the startup is to survive.

• User appreciation of one particular product feature: Sometimes, a particular feature of a product may be viewed by its customers as offering considerably more value than other accompanying features. In this scenario, it may behove the startup behind the product to reconfigure the product in order to position this feature as the core of their offering.

• Unanticipated use of product by users: A startup may market its product as possessing utility that can address an unmet need that they believe exists in particular markets. However, when its customers use the product in a manner partially or completely unrelated to the purpose for which the company promoted it, then it makes sense to undertake a pivot and reposition the product.

• Wrong time: A startup may bring a product to market that is quite literally ahead of its time – and possibly also behind its time - meaning that the market may not be cognitively aware of the problem that the product is designed to solve. The startup may therefore have to pivot to an alternative solution with potential for better reception and uptake.

• Positive response from an unforeseen customer segment: When one segment of a startup’s market exhibits a strong and unexpected demand for its offering, then conventional wisdom suggests that the organisation should narrow its focus to this group and go fishing where there are fish to be caught.

• Running into legal issue: Patent infringement and copyright violations are just two of the elements of intellectual property legislation that can frustrate a startup’s efforts and render them unsuccessful. Clearly in this scenario the company must refocus its efforts on areas that offer greater legal certainty and more assured returns.

• Side project more successful than main project: Many projects have supplementary arms attached to them. It is not uncommon for one or more of these adjunctive components to prove more popular with the target audience than the core endeavour is, necessitating a pivot.

• Targeted market narrowing: This represents the concept of a pivot being triggered as a result of a significant contraction in the scope and/or volume of a startup’s intended market.

2.1.2 Internal Pivot Trigger Factors Internal triggers comprise the following: • Flawed business model: An unsound business model can exist in a number of

forms; one whose costs are too high, revenues too low or even one that demonstrates an inability to acquire sufficient customers. In such a scenario, a startup may have to pivot via innovative means in order to develop and implement a feasible substitute.

9

• Identification of a bigger customer need through solving an internal problem: Sometimes a startup realises, while endeavouring to improve their core product, that the element(s) requiring improvement is/are shared by the client. These circumstances may therefore demand that they reassess their product offering and pivot towards a solution that better meets their customers’ needs.

• Unscalable business: An objective of most digital startups is – or should be – to develop a product that has the potential to be scalable. If an organisation realises that their market offering is not thus, then they should rapidly pivot in the direction of one that is.

2.2 Scaling This part of the thesis examines some of the existing literature surrounding the topic of scaling in digital startups.

Huang et al. (2017) describes scaling as a generative process by which a venture’s user base increases significantly between two points in time. Sahay and Walsham (2006) on the other hand, offer the following, more expansive definition; ‘the process through which a product or process is taken from one setting and expanded in size and scope within that same setting or/and also incorporated within other settings.

Scaling often follows, and is enabled by, a successful pivot. The work of Brynjolfsson and Saunders (2010) as well as that of Henfridsson and Bygstad (2013) and Yoo, Henfridsson & Lyytinen (2010) explain that digital scaling differs from its traditional counterpart due to its ability to utilise and build upon existing digital assets. This can create an enhanced speed of scaling relative to that of yesteryear.

In their paper, Henfridsson and Bygstad (2013) continue their description of scaling by focusing on what they refer to as the scaling mechanism; i.e. the process by which startups are able to expand their reach via their digital infrastructure. As a result of offering enticements to potential partners – who then become actual partners – organisations are able to expand the reach of their digital infrastructure and reach previously untapped customers. The process is therefore self-perpetuating. It also creates a level of innovation space by providing the infrastructure elemental augmentation required to enhance reach.

Sahay and Walsham (2006) emphasise in their work that while numbers and size comprise elements of what scaling is, the phenomenon consists of far more. One must explore beyond the superficial and consider the dynamics which facilitate the spread, enhancement, scoping and enlargement of the heterogenous networks that encircle the technology. Finally, they demonstrate that the ability of an organisation to successfully scale may be dependent on such factors as technology, people and processes, as well as the contextual setting in which these influences are connected.

Picken (2017) outlines in his work several organisational requirements and actions – associated with what he terms rapid scaling – that a startup must possess and execute to achieve success. Compared to earlier points in a startup’s journey he emphasises the need to substantially enlarge the company’s resource base while utilising processes and alliances to expand the venture. This must all take place within the confines of a corroborated commercial hypothesis and sustainable plan. He also identifies the internal changes that

10

must occur – relative to the startup’s earlier form – for successful scaling to occur; specifically, modifications to such staples as configuration, regulation and process. Finally, he elucidates the objective of scaling as “rapid growth in order to achieve competitive scale and establish sustainable market leadership”.

Furthermore, Huang et al. (2017) identifies three distinct mechanisms in their work by which digital innovation can result in rapid scaling; namely data-driven operation, instant release and swift transformation.

2.2.1 Data-driven operation Data-driven operation refers to the process by which potential opportunities – and corresponding risks - can present themselves via high volume data analysis. This analysis can facilitate a startup’s ability to frame, hedge and monitor these potential upsides and downsides, before and during rapid scaling.

The concept of data-driven operation is underpinned by the following three distinct components: • Data profiling: The process whereby user data is employed by a startup to distinguish

and consider potential untapped areas of opportunity; for instance, clusters of new users.

• User hedging: This encapsulates the practice of a startup utilising as many diverse data sources as possible to produce a balanced risk profile for each innovative activity it considers to be part of its development.

• Fine-grained monitoring: This final activity is the scrutiny, at as granular a level as possible, of a startup’s user metrics to identify any areas of operational concern.

2.2.2 Instant release Instant release refers to the ability of digital undertakings to significantly curtail the amount of time required between concept generation and implementation. This can be achieved both by executing user trials concomitant with a new product launch as well as via rapid modifications to more established artefacts. Three foundational activities whose existence are required to facilitate instant release during rapid scaling are: • Launching: This refers to the coordination and focusing of internal resources, while

emphasising the primacy of data-driven operations, in an effort to energise the launch of digital innovations.

• Concurrent trialling: When a startup concomitantly trials multiple different innovations as well as different versions of these in an effort to generate multidimensional feedback from its user base, this effort is referred to as concurrent trialling. This exercise facilitates the implementation of data driven changes on a continuous basis.

• Reactive modification: This is a term that portrays the instantaneous amendments executed by a startup to an established product. These modifications are formulated and driven as a result of the behaviours exhibited by users.

11

2.2.3 Swift transformation Swift process refers to the scaling-up dynamics whereby a digital technology can either be recharacterised in terms of how it is perceived or repurposed and/or reconfigured to solve a new business problem. The three foundations of swift transformation are: • Contextualising core digital technology: Essentially a repurposing exercise

whereby a startup’s primary digital technologies form the basis for the conceptualisation of new innovations that facilitate exploration into novel and henceforth unexplored commercial territories.

• Projecting novel value-in-use: This involves startup management anticipating and demonstrating the unique advantages associated with a product/and or service that they believe will draw the attention and patronage of both novel and current users.

• (Re)defining identity: As startups alter the focus of their commercial activities as well as what they believe in and stand for, so of course must their unique character and distinctiveness change. This adjustment is known as defining or redefining identity.

2.3 Failure/Disintegration This section of the paper discusses some of the extant research around the topic of failure/disintegration in startups.

In the context of this thesis, and as explained by the work of Ries (2011), failure/disintegration may be considered as a possible outcome when an organisation fails to pivot, or alternatively pivots in the wrong direction. Many organisations celebrate failure, primarily for the resultant learnings that may be absorbed and repurposed; Da Silva et al. (2019) remind us of Thomas Edison’s wise words, “I have learned fifty thousand ways it cannot be done and therefore I am fifty thousand times nearer the final successful experiment”. Gartner and Ingram (2013) makes the case that entrepreneurs tend to agree with Edison and that failure is viewed as a positive. Clearly however, the failure of an entire startup – with its corresponding implications for employment, personal financial status, reputation and future endeavours – is not a desirable outcome, no matter what wisdom may be acquired from the event.

Startups by their very nature – which Cantamessa, Gatteschi, Perboli & Rosano (2018) define as high-risk and high-reward – are prone to a relatively high failure rate. Estimates vary on exactly what this rate is, however 90% is an oft quoted figure (Marmer, Herrmann, Dogrultan & Berman, 2012).

Failure may refer to both an organisation or solely to a particular product or service that they offer. Blank (2020) proposes the idea that the principal reason behind the failure of digital startups is not the technology that underpins their product offering but rather a lack of customers. This deficiency results in an inability to generate the relevant user data required to facilitate the development of both the product and the market at which it is aimed. Blank (2013) further explains that startup failure rates can be reduced by applying Eric Ries’s Lean Startup methodology; specifically, the fact that failure is an expected outcome that can be minimised via customer feedback leading to either constant iteration and adaptation of their concept or a pivot away from it to one with more potential.

12

The work of Ries (2011) debunks the mythical link between startup success and creativity, work intensity and determination. He further explains that the constraints associated with traditional management techniques and philosophies – such as fear of failure, lack of flexibility and short termism – are incompatible with startups and are therefore liable to contribute further to the underlying risk of failure. He argues that a new discipline of entrepreneurial management is required to overcome this deficiency. The work of Crowne (2002) identifies and discusses several specific causes for software startup failure, including, inexperienced developers, lack of a product owners and a dearth of cohesion between the organisational strategy and the components of the product offering.

The work of Nair and Blomquist (201) make the case that digital startups within an incubator setting – the objects of this work – are less prone to failure than those outside. This is a result of a number of strategies, specifically: • Focusing on the team around a scalable idea: While startups clearly require an

idea of some description to bring to market, incubators tend to focus more on the qualities of the startup team and whether they possess the potential to scale up the idea to a commercially viable level, as opposed to the perceived innovativeness and superiority of the actual product and/or service.

• Business model validation: Incubators are very aware that, more often than not, the cause of a startup failure lies within its business model. They therefore encourage resident startups to repeatedly, and as early as possible in their lifecycle, take a minimum viable product to potential customers, obtain feedback and incorporate that feedback. This process pressure tests their business model.

• Business model development: As a startup progresses through its incubator journey, it is imperative that it demonstrates proof of concept at various points in order to maintain its funding streams. Incubators are able to provide startups with the myriad of resources they require to achieve these points, the support to pivot to a new idea if the concept cannot be proven and facilitated access to streams of finance to support them while they undertake these activities.

• Business model scalability: Most incubators want their startup residents in and out of the premises as rapidly as possible. Two scenarios can result in this objective being met; the ability of a business model to scale and the inability of a business model to scale. Incubators are able, through coaching, mentoring, and network access, to facilitate the construction of a scalable and feasible commercial model.

3. Research design The design of this research project is summarised in the following sections; research approach, case study elucidation & sampling method, data collection, data analysis and ethical issues. Each of these parts will now be discussed and examined with, where appropriate, a specific emphasis on the factors – and supporting publications – that justify the utilisation of the methodological components that were chosen over various alternatives.

13

3.1 Research approach This section will explain and expand upon the research approach that was adopted during the authoring of this work.

This work is qualitative in nature due to the fact that it is focused on texts and does not concern itself with topics of a numerical nature; hence it cannot be quantitative. Philosophically, it belongs to the interpretivist school of thought since, as explained by the work of Holmström and Sawyer (2011), it seeks to unmask the interviewees’ views of reality; specifically gathering their thoughts on the enabling and inhibiting factors that impact upon the evolution of their digital startups. It does not therefore fit the profile of a study that would lie within either the positivist or critical schools of thought, as defined by both Myers (2013) and Dubé and Paré (2003). Since the study examines three different incubators – and therefore three different cases – the research method utilised is that of multiple-case study, as recommended by Gustafsson (2017) and defined by the same author as a method that “explores real-life cases over time, through detailed, in depth data collection involving multiple sources of information”. However, an adjunctive definition of the method comes from the work of Mills, Durepos & Wiebe (2010) who define the method as “case study research in which several instrumental bounded cases are selected to develop a more in-depth understanding of the phenomena than a single case can provide”.

Dubé and Paré (2003) comprehensively explain in their work how this method can be effectively combined with interpretivism in an IS research setting. Multiple-case study was selected as it appeared to represent the most appropriate methodology to best meet the objective of the research; specifically, to both consider the extant research on some of the enabling and inhibiting factors in digital startup evolution while also narrowing that focus specifically to an incubator setting. The work of Hannah and Eisenhardt (2018), supported by that of Eisenhardt, Graebner & Sonenshein (2016) explains the appropriateness of multiple-case studies for questions of process, which this study clearly represents. Myers (2013) explains in his work that case studies are appropriate for modern-day and actual scenarios where the researcher does not seek to – and in any case is unable to – control any aspect of the research process. He further makes the point that a focus on how and why something takes place is a key characteristic, further bolstering the case for this method.

Case studies are not without their critics. The work of Dubé and Paré (2003) communicates an exhaustive list of deficiencies related to the format, although many of these relate to poor execution, as opposed to the method itself. Myers (2013) elaborates on the fact that understanding which case study data to retain and which to discard can be a challenging business; for those possessing a scarcity of experience, everything may seem relevant.

The advantages, disadvantages and potential employment of a number of other research methods – comprising Grounded Theory, Action Research and Ethnography – were considered, with the assistance of the work of Yin (2018) and Myers (2013) before the final methodological choice was made. Within the parameters and constraints of this study, while grounded theory may offer a certain intuitive attraction to the inexperienced researcher, its multi-layered intensive coding requirements and focus on theory

14

development make it an inappropriate choice. Action Research on the other hand, while lauded for its relevance characteristic, did not fit with the research question this work seeks to answer; it’s focus is not on intervention and solving a problem, both of which are key features of action research. Finally, ethnography, while more likely than multiple-case study to provide data of a richer and more in-depth character, would not have been a good fit with the objectives of this study due to its time constraints and potential requirement for a publishing format more voluminous than that allowed by the confines of this thesis.

3.2 Case study elucidation and sampling This section will expand on the details of the multiple-case study undertaken while also describing the sampling techniques that were employed during the thesis development.

The research phase focused on thirteen startups being nurtured by three business incubators in Umeå, Sweden. The incubators partnered with for this work were Umeå Biotech Incubator (UBI), BIC Factory (BICF) and eXpression Umeå (EU). While there are elements of similarity between these organisations in terms of their geographic location and ownership structure, each is focused on very different industries. Typically, startups stay within the incubators or a period of two years, however there are opportunities to extend this period if both parties are in concordance upon the merits of doing so. The multiple incubator option – as opposed to simply studying one entity – was therefore chosen in order to generate data with as wide a range of utility as possible. Contact with key stakeholders in each startup was facilitated by senior managers within each incubator and initial contact with the incubators was enabled both by myself and Professor Jonny Holmström of Umeå University’s Informatics Department.

UBI positions itself as a biotech incubator operating within the Northern Swedish life science industry. They mentor successful applicants through a multi stage development process – including concept verification, financing and growth – designed to ensure they achieve commercial success upon graduation.

EU focus their efforts on a wide and diverse range of subdisciplines within the creative and cultural industries. They aim to advance their partner organisations via the provision of four distinct developmental programs – Express, Creative, Future Retail and Design – whose implementation are enhanced via the innovative and stimulating environment within which day-to-day activities take place.

BICF is a general business incubator specifically focused on young entrepreneurs. They provide two years of working space lease, commercial mentoring, as well as the possibility of their startups constructing and benefiting from a robust business network with a range of potentially useful partners.

The constrained resource volumes – of both a timeous and manpower related character – associated with attempting to ascertain some of the enabling and inhibiting factors in digital startup evolution led this thesis, as suggested by the work of Etikan, Musa & Allkassim (2015) in the direction of nonprobability sampling techniques. After further consideration of the advantages and disadvantages of a number of techniques grouped within this sampling category – and exhaustively explained by the work of Bernard (2017) –the decision was made to proceed with purposive sampling. Tongco (2007) writes that

15

purposive sampling, which they explain is also referred to as judgement sampling, is an appropriate choice where the informants are chosen as a result of certain characteristics they possess; in this case the fact that they are senior managers and or/owners of startups that have pivoted and exist within an incubator. This choice is further reinforced by the work of Bernard (2017) who specifically cites the applicability of purposive sampling for various types of case studies, while also pointing out the necessity for utilising this technique in populations whose members do not exist in abundance. While there may be plenty of startups globally, the scarcity of such organisations within incubators in a narrowly defined geographic setting such as Umeå appears to further reinforce the aptness of this technique.

Alternative nonprobability sampling techniques that were considered – and that are further discussed in the voluminous work of Bernard (2017) – included quota sampling, convenience (or haphazard) sampling, and chain referral (comprising snowball and respondent-driven) sampling. While quota sampling – as the name suggests – is appropriate when a researcher is looking to obtain specific proportions of populations, that was not the case here and therefore it was not deemed suitable. The underlying principle of convenience sampling – namely interviewing anyone who can be easily accessed – was also not considered a suitable technique due to the very specific focus of this study, as explained in the research question. Finally, chain referral – where reliance is placed on disclosure and subsequent referrals by interviewees of potentially suitable subjects – appeared to leave too many variables to chance in what was a time constrained project.

3.3 Data collection This segment of the thesis examines the data collection principles that were utilised during the thesis development.

The primary data sources were a series of thirteen semi-structured interviews with either startup senior managers or owners. Myers (2013) explains that it is almost an essential requirement to employ interviews when conducting case study research. This question format was chosen to allow for both the flexibility and creativity on the part of the interviewer and free expression on the part of the interviewee, that Diefenbach (2009) explains it facilitates. The applicability of semi-structured interviews specifically for small scale research is endorsed by the work of Drever (2003). The interview script was divided into both individual questions and three sections as recommended by the work of Miles and Huberman (1994).

The interviews were conducted via Zoom as a result of the ongoing global novel coronavirus crisis. The audio-visual content of the interviews was recorded in preparation for thematic analysis. Several of the interviewees were aware of and specifically referred to the work of a number of authors whose publications underpin this thesis. However, in an effort to minimise the influence of these authors upon their responses, I did not proactively raise such matters with any participant.

The first part of the interview guide focused on the background of both the interviewee and other members of their senior management team. Specific enquiries were made regarding whether their digital startup had ever experienced rapid scaling – definitions of

16

this phenomenon were shared at this point in an effort to facilitate comprehension – or failure. The second part of the discussion concentrated on multiple pivoting related topics. At the commencement of the second section of the interview, definitions of pivoting were shared with the participants in order to reduce confusion and maximise understanding of the meaning of the term. The concluding section discussed their experiences within the incubator, with an emphasis on both pivoting as well as the areas of their business which had been most impacted by the influence of their respective incubator partners. The participant interviews varied in length from forty to seventy-five minutes with a mean duration of forty-nine minutes. More specific information regarding the participants and their interviews is detailed below in table 2. In an effort to protect the anonymity of the digital startups who participated in this study, and bearing in mind the relatively small size of the sample, the specific fields within which each of these organisations operate has not been identified. However, the industries represented comprise CE marking, creative, retail, internet-service provision, life sciences, media production, nutrition and software.

Participants Interview Duration (min./sec.) Incubator

Startup A 46.19 BICF

Startup B 49.51 UBI

Startup C 45.57 BICF

Startup D 49.57 EU

Startup E 49.24 BICF

Startup F 48.37 BICF

Startup G 50.20 UBI

Startup H 75.48 UBI

Startup I 40.36 BICF

Startup J 58.24 EU

Startup K 39.44 BICF

Startup L 48.24 EU

Startup M 40.05 UBI

Table 2: Interview details

The fact that, as explained above, the participants were not chosen on a random basis – as well as the circumstance of this being an interpretive study raises the spectre of potential researcher bias. Mehra (2002) explains in their work that managing bias is considered to be an important factor for researchers to consider, while the work of Collier and Mahoney (1996) explains that not doing so can cause the validity of research to be undermined. The work of Holmström and Sawyer (2011) explains that the phenomenon of bias may negatively influence the results of a case study – due to the relative intimacy of the relationship between researcher and subject as well as the frequency of contact – as a result of respondents being less than frank and honest in their responses to questions. This can be due to an attempt to protect either the participant or their organisation. I am confident that, in this thesis, the potential for bias was minimised both due to the brevity and scarcity

17

of the interactions between myself and the interviewees as well as the strict pledges of confidentiality provided to them, relative to the identification of both them and their organisations in the final draft of this thesis.

3.4 Data analysis This section of the thesis will present the cognitive processes that were reflected upon when analysing the gathered data.

Thematic analysis, according to the work of Braun and Clarke (2012), is ‘a method for systematically identifying, organizing and offering insight into patterns of meaning (themes) across a data set’. To achieve the objective of this thesis – identifying the enabling and inhibiting factors in digital startup evolution within an incubator setting – thematic analysis was considered to be the most appropriate analysis method. There were several reasons for this. The author’s background – with its relatively limited experience involving the use of this analytical method – combined with the requirement for common theme identification were important factors. The ability of thematic analysis to offer flexibility and simplicity – attested to by the work of Braun and Clarke (2006) – was also an important consideration. Finally, the fact that the material being analysed can be considered as in-depth interviews makes it, as explained by the work of Guest, MacQueen & Namey (2012), a suitable media format for thematic analysis to be applied to.

Other data analysis methods – described in the work of Bernard (2017) and Leech & Onwuegbuzie (2008) – that were briefly considered and subsequently discarded from consideration included hermeneutics, semiotics and narrative analysis. While each has distinct advantages, these plusses were not relevant to the research question under consideration Hermeneutics is undeniably a useful technique when the analysis goal is to gain an enhanced understanding of people, however the focus of this thesis was much more on individual commercial entities; it was therefore not considered to be a fit and proper method for the job at hand The interpretation of signs and symbols was not the focus of this work and therefore semiotics was not considered appropriate. Finally, in the case of narrative analysis, the capturing of stories was not considered the optimal data collection technique for this scenario. Narrative analysis was therefore also not the best fit.

The primary objective of the analysis phase was – in keeping with the nomenclature of the chosen analytical methodology – to extract themes that could add value to the process of addressing the research question. During an initial review, a number of preliminary codes were identified and applied manually. A more in-depth review was then undertaken utilising the ATLAS.ti analytical software package. The identified themes were incubator influence, pivoting, scaling, failure/disintegration and

. Each of these is strongly related to the others and a diagrammatical representation of these relationships is shown in figure 1.

An additional objective was to combine the key findings from the interviews with any corresponding content from section 2, in order to start developing some ideas around the research streams of pivoting, scaling, and disintegration/failure. A journey of iterative coding was therefore embarked upon, during which a number of embryonic ideas were

18

recognised and compared & contrasted against the previously published research of other authors.

3.5 Ethical issues This section will examine the ethical issues that were considered during this endeavour.

The work of Bernard (2017) explains that all research has an ethical factor that must be considered, with significant potential negative outcomes being at least a possibility if they are not. Therefore, in accordance with the writings of Ritchie, Lewis, McNaughton Nicholls & Ormston (2013), I endeavoured to consider any potential issues that might arise before, during and after the interviews.

All potential study participants (fourteen in total) were contacted twice via email to request an interview. If no response was received, contact was discontinued. Fortuitously however this was an exceptionally rare event as only one potential interviewee failed to respond after two contact endeavours. The purpose of the study – as well as the rationale for it – was clearly communicated within the emails in combination with a comprehensive explanation of exactly what I was seeking to obtain from the interviews. The expected duration of the interview was also disclosed. Once a time and date were agreed, a zoom invite and calendar appointment were sent to the interviewees. I was previously acquainted with one of the startup owners as well as some confidential information regarding his organisation, as a result of working with him on a project that formed part of a previous course within the Masters in IT Management curriculum. However, I am not aware of any ethical issues that could result from this previous experience; the previous work and the information I had access to in no way related to the topic under discussion in this thesis. I therefore have no reason to believe that either the integrity or accuracy of the results discussed in this thesis have been impacted in any way.

At the beginning of each interview, the purpose of and rationale for the thesis were communicated again and participants were asked if they had any questions that I could answer. Consent to record the discussions – under the proviso that they would not be shared with anyone other than academic staff at Umeå University – was sought and obtained. Verbal consent to utilise any data generated in an anonymised manner was sought and obtained. The anticipated duration of the call was also explained again. Every effort was made to make interviewees feel at ease. Interruptions were kept to a minimum and not initiated by me unless it became clear that the conversation was veering completely off topic or that the question had been misinterpreted. At the conclusion of the call, those taking part were asked if they had any additional questions for me or any final thoughts on the topic that they wished to contribute to the study

The pledge of confidentiality that was promised to the interviewees has been respected and their privacy has been protected as far as possible during the reporting of the results. Clearly with such a small sample size, complete confidentiality cannot be absolutely guaranteed. Several participants requested a copy of the finished thesis and I committed to ensuring that each of them will receive this, assuming a successful outcome.

19

4. Results This section of the thesis encompasses the empirical findings of the research. Five distinct themes related to the research question emerged from the interview content coding exercise, all of which exhibit robust interdependent relationships with each other. These five – incubator influence, pivoting triggers, scaling, failure/disintegration and inertia – and summaries of the results related to them are presented and in greater detail below.

4.1 Incubator influence This section will relate the opinions expressed by the interviewees regarding the influence exerted by the incubators whose auspices they operate within upon the operational dynamics of their digital startup.

All thirteen of the organisations interviewed were, to a greater or lesser extent, satisfied with at least some aspects of their incubator experiences. They appeared to understand the inherent value underpinning their association with these umbrella organisations. Of specific interest to the startups were the wide range of external (and complimentary) services that they had access to as a result of their incubator placements. Mention was made of such services as accounting, financing, legal (contract and IP), marketing and regulatory. Internal coaching was generally considered to be of equivalent value to the interviewees as the external services and included advice and input on such matters as brainstorming, investor management, IPO preparation, motivation, presentation skills, price-setting, prioritisation, productivity optimisation and sales training. Finally, the public exposure, networking opportunities – both internal and external in character – that were enabled via the incubators, as well as the financial upside associated with such benefits as subsidised office space, were frequently mentioned. Startup G summed up their thoughts on the importance of incubation to their business by stating “without them, from the beginning, I don’t think we would have managed to take this to where we are now”. This sentiment was supported by startup C who opined that “I’ve been very happy with the support I’ve received. The coaching has made a real difference to my company.”

However, four of the interviewees also expressed a degree of dissatisfaction with their incubator mentorship related to the areas of specific and relevant expertise and skill-sets. While the startups were grateful for the wide range of coaching available, a recurrent theme raised by a minority but vocal group of startups was the lack of relevant experience possessed by some managers and/or coaches within the incubators. There was an underlying belief that the specialised nature of these organisations meant that anyone possessing experience of a more general nature – or derived from a separate industry altogether – was not best suited to providing effective counsel. Startup D for instance felt that “The most valuable form of mentorship you can get is from people who have done what you’re trying to do and I kind of feel that there’s none of them available. This was reinforced by startups C and H who opined respectively that “It’s always people, who maybe in a best case have been running a company maybe fifteen or twenty years ago, but maybe don’t have much experience of today’s environment” and “The ones that have the experience, they don’t work for a governmental funded agency in any country”.

20

All thirteen of the interview subjects reported a lack of mentorship related to pivoting and pivot triggers. While specific mention was made of incubator coaches offering occasional aid with strategy development and in turn challenging that same strategy, no organisation was able to relay a scenario where an incubator manager and/or coach had suggested the execution of a pivot.

The summary of what was gathered from this section of the questionnaire is that, as demonstrated by the provision of many of the services referred to above, the focus of the incubators is on assisting the organisations both tactically and operationally; with limited desire on the part of the coaching team to disrupt the strategic status quo. Their efforts and corresponding influence in this matter are both considerable and greatly appreciated by the startups who are impacted by them. However, their influence does not extend to pivoting. Strategy is discussed however the adjustments and refinements associated with pivoting are not a topic of conversation between the incubators and their incubatees. Examples of responses reinforcing this point included, from startup A “I wouldn’t say they have done that, they have left that to me”, from startup E “They haven’t given any advice like that, the advice we usually get is to focus more” and finally, from startup F “I don’t think so, no, it’s been more about pushing me forward than rethinking”.

4.2 Pivoting This section records the participants’ recollections upon both the pivoting activities they have undertaken as well as their opinions regarding identification of the root causes that led to changes in organisational strategy.

All of the organisations interviewed – with the exception of digital startup I – claimed to have executed at least one change of strategic direction that they felt qualified as a pivot. While each organisation claimed to have executed at least one pivot, a number of these would not qualify, even under the wide variety of definitions that exist in the extant literature. These events have not therefore been included in the results section

The range of pivots offered as examples were extremely disparate. Examples of activities specifically related to refining the strategy of a functional area of the business included such items as platform change, product divestiture, new market entry, marketing digitalisation, business model innovation, expansion/modification of service/product, replacement of key partner(s), director removal, pricing strategy refinement, new customer focus, sales channel addition/elimination, production insourcing and substituting consultants for employees.

One comprehensive pivot undertaken was that of startup A who completely reconfigured their customer facing platform in order to be able to offer their clients a level of portal personalisation that their previous infrastructure was not capable of at a moderate cost. They described the implementation process as “very successful; it’s been above our expectations”.

Startup E’s pivot was radically different from that of A and focused on the people who undertake the development work upon which their organisation relies. They moved away from partnering with a Swedish based IT consultancy that assigned their work to a wide and inconsistent range of developers and instead partnered with a Vietnamese based

21

consultancy that recruits and manages staff – that are selected by the startup – specifically and exclusively for their projects. They believe this arrangement offers them a number of advantages and explained “We got developers that are really good, they feel they work for us. In their minds they are our employees. That’s what we want to achieve, we don’t want them to be employees for a consultancy”.

Company I on the other hand claimed to have been at the point of pivoting several times. They took the view that their original (and current) course remained the most efficacious one for the financial success of their organisation; “We have been on the border many times, but actually we have believed very hard in our main plan”.

Finally, startup L executed a pivot that involved a change in their positioning within their industry’s supply chain continuum. They moved their entire business model from being a media developer whose work was then distributed by others to one whereby they distributed content developed by other organisations.

4.2.1 Pivoting triggers The factor(s) identified by the digital startups as having triggered the pivots were also extremely diverse. These included such influences as lack of sales, customer feedback, customer needs, novel coronavirus, excessive costs, supply chain disruption, loss of vendor oversight, creative inertia, regulatory legislation and lack of relevant internal skillsets.

Customer feedback was a recurrent theme during the interviews, with three organisations – startups B, C and G – all presenting their customers with either a minimum viable product (MVP) version of their product or a detailed explanation of it in order to gather feedback.

Financial metrics, including cost reductions or revenue increases, were cited by several of the participants as the key rationale underpinning their pivots. A good example of this was startup L who, in addition to the pivot cited above, also moved their distribution channels from physical settings to an exclusively online presence in an effort to reduce their cost base and increase their profitability.

Startup D explained that the presence of much of their vendor partner network in a myriad of developing countries resulted in the tendency for communication and fulfilment issues to arise, necessitating a constant change in their supply chain strategy.

Startup H raised a scenario where they realised that they had niched their innovative patient focused technology and that its utility could be expanded across a far wider population than previously envisioned. They expounded, “the solution was true to pretty much any context within healthcare; it opened up possibilities that we hadn’t really been thinking about.”

European Union mandated regulatory legislation changes resulted in enforced changes – and consequently a significantly expanded cost burden – to the product development strategy of both startups G and M. They explained it thus, “Unfortunately it’s something that’s completely out of our control but it has demanded a complete rethinking of our strategy”.

The colossal impact of novel coronavirus on the social and economic wellbeing of the planet has, via necessity, undoubtedly triggered some very recent and swift pivoting on the part of a number of Umeå based startups. Some of the organisations interviewed were

22

prepared to acknowledge the disruption this had caused to their business while acknowledging the opportunities it had also created. Startup C for instance had transformed its key marketing tactics from a print-based endeavour to a multi-faceted and integrated digital campaign. However, their ability to attend key customer meetings and conferences had also been curtailed. While they appreciated the cost reduction opportunities offered by their transition to digital marketing, they felt that this benefit was overshadowed by the inability to interact directly with their customers.

Other digital startups acknowledged the fragility of startup organisations that were experiencing financial difficulty after only a few weeks of disruption; startup B commenting “it makes you wonder how these companies are being run if they’re becoming insolvent after only a few weeks….” It must of course be remembered that most of these interviews took place in April 2020, before the enormity of the COVID crisis had been fully revealed to the world.

In summary, while the majority of digital startups do seem to be pivoting, not all who believe they are executing pivots, in reality are. The breadth of both the pivot types and the triggers that underpin their execution are diverse and relate to a number of different aspects of the environment in which the companies exist and operate within.

4.3 Scaling This section will narrate the findings of this thesis regarding the scaling activities – and lack thereof – undertaken by the digital startups that were interviewed.

With the exception of Startup B, none of the digital startups interviewed had experienced rapid scaling of any variety during the organisational lifespan. A number of the organisations felt that it was simply too early in their lifecycle to do so – or the period of time post pivoting was too short – for any scaling activities to have taken place. However, by no means was this the case for all. In other cases, simple manpower issues meant that scaling was not practical. Examples of this were exhibited by the digital startups – whose chief marketable commodity is time – who only have one employee/owner whose hours can be billed and have no plans to increase their headcount.

Startup A, whose pivot is discussed above, acknowledged that one of the primary motivators underpinning their decision to pivot to a proprietary platform that facilitated client personalisation was the potential ability to scale their business in the future. They contrasted their previous business model of creating stand-alone educational courses, which were owned by their clients, with their current activities that are centred around creating proprietary content which can be branded for their client’s needs. Specifically, they stated “Our main focus will be to have ownership of the courses, and that’s because of the scalability”.

Startup D emphasised that the characteristics inherent in some of the components that comprise their business model – which facilitates access to physical products via digital means – meant that their ability to successfully scale would be severely constrained. In their case, the current ability of their supply chain to rapidly increase manufacturing output would be problematic, causing them to assert “you are kind of limited by what your production partners can produce”.

23

Startup E implied that, as things currently stand with their business, they would prefer not to undertake rapid scaling. This is due to a lack of internal management expertise that they believe would inhibit their ability to manage such an undertaking. They explained, “That’s what we would like to achieve, but we wouldn’t know where to start with something like that.”

Startup B on the other hand, over the last twelve months, has seen exponential growth in their sales numbers. They attribute this successful outcome primarily to their ability to gather and publish long-term data proving to their potential client base that the operation of one of their proprietary solutions – which at the time only existed as an MVP – was able to scale up in a manner that had not previously been recognised or understood. This digital startup also emphasised the internal challenges associated with publishing information based solely upon data generated from an MVP and not a final product; specifically, from the more technically minded team members who believed such actions to be a significant error of judgement.

A number of the digital startups interviewed are not scaling in any sense of the word. There are several reasons for this; a number of them either do not have the ambition, have not formed the necessary partnerships, lack the skills and capabilities, or do not feel that this is the right time in their organisation lifecycle to undertake such an exercise.

4.4 Failure/Disintegration This section will focus on the concept of digital startup failure/disintegration as discussed with the interviewees.

None of the digital startups interviewed were prepared to admit to any failure of their organisations either in their current form or in a previous iteration of the organisation. Moreover, none of the senior managers interviewed had experienced any failures in previous or concomitant commercial activities. Far more than any other question, this part of the interview provoked the least amount of discussion and willingness to engage.

Of key significance within this topic was of course the definition of failure/disintegration held by the respondents. Some, such as startup E for instance, seemed to associate the concept with the financial failure of the enterprise, characterised by a failure to keep trading as a result of being unable to pay their debts. Others, such as startup D however, felt that failure could be epitomised simply as a result of arriving at a point in the organisation’s journey where it had become clear that their idea was not going to enjoy widespread commercial success; possibly due to the relatively young age of the company. Finally, there were who acknowledged that certain initiatives within their digital company had failed, but these had not led to the failure of the entire organisation; specifically, startup K described how “I have had failed projects but nothing more”. This was reinforced by the experiences of startup H who expounded that “there’s stuff we did that we look at now and think, how could we…”.

Some respondents were eager to emphasise the previous successes that either they or their colleagues had enjoyed when the matter of failure was raised. This phenomenon was exhibited both by startup M who explained that “my cofounder he’s been in two successful

24

startups, so I guess it’s a good thing” and startup I who started that “I’ve been involved in a number of successful companies before this one”.

Furthermore, a number of the organisations interviewed claimed to have either been aware or even been in close proximity to digital startups who had failed. A member of startup L’s team had experience of actually supervising such companies via his previous employment by an incubator - in addition to managing his own company – and explained “I worked in an incubator where we had success stories and other not such successful stories”.

Startup L also emphasised that they believed failure was comparatively less prevalent within the creative and cultural industries as a result of the relatively low level of overheads and debt incurred by many digital startups within these fields. He further explained that where external investments were accepted by startup entities within these industries it was often in the form of non-repayable grants; “if they were to take on money it will usually come as soft money and not loans or VC, so that’s why there is a less failing culture within creative businesses.”

Finally, a partial contradiction to the views expressed earlier in this section emerged when I proposed the theory that failure should be seen – as it is on other parts of the world - as at least a partial positive due to its potential learning benefits. Most interviewees agreed with me that commercial failure could indeed be a positive thing. However, whether this was a true reflection of their views or simply the adoption of a nonconfrontational position is difficult to ascertain with any degree of certainty.

Failure/disintegration appeared to be a rare (or even extinct) creature in this part of the interview. Although different parties assigned different meaning to the word, it simply was not an outcome that triggered a great deal of engagement and actually in some cases led to (brief) discussions concerning either digital startup successes or failures experienced by other parties. Clearly, lack of debt also plays a significant factor here.

4.5 Inertia This section discusses the topic of inertia which was prominently displayed by the participating digital startups.

One of the most common, interesting and unexpected themes to emerge was that, despite almost every organisation claiming to have executed a pivot, only one of them had subsequently (or even previously) experienced scaling or failure/disintegration. Among the participants, the default action appeared to be to maintain a growth/lack of growth trajectory similar to that exhibited pre-pivot. This relative lack of dynamism was perhaps best exhibited by startup E who, as a result of their current lack of ability to scale, explained that pursuing anything other than inertia would “be a problem for us right now”.

One rationale for inertia that emerged during the interviews revolved around the perceived difficulties associated with funding an escape from such a state. Some of the participants are entirely self-funded by the owners or, as in the case of startup F, close friends and family. However, others – such as startups G and M for instance – rely on such sources as ‘soft’ money, venture capital and additional investments from their board of directors and current shareholders in order to continue to operate at their current levels of

25

activity. In the former funding scenario there was a view expressed by such startups as C and E that the currently suboptimal global economic environment necessitated a prudent approach and that their focus should be on survival, albeit in a state of relative inertia. Startup C declared “Now, with the coronavirus, I’ve had to reduce my costs as much as possible to be able to survive”, while Startup E, who had planned to expand their offer to customers outside the European Union asserted that “until coronavirus passes and we’re able to seek investments to strengthen our financial position, those plans are on hold”. In the latter funding scenario, there was a belief expressed that the same environmental factors would make adequately funding an escape far more challenging than normal. Startups G and M were acutely aware that they might need to cede control of their enterprises in order to finance an escape from inertia and in the words of startup M, ‘that’s not a decision that I’m in a big hurry to take”.

Startup F explained that they were reluctant to compromise their creative principles and aspirational branding, by undertaking endeavours that they felt were in conflict with the work that they most wanted to do. In fact, they had actively declined work that they felt interfered with these goals. They also expressed their reluctance to interfere with what they the felt to be an acceptable level of work-life balance and recognised that for the company to achieve its full potential, it would be necessary for this equilibrium to be disrupted. They averred, ‘If it was all about the growth and the money, I wouldn’t be here in Umeå, I would be in Stockholm.

Finally, it can be argued that the findings relayed above on the topic of (lack of) incubator influence are also relevant to the topic of inertia. While no participant specifically cited a link between their lack of scaling or failure and the behaviour of their respective incubators, it seems plausible and logical that the dearth of guidance regarding pivoting could be extended and connected to the scarcity of these dynamics in a post pivoting scenario.

In summation, inertia emerged as a common and prevalent theme of the results. It appears to exist for a diverse range of reasons including finance, creative principles, the novel coronavirus and lack of motivation. Inertia’s existence is strongly related to and influenced by the presence of the other emergent themes in this thesis.

5. Discussion This part of the thesis will explain how the empirical findings unearthed by this work

contribute to the research area of digital entrepreneurship. Furthermore, this section will compare and contrast these results with the extant literature, while at the same time suggesting potential opportunities for future research contributions. It will also briefly reflect upon the limitations of the study.

The work of Islam, Buxmann & Ding (2017) explains that “digital technologies offer multiple opportunities for firms but they also involve many challenges”. While there is a small volume of extant research in existence regarding the factors that, in digital startup evolution, both facilitate and impede access to these opportunities, there appears to be a scarcity of significant publications – with the exception, of course, of the excellent work of

26

Mankevich and Holmström (2016) – that examines this topic in the context of an incubator setting. It seems logical to me that the earlier such knowledge can be obtained and applied to digital startups, the more attractive the long-term prospects may be for them. This may, in turn, also result in favourable outcomes for the geographical setting in which they find themselves, whether it be through enhanced tax receipts, increased employment, improved investments & collaborations and a whole range of other valuable societal benefits. Assuming this premise to be true, incubators represent an excellent opportunity to both study and implement the resulting research outcomes. Therefore, to recap, the research question which this work seeks to answer is: What are the enabling and inhibiting factors in digital startup evolution within an incubator setting.

It may contribute to a more thorough comprehension of this thesis, since the foundation of this work is digital entrepreneurship, to briefly broach again the topic of the key differences between traditional entrepreneurship and its digital successor. Clearly, as explained by the work of Nambisan (2017), Henfridsson and Yoo (2014) and Fang et al. (2018), while there is a degree of common ground between the two domains, there are also differences of a not inconsiderable magnitude. These include a novel collection of complementary assumptions; for instance, entrepreneurial practices and results possessed of enhanced fluidity and/or fewer constraints as well as entrepreneurial agency characterised by reduced predefinition and increased distribution.

This section is further divided into a number of sections. 5.1 will discuss, in greater detail, the Ideation Dynamics Model represented by figure 1, while section 5.2 will expand upon the topic of inertia identified earlier in this work. Section 5.3 will contribute to a greater understanding of pivoting, while 5.4 will endeavour to achieve the same objective in relation to failure/disintegration. Finally, section 5.5 will examine a number of implications for practice by both incubators and digital startups.

5.1 Ideation Dynamics Model (IDM) This section will examine the various elements – and their interrelationships – contained within the IDM, represented by figure 1 below.

The IDM, in essence, represents a series of paths that a digital startup can follow. It was named thus as it represents the dynamics that an idea should adopt as it progresses through its life cycle from initial concept to commercial success. Two of these paths – inertia and failure/disintegration – represent barriers to successful outcomes. Utilising the language of the research question, these are factors that inhibit the evolution of digital entrepreneurship. Scaling on the other hand represents an enabling factor, one whose presence is essential in order a digital startup to achieve its full potential. The schematic is intended to communicate the concept that a future product and/or service – represented by the idea box – should be tested by a digital startup with its target audience to ascertain whether a pivot is required. This helps to ensure that the idea, when it takes the form of a product or service, meets the needs of its intended market. Following on from the pivot, the results suggest, in combination with the extant research, that there appear to be three options for the organisation; scaling, inertia or failure/disintegration. Scaling is clearly the

27

most favourable outcome and failure/disintegration the least desirable, while inertia occupies the middle ground in a position far closer to the latter option than the former.

Figure 1 proposes that whichever of the three paths is chosen, the process should exhibit the characteristics of a continuous closed loop system. Therefore, whether the idea is scaled, suffers from inertia or fails/disintegrates, the digital startup should always be searching for its next concept to evaluate and guide through the IDM. The experience and learnings that will undoubtedly be obtained from each product development lifecycle will feed back – and hopefully offer considerable potential benefits – into the succeeding development cycle.

As far as I have been able to ascertain, while each of the three choices – scaling, inertia and failure/disintegration – have been examined, either as stand-alone items or in different configurations to the model I present, no similar model has yet been published in any of the extant digital entrepreneurship literature that I have reviewed for the purposes of this work. I feel therefore that the IDM makes an important contribution to the existing Lean Startup model of Ries (2011) by introducing an enhanced level of sophistication and complexity, and therefore hopefully facilitating more accurate modelling outcomes. Awareness, knowledge and understanding of this simple and basic model has the potential to provide important developmental support to budding entrepreneurs as well as needed guidance to incubator managers and coaches. This support may – even if just by raising awareness of the existence of these phenomena – help to maximise scaling while minimising inertia and failure/disintegration. This in turn may lead to economic benefits for digital startups, incubators and the wider macroeconomic environment of the geographical location within which the direction is implemented.

Figure 1: Ideation Dynamics Model

28

5.2 Inertia This section will discuss the topic of inertia – a component of the IDM – that was identified as an outcome of digital entrepreneurship endeavours, during the research phase of this thesis. Very little extant research appears to exist regarding the subject of inertia within the digital startup community.

It is however interesting to consider this unexpected finding in the context of the work of Ries (2011) whose views regarding the organising logics of digital startups are very simple and well documented; their options are either scaling or failure. Inertia is not an entrepreneurial factor that appears to be worthy of featuring in his thinking. However, this research clearly suggests the phenomenon exists within at least some digital startups.

When examining the reasons for the existence of inertia within digital startups, one may consider the work of Kaplan and Henderson (2005) who suggest that there are links between organisational inertia and internal incentive systems. On other words, if there is an absence of motivating factors to exert excessive efforts aimed at breaking free from inertia it is not unreasonable to expect it to represent the organisational status quo for some digital startups. It seems therefore at least possible that the relatively comfortable environment in which some of the study participants find themselves – a situation where they have access to a significant breadth of training, advice, development and financial resources, combined with inexpensive or free office space – may result in a situation where the temptation to embrace operational and strategic stagnation becomes irresistible.

The work of Ruef (2006) suggests that what he refers to as entrepreneurial inertia may result in an increased likelihood of boom and bust cycles and therefore, potentially destabilisation of entire industries. This can happen due to, for instance, the perception by investors that industries plagued with inertia do not offer a favourable return on investment. In addition, the work of Gong et al. (2009) advances the view that the presence of what they refer to as “absorptive inertia” – an absence of willingness to absorb relevant information from external network sources – may be directly related to and influenced by the level of experience within the organisation. They also found that the phenomenon was linked to the perception of arrogance by some startup customers. Since a number of the interviewees in the study were relatively inexperienced and some level of inertia was seen in almost all cases, it would seem to be disingenuous to suggest that the findings mirror these results. All of the study interviewees exhibited a positive dearth of arrogance in their interactions with me. However, for the more experienced startup managers in the group it may well be the case that inertia is caused by a lack of knowledge transfer from outside the organisation.

5.3 Pivoting Some of the extant research that deals with pivoting and the triggers that cause it will now be examined in combination with the research findings in this part of the thesis. A deeper understanding of pivoting than that previously discussed can be obtained by deep immersion in the various writings of Eric Reis who, while not inventing the concept of a pivot, was certainly the first author to label it as such. The work of Ries (2011) presents pivoting as an essential component of a startup’s raison d’etre; “to turn ideas into products,

29

measure how customers respond, and then learn whether to pivot or persevere”. While pivots will change the strategy that a startup implements in order to pursue a vision, it is rare for pivoting to change the actual vision.

The work of Bajwa et al. (2016), Bajwa et al. (2017), Ochoa-Zambrano and Garbajosa (2017), Bohn and Kundisch (2018), Hirvikoski (2014) and Terho et al. (2015) have established between them that at least fourteen different types of pivot exist. There is some minor divergence among these authors as a result of the work of Bajwa et al. (2017) which identified three novel pivot types and Hirvikoski (2014) whose presentation of a social pivot type has not been universally accepted by other authors. However, other than these differences, concurrence and concordance exists.

The results of this research suggest that a relatively wide range of pivot types were executed by digital startups within the three incubators studied; the most common being – in no particular order of popularity – complete, customer segment, customer need, business architecture, value capture, channel pivot, social pivot and technology pivot. The extant research presents an unnuanced view of pivots, suggesting an almost overly romantic and simplistic view of the need for pivoting. The contribution of this work offers a slightly more nuanced view. However, the most important aspect of the results was not necessarily the existence of these pivots but how the startups arrived at the decision that such a course of action was required. It would seem reasonable to assume that startups within an incubator setting would be guided towards such an essential developmental component by stakeholders within the incubators; however, that does not appear to have occurred in any meaningful way. Whether this omission is a simple oversight, is because the incubator coaches and managers do not see this as their role, or is due to the fact that they lack the expertise and/or experience to undertake this task is not clear.

Another aspect of pivoting that was considered during this research project is the environmental triggers that lead to it. Fourteen of these – eleven external and three internal – courtesy of the work of Bajwa et al. (2016) and Bajwa et al. (2017), were explained earlier. This research identified a diverse group of externally positioned triggers comprising technology challenges, negative customer feedback and running into legal issues. From an internal perspective, the identified triggers were a flawed business model and an unscalable business. As mentioned earlier, a number of recent pivots were directly attributable to the impact of the novel coronavirus. When reviewing the list of fourteen pivot triggers identified in the literature, it is difficult to find a trigger category within which this impediment belongs. It has undoubtedly triggered pivots but appears not to be classifiable within the current descriptors. However for the most part, the extant literature and the findings are in alignment on both the existence and categorisation of pivot triggers.

5.4 Failure/Disintegration This part of the thesis will discuss the empirical findings on failure/disintegration in the context of extant research related on this subject.

Ries (2011) explains that there will always be a proportion of startups that fail/disintegrate; a figure that is estimated by the work of Cusumano (2013) and Marmer

30

et al. (2012) to be at least 90%. It is difficult to conceive of any serious researcher who would debate this.

However, a topic that appears to be more likely to generate debate is whether failure should be seen as a net positive or net negative outcome. The work of Mitchell, Mitchell & Smith (2004) hypothesises that ‘Entrepreneurs who have failed will have higher levels of expertise relative to experience than those entrepreneurs who have not failed”. The work of Cassar and Craig (2009) tempers this enthusiasm for failure somewhat by demonstrating that there may be a tendency for entrepreneurs to paint a rosier picture of the circumstances and reasons resulting in their failure than is justified by the facts of the matter. It is possible that the viewpoint taken on this matter is highly dependent on the nature of the failure. What were the key learnings extracted from the endeavour? Were the reasons bad luck or poor judgement? Has the failure constrained the potential for future entrepreneurial efforts by the digital startup or its owners?

One of the important findings from this study relates to the lack of consensus regarding what failure actually is; it appears to mean different things to different people. While the interviews conducted for this thesis did not uncover any real evidence of failure/disintegration, it certainly uncovered a lack of consistency in the meaning ascribed to the terminology by the study participants. It also seems possible that were the extremely limited geographic scope of this study to be expanded to a different country or continent then there may be a far more significant dearth of conformity, both in the meaning of the terms as well as the stigma that is attached to it by both the digital startup’s senior management and society at large.

An important debate that exists in digital entrepreneurship is whether not opportunities are simply discovered by digital entrepreneurs or must be manufactured in order to avoid the spectre of failure/disintegration. The work of Shane (2012) supports the belief that while the opportunities themselves are objective and identified by those possessed of an entrepreneurial spirit, what is created and more subjective is the response in terms of how resources are recombined by entrepreneurial organisations to take advantage of these opportunities. However, the work of Alvarez and Barney (2007) takes a partially opposing view, suggesting that it is possible for opportunities to be the outcome of entrepreneurial endeavours. The high level of inertia exhibited by the digital startups contained within this study suggests that they may subscribe to the former view although this question was not specifically referred to during the interviews conducted.

In summary, failure/disintegration is clearly a factor of life for the vast majority of digital startup endeavours, digital or otherwise. However, it was not found to exist within the population that formed the foundation for this work. Terminology interpretation and cultural stigma reasons aside, this suggests there may be other factors at work, such as inertia. Whether inertia induces failure avoidance behaviour or vice versa is an interesting but, at this stage, unresolved question.

31

5.5 Implications for practice This section will discuss a number of potential operational and strategic implications arising from this study which may be of value for key stakeholders within both digital startups and incubators.

This work has already discussed the four strategies advanced by Nair and Blomquist (2018) to minimise organisational failure/disintegration; focusing on the team around a scalable idea and business model validation, development and scalability. Avoidance of failure is of course a desirable outcome. However, if digital startups within incubators are to evolve and reach their potential, the aspiration and focus must not be simply to elude failure, but also to pivot, scale and avoid inertia. This must be an area of focus – both for the incubators and their business coaches – that is seeded throughout their interactions with digital startups. The startups themselves must overcome their apparent fear of failure and recognise the fact that the extant data suggests they probably will fail/disintegrate. However it is this failure that will afford them the opportunity to try again with an enhanced skillset as well as deeper reservoirs of resilience, determination and experience.

While scaling is mentioned twice in the strategies referred to in the above paragraph, the results of this study suggest that considerably more emphasis needs to focused on it as an objective, in alignment with the thoughts of Ries (2011) and Picken (2017). The work of Picken (2017) specifically puts forth the opinion that the lifecycle of an entrepreneurial venture should consist of such distinct phases as startup, scaling and exit. It seems however that the vast majority of digital startups are not progressing beyond the first phase and exit – never mind scaling – remains but a lofty and distant objective. The spotlight of incubator endeavours must be robustly shone on this growth mechanism. An eminently practical route– and one that perhaps offers a quicker journey to the destination of organic growth – for digital startups looking to grow could be the previously discussed partnership advice of Henfridsson and Bygstad (2013), as exemplified by their “Scaling Mechanism”.

The implications for practice discussed in the previous two paragraphs should go some way to reducing what may be viewed as the scourge of digital startups and incubators; inertia. In addition however, it is incumbent upon incubators to ensure that the comfortable environment offered to their incubatees is not taken for granted and does not result in a scenario where stagnation becomes the dominant feature, while pivoting & scaling exist merely as secondary elements. The potential pitfalls associated with the state of inertia – that were explained earlier via the work of Ruef (2006) and Gong et al. (2009) – must be borne in mind by all parties. Aggressive short and long-term growth goals must be set, monitored, adhered to and achieved by incubators, coaches and startups in synergistic and coordinated cooperation. If these targets are not – at the very least – met then there must be questions asked, actions taken and consequences managed.

Incubator business coaches offer guidance on operational and tactical matters and the incubators themselves offer a wide range of services, opportunities and benefits. It is right that they do so, however it must not be at the expense of more strategically focused initiatives such as pivoting. Tremendous execution is meaningless if it relates to the wrong plan. It is therefore essential that pivoting is raised far higher up the agenda of both incubators, coaches and digital startups. Ideas must be tested, assessed and discarded and

32

this must be repeated until a feasible market offering is found. The famous words of Ries (2011) – “Build, Measure and Learn” – must become the mantra for all parties in all that they do.

5.6 Limitations By necessity, and as with all studies of an academic and/or scientific nature, a number of limitations exist within the design of this study. These will be acknowledged and discussed within this section.

The physical location within which this research was undertaken is a very narrow one. While the three incubators studied all exist as separate entities with different focal points, there are common threads running through them in terms of controlling stakeholders and ownership structure. It is reasonable to assume therefore that they have more features in common on a practical level than might be exhibited by equivalent organisations in other parts of Sweden or within a wider geographical locations. Any widening of the geographical context for similar research may well yield a different set of results.

While thirteen participants appears to be an adequate number of respondents for the purposes of this study, it cannot be argued that a larger and more diverse group of interviewees would likely yield data of greater richness and depth than it has been possible to generate during this exercise. In this scenario at least, bigger probably does equate to better.

6. Conclusions and further research The objective in this study has been to discuss the environmental components that both facilitate and hinder the advancement of digital entrepreneurship endeavours that take place within business incubators. This resulted in the formulation of the following research question: What are the enabling and inhibiting factors in digital startup evolution within an incubator setting?

In order to address this question, a multiple-case study was executed, that involved three incubators, located within the city of Umeå in Northern Sweden. Two of the incubators – UBI and EU - partner with organisations that operate respectively in the life science and creative industries. The third, BICF, has no particular industry focus. Thirteen semi structured interviews were undertaken with digital startups under the auspices of the incubators. The research efforts produced both some anticipated and unanticipated results.

One enabling factor in the former category was the finding that the digital startups claimed to both understand what a pivot is and to have undertaken the process of executing one or more of them. The range of pivot types undertaken is broad in character. An equivalent level of variety exists when one considers the triggers that lead to the requirement for pivots. The presence of incubators as important stakeholders in the evolution of digital entrepreneurship is also clearly a positive and enabling factor; although not one that is unable to benefit from adopting a more strategic view of their charges’ fortunes, ideally executed and facilitated by a team of coaches possessing a higher level of real world and relevant experience. Scaling should be both the ultimate enabling factor and

33

goal for any fledgling digital organisation, however other than in one case, this is not being achieved and therefore most of the organisations I interviewed are undoubtedly not achieving their full potential.

A number of inhibiting factors also appear to exist. The widespread inertia that was found to exist is both surprising and troubling, whether it be caused by a scarcity of resources, incubator indifference, flawed entrepreneurial attitude or an excessively comfortable existence. One may wish to consider as an example, the case of some PhD students employed by Swedish Universities who, in contrast to their colleagues at similar establishments in the U.K. or U.S., are relatively well paid and under no particular obligation to publish a specified number of papers. As a result, their corresponding research output is, generally speaking, far less prolific than students attending universities in a number of other countries. It is possible to take the view that failure/disintegration, by some measures, is a preferable outcome to one where organisations simply tread water and possess no aspirations to scale their business. Failure/disintegration at least reduces further wasted investments of energy, money and our most precious resource of all; time. While failure itself is certainly an inhibiting factor – and one viewed with considerable negativity in Sweden – it can also be seen as one with considerable mitigating factors, not least the considerable experience and learning that it affords those who encounter it.

Within these enabling and inhibiting factors there undeniably exists ample scope for further academic research endeavours. Obtaining an enhanced understanding of the behavioural and support characteristics that, if exhibited by incubators, may result in less inertia and failure/disintegration, would undoubtedly result in greater efficiencies potentially benefitting all the principal stakeholders. Further investigation of the same research question spanning a wider geographical setting – such as Scandinavia or the EU for instance – could also be an interesting and valuable project to undertake. This could facilitate the validation of both these research findings and the resulting IDM, as well as provide the opportunity to investigate the presence of any regional variations and the corresponding reasons underpinning them. Finally, a longer-term head-to-head study that compares and contrasts the fortunes of digital startups and incubators who follow the principles of the IDM with those who do not, may be an interesting undertaking to carry out.

34

Acknowledgements It is not possible for this work to be viewed as complete without the inclusion of a paragraph or two expressing my profound thanks to a number of individuals.

At the top of the gratitude list must come my thesis supervisor, Professor Jonny Holmström. Throughout the two years of my studies at Umeå University, Jonny has proved himself to be an invaluable and unsurpassable source of education, encouragement, guidance, amusement and thoughtful insight; always, of course, combined with an abstract thought or two regarding the prospects for his beloved Liverpool Football Club in the current or upcoming season. Without his large network, responsiveness, swiftness and clarity of thought, this work would not have been completed in its current form. Any errors, oversights or shortcomings in this work lie firmly in my sphere of influence and not his.

Thanks must also go to my principal contacts (and access facilitators) at the three incubators who participated in this study; Annakarin Nyberg of eXpression Umeå, Jennie Ekbeck of Umeå Biotech Incubator and last, but definitely not least, Lena Öhlund & Kerstin Edvardsson of BIC Factory. You all make an invaluable contribution to the hopes and dreams of those within your incubators and you should be proud of that.

Due to the guarantees of confidentiality granted to all the participants in this study I am unable to thank them individually. I must therefore content myself with a collective vote of appreciation. It takes considerable courage, energy, drive and determination to make the choices you have made and I wish you the very best of luck with your current and future endeavours. You deserve the very best of success and I hope that is what you receive.

A huge debt of thankfulness must also go to Ms. Johanna Katarina Rindforth of London and Småland whose emotional support, occasional stern lectures, praise and inspiration have played a massive part in the completion of this project. Thank you for everything Jo. I am a better person because of you.

Finally and most importantly, I must acknowledge the exceptional and unwavering support of my parents; Mrs Sandra Page and the late Professor John Graham Sayre Page, who unfortunately passed away on January 24th of this year. I am privileged to have had him as my father. His advice and encouragement to retrain was one of the principal reasons that I took the decision to relocate to Sweden and undertake this Master’s degree in IT Management and it is to my eternal regret that he did not live long enough to witness its completion. RIP Dad and thank you for everything.

Andrew Page Umeå September 9th, 2020

35

References

Alvarez, S.A. and Barney, J.B. (2007) Discovery and Creation: Alternative Theories of Entrepreneurial Action. Strategic Entrepreneurship Journal, 1(1-2), pp. 11-26

Bajwa, S.S., Wang, X., Duc, A.N. and Abrahamsson, P. (2016) How Do Software Startups Pivot? Empirical Results from a Multiple Case Study. In Maglyas A., Lamprecht AL (Eds.), Software Business. Lecture Notes in Business Information Processing. Cham: Springer, 240, pp. 169-176

Bajwa, S.S., Wang, X., Nguyen Duc, A. and Abrahamsson, P. (2017) “Failures” to be celebrated: an analysis of major pivots of software startups. Empirical Software Engineering, 22, pp. 2373-2408

Bandera, C. and Thomas, E. (2018) The Role of Innovation Ecosystems and Social Capital in Startup Survival. IEEE Transactions on Engineering Management, pp. 1-10

Bernard, H. R. (2017) Research methods in anthropology: Qualitative and quantitative approaches (6th ed.). Walnut Creek, CA: Alta Mira Press

Blank, S. (2013) Why the lean start-up changes everything. Harvard Business Review, 2013(05), pp. 1-9, Available at: https://hbr.org/2013/05/why-the-lean-start-up-changes-everything, Accessed August 20, 2020.

Blank, S. (2020) The four steps to the epiphany: Successful strategies for startups that win (3rd ed.). San Francisco: CafePress.com

Bohn, N. and Kundisch, D. (2018) Much more than ‘same solution using a different technology’: Antecedents and consequences of technology pivots in software startups. In Drews., Funk, B., Niemeyer P. and Xie, L. (Eds.), Exploring Innovation Practices for B2E Initiatives in the Digital Age, Lüneburg: Leuphana Universität, pp. 526-537

Bosma N. and Schutjens, V. (2011) Understanding regional variation in entrepreneurial activity and entrepreneurial attitude in Europe. The Annals of Regional Science, 47(3), pp. 711-742

Braun, V. and Clarke, V. (2006) Using thematic analysis in psychology. Qualitative Research in Psychology, 3, pp. 77-101

Braun, V. and Clarke, V. (2012) Thematic analysis. In Cooper, H. (Ed.), The Handbook of Research Methods in Psychology. Washington, D.C.: American Psychological Association

Brenner, T. and Fornahl, D. (2008) Regional path-dependence in start-up activity, (Papers in Evolutionary Economic Geography 08.12), Utrecht University, Utrecht.

Brynjolfssson, E. and Saunders, A. (2009) Wired for Innovation. Cambridge, MA: MIT Press

Cantamessa, M., Gatteschi, V., Perboli, G. and Rosano, M. (2018) Startups’ roads to failure. Sustainability, 10(7), pp. 2346-2364

Cassar, G. and Craig, J. (2009) An investigation of hindsight bias in nascent venture activity. Journal of Business Venturing, 24, pp. 149-164

36

Cohen, B. Amorós, J.E. and Lundy, L. (2017) The generative potential of emerging technology to support startups and new ecosystems. Business Horizons, 60 (6), pp. 741-884

Collier, D. and Mahoney, J. (1996) Insights and Pitfalls: Selection Bias in Qualitative Research. World Politics, 49 (10), pp. 56-91

Comberg, C., Seith, F., German, A. and Velamuri, V.K. (2014) Pivots in Startups: Factors Influencing Business Model Innovation in Startups. Conference on Innovation for Sustainable Economy & Society, Lappeenranta: Lappeenranta University of Technology Press, pp. 1-19

Crilly, N. (2017) “Fixation” and “the pivot”: Balancing persistence with flexibility in design and entrepreneurship. International Journal of Design Creativity and Innovation, 6, pp. 1-14

Crowne, M. (2002) Why Software Product Startups Fail and What to Do About It. Proceedings of the IEEE International Engineering Management Conference, Cambridge: UK

Cusumano, M.A. (2013) Evaluating a Startup Venture. Communications of the ACM, 56(10), pp. 26-29

Da Silva, F.R. Fabrício, R., Da Silva Pinto, R., Galegale, N.V. and Kabane, G. K. (2019) Why technology-based startups fail? An IT management approach. Paper presented at the Production and Operations Management Society, POMS 26th Annual Conference, Washington D.C., May 8-11

Davidson, E. and Vaast, E. (2010) Digital entrepreneurship and its sociomaterial enactment. 43rd Hawaii International Conference on System Sciences, Washington D.C.: IEEE Computer Society, pp. 1-10

Diefenbach, T. (2009) Are case studies more than sophisticated storytelling? Methodological problems of qualitative empirical research mainly based on semi-structured interviews. Quality and Quantity, 43(6), pp. 875

Drever, E. (2003) Using Semi-Structured Interviews in Small-scale Research: A Teacher’s Guide. Glasgow: Scottish Council for Research in Education

Dubé, L. and G. Paré. (2003) Rigor in information systems positivist case research: Current practices, trends, and recommendations, MIS Quarterly, 27(4), pp. 597-635

Eesley C.E. and Wu L. (2020) For startups, adaptability and mentor network diversity can be pivotal: Evidence from a randomized experiment on a MOOC platform. MIS Quarterly, 44(2), pp. 661-697

Eisenhardt, K. M., Graebner, M. E., & Sonenshein, S. (2016). Grand challenges and inductive methods: Rigor without rigor mortis. Academy of Management Journal, 59(4), pp. 1113-1123

Eisenmann, T., Ries, E., and Dillard, S. (2013) Hypothesis-driven entrepreneurship: The lean startup. Brighton, MA: Harvard Business School Publishing

37

Etikan, I., Musa, S.A. and Allkassim, R.S. (2015) Comparison of convenience and purposive sampling. American Journal of Theoretical and Applied Sciences, 5(1), pp.1-4

Fang, Y., Henfridsson, O. and Jarvenpaa, S. L. (2018) Editorial on Generating Business and Social Value from Digital Entrepreneurship and Innovation. Journal of Strategic Information Systems, 27(4), 275-277

Gartner, W. B. and Ingram, A. E. (2013) What do entrepreneurs talk about when they talk about failure? Frontiers of Entrepreneurship Research, 33(6), pp. 1-14

Giardino C., Wang X. and Abrahamsson P. (2014) Why early-stage software startups fail: a behavioral framework. ICSOB, pp. 27-41

Gong, Y., Baker, T., and Miner, A. S. (2009) Failures of Entrepreneurial Learning in Knowledge Based Startups. Frontiers of Entrepreneurship Research, 26(15), pp. 1-12

Guest, G., MacQueen, K. M. and Namey, E. E. (2012) Applied thematic analysis. Thousand Oaks, CA: Sage

Gustafsson, J. 2017, Single Case Studies vs. Multiple Case Studies: A Comparative Study. Retrieved from: http://www.diva-portal.org/smash/get/diva2:1064378/ FULLTEXT01.pdf. Accessed August 29, 2020.

Hannah, D. P. and Eisenhardt, K.M. (2018) How firms navigate cooperation and competition in nascent ecosystems. Strategic Management Journal, 39, pp. 3163-3192

Henfridsson, O. and Bygstad, B. (2013) The Generative Mechanisms of Digital Infrastructure Evolution, MIS Quarterly, 37(3), pp. 907-931

Henfridsson, O. and Yoo, Y. (2014) The Liminality of Trajectory Shifts in Institutional Entrepreneurship. Organization Science, 25(3), pp. 932-950

Hirvikoski, K. (2014) Startups pivoting towards value. Data and value-driven software engineering with deep customer insight. In Jürgen Münch (Ed.), Proceedings of the seminar no. 58314308, Finland: University of Helsinki, pp. 1-7

Holmström, J. and Sawyer, S. (2011) Requirements engineering blinders: Exploring information systems developers’ black-boxing of the emergent character of requirements. European Journal of Information Systems, 2081), pp. 34-47

Huang J., Henfridsson O., Liu M.J. and Newell S. (2017) Growing on steroids: Rapidly scaling the user base of digital ventures through digital innovation. MIS Quarterly 41(1), pp. 301-314

Islam, N., Buxmann, P. and Ding, D. (2017). Fostering Digital Innovation through Inter-Organizational Collaboration between Incumbent Firms and Start-Ups. Proceedings of the 25th European Conference on Information Systems, Portugal: Guimarães, pp. 1029-1043

Kaplan S. and Henderson R. (2005) Inertia and incentives: Bridging organizational economics and organizational theory. Organization Science, 16(5), pp. 509-521

38

Kirtley, J. and O'Mahony, S. (2020) What is a pivot? Explaining when and how entrepreneurial firms decide to make strategic change and pivot. Strategic Management Journal 2020(1), pp. 1-34

Leech, N. L., and Onwuegbuzie, A. J. (2008) Qualitative data analysis: A compendium of techniques for school psychology research and beyond. School Psychology Quarterly, 23, pp. 587-604.

Mankevich, V. and Holmström, J. (2016) Gateways to Digital Entrepreneurship: Investigating the Organizing Logics for Digital Startups. Academy of Management Proceedings, 2016(1)

Marmer, M., Herrmann, B. L., Dogrultan, E., Berman, R., Eesley, C. and Blank, S. (2012). The startup ecosystem report 2012. USA: Startup Genome, Available at: https://startupgenome.com/reports/global-startup-ecosystem-report-2012, Accessed August 29, 2020

Mehra, B. (March, 2002) Bias in Qualitative Research: Voices from an Online Classroom. The Qualitative Report, 7(1), Available at http://www.nova.edu/ ssss/QR/QR7-1/mehra.html, Accessed August 29,2020.

Miles, M. B. and Huberman, A. M. (1994) Qualitative data analysis: An expanded sourcebook (2nd ed.). Thousand Oaks, CA: Sage Publications

Mills, A. J., Durepos, G., & Wiebe, E. (2010) Encyclopedia of case study research. Thousand Oaks, CA: SAGE Publications, Inc

Mitchell, R., Mitchell, J. and Smith, J. (2004) Failing to succeed: New venture failure as a moderator of startup experience and startup expertise. In Zahra S. et al. (Eds), Frontiers of Entrepreneurship Research 2004, Massachusetts, Ma: Babson College

Myers, M. D. (2013) Qualitative research in business and management. London: Sage

Nair, S. and Blomquist, T. (2019) Failure prevention and management in business incubation: practices towards a scalable business model. Technology Analysis and Strategic Management, 31(3), pp. 266-278

Nambisan, S. (2017) Digital entrepreneurship: Towards a digital technology perspective of entrepreneurship. Entrepreneurship Theory and Practice. Entrepreneurship Theory and Practice, 41(6), pp. 1029-1055

Ochoa-Zambrano, J. and Garbajosa, J. (2017) An analysis of the Bluetooth Terminal development pivots from Lean Startup perspective: Experience and Lessons Learnt. In Tonelli R. (Ed), XP '17: Proceedings of the XP2017 Scientific Workshops, New York: Association for Computing Machinery, pp. 1-5

Picken, J.C. (2017) From start-up to scalable enterprise: laying the foundation, Business Horizons, 60(5), pp. 587-595

Ratinho, T., Amezcua, A., Honig, B. and Zeng, Z. (2020) Supporting entrepreneurs: A systematic review of literature and an agenda for research. Technological Forecasting & Social Change, 154, pp. 1-20

39

Ries, E. (2011) The Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses. New York: Crown Business

Ritchie, J., Lewis, J., McNaughton Nicholls, C. and Ormston, R. (2013) Qualitative Research Practice: A Guide for Social Science Students and Researchers. (2nd ed). Thousand Oaks, CA: Sage

Ruef, M. (2006) Boom and bust: The effect of entrepreneurial inertia on organizational populations. Advances in Strategic Management, 23, pp. 29-72

Sahay, S. and Walsham, G. (2006) Scaling of Health Information Systems in India: Challenges 14 and Approaches. Enhancing human resource development through information and communications technology, 12(3), pp. 185-200

Shane, S. (2012) Reflections on the 2010 AMR Decade Award: delivering on the promise of entrepreneurship as a field of research. Academy of Management Review, 37, pp. 10-20

Terho, H., Suonsyrjä, S., Karisalo, A. and Mikkonen, T.O. (2015) Ways to cross the rubicon: Pivoting in software startups. Lecture Notes in Computer Science, Cham: Springer, 9459, pp. 555-568

Tongco, M. D. (2007). Purposive Sampling as a Tool for Informant Selection. Ethnobotany Research and Applications, 5, pp. 1-12

Unterkalmsteiner, M., Abrahamsson, P., Wang, X., Nguyen-Duca, A. and Shahd, S. (2016) Software startups – a research agenda. E-Informatica Software Engineering Journal, 10(1), pp. 89-124

Yin, R. K., (2018) Case Study Research and Applications: Design and Methods. 6th ed. London: SAGE Publications Inc

Yoo, Y., Henfridsson, O. and Lyytinen, K. (2010) Research Commentary—The New Organizing Logic of Digital Innovation: An Agenda for Information Systems Research. Information Systems Research, 21(4), pp. 724-735

40

Appendix 1: Interview Guide Background of Company and Key Managers

1. Can you briefly describe the history of your company from its founding to the present?

2. Can you briefly describe the background of your senior management team?

3. Can you briefly explain your current strategy? 4. How long have you been part of this incubator?

5. Can you describe the major decisions that your organisation has taken during its lifespan?

6. What was the main obstacle you faced when your company started?

7. How have you overcome/attempted to overcome this?

8. What is your assessment of the firm’s current performance? 9. Have you personally and/or your organisation experienced rapid scaling at any point

in your lifecycle?

10. Has your startup experienced failure or have you previously experienced a failed startup?

Pivoting

1. Are you familiar with the definition of a pivot? 2. Can you provide a brief description of your pivot(s)? 3. At what point in your company journey did it/they take place? 4. Who participated in the decision-making activity? 5. What was/were the trigger(s) that led to your pivot? 6. Can you describe the mechanism/process by which the triggers led to the pivot? 7. How did you evaluate pivot alternatives? 8. What data/data sources did you utilise in selecting the option you proceeded with? 9. What were the pivot outcomes? 10. Did you validate the outcomes of your pivot? If so, how? What were the results?

Incubator Influence

1. Can you describe the feedback/guidance you have received from incubator business coaches and managers?

2. What specific areas of your business operations has it related to? 3. Has any of it related to pivoting?