Department of informatics Master thesis, 30 hp
IT-Management
SPM 2020.07
Making contact
A case study of a digital platform in
the construction industry
1
Abstract
Digital platforms are increasingly disrupting established status quos in traditionally stable markets. This study focuses on the launch and use of a digital service platform in the construction industry in Finland. The study encompasses a practical, action research case study of the platform itself and the context in which it exists. In order to clearly describe observed phenomena from the case, we utilise well-established concepts both from the field of economics and information systems to utilise a relevant vocabulary for describing the findings. The purpose is to answer our research question: what are the perceived market frictions for Small and Medium-sized Entreprises within the construction industry affecting networking and contracting? Also, in what ways can digital platforms mitigate these frictions? The study identifies several underlying frictions in the industry relating to network saturation, disintermediation, transaction costs and relate these to the functionality of the platform attempting to alleviate these frictions.
Keywords: Digital platforms, construction, market frictions, matchmaking
Acknowledgements
2
1. Introduction
Digital service platforms are becoming an increasingly pervasive phenomenon. These digital platforms connect buyers and sellers in the area of commoditised work, i.e. buying or selling a service conducted by an individual or by smaller firms on a digital marketplace (Aldrich, 1999; Lusch & Nambisan, 2015). The commoditisation of software-based services, e.g. programming and graphical design, have seen an adoption-rate higher than most industries (Asadullah, Faik & Kankanhalli 2018). In more traditional service industries, e.g. construction and infrastructure, adoption and deployment of such platforms have been slow and tedious (Argawal, Chandrasekaran & Sridhar 2016; Alaloul, Liew, Zawawi, Noor & Mohammed 2018). The main hindering factors for the adoption and development of digital platforms within these industries are the high level of manual labour involved in execution, scepticism and resistance entrenched in the mindset and slow adoption of digital tools (Argawal et al., 2016). An important characteristic of how the construction industry operates is contracting and sub-contracting; these are by no means unique for the industry as contracting is an important factor in most industries. However, these characteristics inherit the traditional legacy entrenched in the industry; many contracts are arranged through informal networks and word-of-mouth recommendations (Alaloul, et al, 2018).
Therefore, getting access to networks for collaboration through contracting or sub-contracting is a tedious and time-consuming bottleneck for companies in the industry who need either short-term strengthening of the workforce or for companies trying to get larger contracts by teaming up (Leviäkangas, Mok Paik & Moon 2017). Some inherent characteristics of digital technology have allowed for some digital platforms to emerge and alleviate many of the issues with short-term contracting, planning, collaboration and networking still plaguing the construction industry (Parker, Van Alstyne & Choudary 2017; Evans & Schmalensee, 2016). So, the industry as a whole seems ripe for disruption, indicating that there is potential for a digital platform to enhance and broaden these aforementioned aspects (Agarwal et al, 2016; Evans & Schmalensee, 2016)
Furthermore, research suggests the construction industry is suffering from its time-lag in digitalisation as it has been increasingly losing efficiency, with many projects going over budget and time
(Agarwal, et al, 2016)
. However, digital platforms are encroaching onthese traditionally analogue industries and will, most likely, disrupt well-established status quos and re-define industrial borders (Oesterreich & Teuteberg, 2016; Argawal et al, 2016). There are interesting findings from previous research on the subject that could apply to these industries (Van Alstyne, Parker & Choudary 2016; Evans & Schmalensee, 2016; Parker et al, 2017). However, there exists a gap in the research for studies aimed at understanding the implementation of digital service platforms in these lagging industries, specifically research trying to understand how these digital platforms can support collaboration, networking and alleviate market frictions (Alaloul et al, 2018; Oesterreich & Teuteberg, 2016).
3
involved in making an economic transaction, which could involve, but is not limited to, time and inconvenience (Williamson, 1981). Barriers, i.e. obstacles hindering contestants from entering a market (Evans & Schmalansee, 2016; Ivarsson & Svan, 2020). Alleviating these frictions has been a primary focus of many digital platforms and their considerable growth over the last decade highlights that there is a willingness to pay for the alleviation of market frictions (Evans & Schmalensee, 2016). Many of the early success stories have revolved around the B2C market, wherein we find such actors as Spotify, Netflix and Uber (Kenney & Zysman, 2016). Seeing the successes of these B2C platforms has drawn much attention from researchers, and it has been the object of study in many reports and articles the last years (Tiwana, Konsinsky & Bush 2010; Hagiu, 2014; Kenney & Zysman, 2016; Evans & Schmalensee, 2016). The continuous growth and expansion of these digital platforms show that they are in demand and that the markets are answering positively to their introduction (Taeuscher & Laudien, 2017; Lusch & Nambisan, 2015).
Therefore, we believe the timing to be right in studying a more traditional sector which is just starting its digitalisation efforts. As the industry has been lagging, and there is some entrenched resistance within, we argue that research specific to the industry is needed to guide the digitalisation efforts. Specifically researching how digital tools and possibilities can be applied in order to help and enhance how construction firms can collaborate, network and contract for workforce reinforcement are vital as digital platforms are becoming increasingly pervasive (Evans & Schmalensee, 2016). Understanding the factors for succeeding in implementation and how to guide the design will be essential to penetrate a sceptic market like the construction industry, we would argue that our research is an important step in furthering the understanding in this field of research.
Therefore, in order to understand our chosen field of research, we have conducted a qualitative case study on a digital platform in a region of Finland, specifically targeting the construction industry. The research questions we will attempt to answer in the study are as follows:
• What are the perceived market frictions for Small and Medium-sized Entreprises (SME’s) within the construction industry affecting networking and contracting? • In what ways can digital platforms mitigate these frictions?
In order to answer our research questions, we interviewed both the platform owner and the SME’s who are using the platform today. The results will be presented in the result section (part 4) and discussed in order to generate clear results from the data gathering.
2. Theoretical foundation
4
through its platform (Williamson, 1981). We, therefore, want to investigate to what degree these frictions are present within the research context. The concept of matchmaking involves the practical implementation of an intermediary platform in a market (Evans & Schmalensee, 2016). It is relevant to study how the design of the platform performs and how the actors involved view its value proposal. Both of these research areas contain concepts and theories that we consider relevant in order to analyse the underlying conditions of the research context of this study, as well as providing a more defined vocabulary for us to convey our findings.
2.1 Construction industry digitalisation
There is an extensive body of research emerging regarding digital platforms (Asadullah, et al, 2018) as they have grown in popularity and continuously transform industries today (Gawer & Evans, 2016). Many papers have investigated the dynamics between smaller business entities and digital platforms in retail by, for example, studying the ecosystems formed by large companies such as eBay, Alibaba or Taobao (Chu & Manchanda, 2016; Tadelis, 2016). Other popular areas of interest have been the trust aspects in the transportation industry or accommodation where companies such as Uber (Mittendorf, 2017) and Airbnb (Zervas, Proserpio & Byers, 2015) are common subjects of study (Sutherland & Hossein Jarrahi, 2018). The industry itself has a substantially lower investment percentage in R&D compared to other industries of the same scale, on average 1% compared to over 4% in aeronautics and 3,5% in the car industry (Agarwal, et al, 2016). A considerable part of the influential research regarding platforms in the construction industry has been done on the concept of Building Information Modeling (BIM) (Oesterreich & Teuteberg, 2016; Singh, Gu & Wang 2011), as well as project and knowledge management (Styhre & Gluch, 2010).
5
2.2 Digital matchmaking platforms
The concept of matchmaking fundamentally deals with the process of connecting and facilitating interactions between actors to amplify and create new forms of value (Ivarsson & Svahn, 2020; Parker, et al, 2017). Matchmakers, i.e. firms conducting this process do this by providing a shared space for actors to meet and exchange services or products, effectively splitting the market into at least two different segments of customers who affiliate with the matchmaker, but essentially only interacts with other active actors in the shared space provided by the matchmaker (Evans & Schmalensee, 2016). For this reason, the shared spaces provided by matchmakers are often referred to as multi-sided markets within the field of economics (Evans, 2003). This implication means that the value-logic of matchmaking firms is non-linear in contrast to firms conducting its business utilising more traditional pipeline supply chains. Where traditional supply chains increase the value offering through efficiency improvement in the activities leading up to the point of transaction (Coyle, Langley, Gibson, Novack & Bardi, 2009), multi-sided markets harness network externalities as a defining component (Holzmann, 2014), increasing its value-offering by expanding in both scale as well as scope by attracting more actors to its shared spaces (Evans & Schmalensee, 2016).
6
for effective platform strategy: initialising and growing the created shared space and
orchestrating the ecosystem of actors emerging as the shared space grows. In order for
platform owners to increase their value offering it is argued that they need to consider how to attract at least two distinct sides of a market that would benefit from interacting on the platform (Rochet & Tirole, 2003; Evans, 2009), and what entry frictions that are intrinsic in the current design of it (Parker, et al, 2017).
2.2.1 Initialising digital platforms & entry frictions
Initialising a platform is often referred to as the “chicken and egg problem”, which is the challenge of a platform owner to attract two multiple sides of a market in the initial phase where none of the sides are currently present and active on the shared space (Parker, et al, 2017). As actors on different sides of the market are reluctant to join a platform without the other side also being present (Hagiu, 2014), a continuous topic of discourse is how to formulate strategies in order to solve this problem of launching a platform where multiple patterns practised by different firms have been identified (Stummer, Kundisch & Decker 2018). Relevant concepts for this study, related to these strategies are subsidised pricing
structures, single target-groups, and user-specific platform design.
Subsidized pricing structures refers to the practice of one type of actor being offered the services at a reduced price or for free (Hagiu, 2014; Parker, et al, 2017). This is, for example, the prevailing price model from the magazine industry where magazines subsidise the cover side as a way to increase readership to incentivise advertisers to join their shared space (Kaiser & Wright, 2004). Single target group is a practice where the platform owner concentrates its initial focus on one customer segment or geographical area (Stummer, et al, 2018). Uber building its initial critical mass within the city of San Francisco is an example of how companies utilise this practice. User-specific platform design refers to the practice of providing tools that serve one customer segment to such a degree that it will join the platform, even though there are no representatives of the other side of the market. As the growth of one actor on the platform increases for the reason of adopting the tools embedded in the platform, other actors will be attracted due to the presence of the first actor-group (Parker, et al, 2017).
Even though a platform has been initialised through strategy that utilises some, or many of the concepts outlined above, an influential facto§ r that can limit further growth are entry frictions. Entry frictions in the context of platform discourse are the barriers that hinder or slow down the process of an actor joining and contributing in the value-creating activities facilitated by the platform (Parker, et al, 2017). Several concepts are related to entry frictions, but for this study we will focus on the monetary price of joining or
performing activities on a platform (Hagiu, 2014), as well as how the platform is designed technically and its implications on an organisation’s current work-practices (Parker, et
7
considered a governing structure which exerts control over an organisation’s work practices. In the platform discourse, control relates to platform design and what type of mechanisms in place that a platform owner utilises as a way of governing the processes and interactions that occur in the shared space provided by the owner (Schreieck, Wiesche & Krcmar, 2016).
2.2.2 Network orchestration & the risk of disintermediation
The concept of orchestration within the context of digital platforms or digitalized ecosystems refers to the facilitation of the processes that stimulate value-generating activity in the shared space (Dhanaraj & Parkhe, 2006). The role of orchestrating is often reserved to the platform owner as they provide the shared space under most circumstances. Orchestrating a platform is based on the formulation of rules and architecture (Van Alstyne, et al, 2016), which dictates to what degree the platform could be considered open or closed in terms of architecture and governance (Smedlund & Faghankhani, 2015). Platforms exhibiting an open architecture provides third-party actors with tools to create new sources of value on the platform, whereas in a closed architecture further technical development is reserved for the platform owner. Open governance refers to a higher degree of rules regarding trade and the sharing of rewards being shaped by the actors on the platform (Van Alstyne, et al, 2016).8
stemming from the lesser core interactions taking place. Platform owners address this aspect by enforcing stricter rules and more closed governance.
Since platform owners and matchmakers act as an intermediary between different sides of a market, a central concept of network orchestration is the inherent risk of disintermediation. Disintermediation refers to conditions and events that weaken the position of an intermediary (Gu & Zhu, 2018; Zhu & Iansiti, 2019), which can occur from both external factors such as disruptive elements in a market (new competitors or innovations) as well as from various ways of interaction on the space provided by an intermediary. An example of the latter could be the choice of two actors to leave a shared space after a match has been facilitated in order to circumvent added costs associated with the participation in the space. One such cost could be commission fees on subsequent activities that the matchmaker also intends to facilitate for the actors. In that scenario, the matchmaker risks failing to capture value from the interacting actors, while the actors capture value from the initial facilitated match between two actors, but decide to continue their business outside the realm of the intermediator.
2.3 Market frictions as the foundation of matchmaking
Matchmaking, whether it is in the form of a mall or a digital platform, centres on the concept of minimising frictions and the role of intermediators in a market. Early 20th
century economists theorised why intermediation is so pervasive in market economies. These economists found that intermediators are efficient for alleviating transaction costs between parties and to lower barriers for entry present in any market setting. Therefore, in order to fully discern why the term matchmaking is so prevalent today, we need to understand the foundational concepts underpinning matchmaking.
2.3.1 Transaction costs
The idea of inherent frictions that are present in any market setting started to arise in academia at the turn of the 20th century. Two Nobel prize-winning articles were
ground-breaking in realising and defining what constitutes these frictions and how they can be analysed. The first article of these that dealt with the analysis of market frictions through a transaction cost analysis was written by Ronald Coase (1937). Coase (1937) main point was an attempt to understand why individuals organise in firms and through his analysis one conclusion he made was that the formation of firms was an attempt to minimise transaction costs in the market. Transaction costs are defined as any cost involved in making an economic transaction (Coase, 1937; Williamson, 1981). Mainly, transaction costs that are of relevance to this thesis are search, inconvenience, trust and information. This highlights an essential fact of transaction costs, they are not limited to strictly financial transaction, but also relate to costs associated with searching, matching and the procurement and allocation of scarce resources (Coase, 1937; Williamson, 1981; von Hayek, 1945).
9
Transaction cost
Definition
Example
Search costs Costs associated with searching for resources to complete a transaction
Resources spent calling potential collaborators until a transaction can be done
Information costs Costs for gaining information sufficient to complete a transaction
Resources spent in gaining satisfactory knowledge about a potential collaborator
Inconvenience costs Costs inconvenience in completing a for the perceived
transaction
Total resources spent on unsatisfactory activities leading up to a transaction
Trust costs Costs associated with gaining a satisfactory level of trust between transacting parties
Resources spent in order to establish a satisfactory level of trust between parties engaged in a transaction
Table 1, transaction costs definitions
Williamson (1981) synthesise their articles and others in order to arrive at a more uniform definition of transaction costs, arriving at how the concept is understood today, which is as was defined earlier; any cost involved in making an economic transaction (Coase, 1937; Williamson, 1981). However, in order to make the definition clearer and more applicable to an organisational transaction, he adds a more granular question to the definition, i.e. do the parties to the exchange operate harmoniously (Williamson, 1981, p. 552)? What this means is, what are the costs associated with the exchange at hand? Are there information asymmetries, i.e. one party having more in-depth knowledge about the transaction, or is there much planning, i.e. time-consuming tasks needed for completion of the transaction (Williamson, 1981; Foster, Graham & Waela Mwolo 2019). The level of transaction costs present in a market-setting has consequences on the organisation of actors (Coase, 1937) (Williamson, 1981). The very emergence of organising in firms can, according to Coase (1937), be seen as a consequence of trying to lower transaction costs. In some modern market settings, e.g. digital platforms, there has been a move from firm-organisation to market-organisation wherein individuals avoid organising in firms and ultimately act directly in the market because transaction costs are low enough to avoid firm organisation (Williamson, 1981).
The main point in utilising the concept of transaction costs is that they are clearly defined and well-established in the field, we draw upon this body of knowledge in order to utilise clearly defined concepts. We will not apply Transaction costs analysis (TCA) as it is out of the scope of this thesis; we instead apply the concepts laid out to understand and describe different phenomena observed in the study. Furthermore, there exist some quite substantial criticisms of Coase’s (1937) theorem as a whole, both in the application and in theory. However, there is seemingly little disagreement in the field regarding the existence of transaction costs (Bylund, 2014). Therefore, we argue that there is enough credibility for the application of the concept of transaction costs in this thesis when analysing our intended market setting.
2.3.2 Barriers for entry
10
eliminate new entrants from entering a market (Caves & Porter, 1977). In general, there are seven main definitions of what constitutes a barrier for entry. Each has built on earlier definitions, either based on criticism or a need for clarification (McAfee, Mialon & Williams 2004).
However, there are some general traits which run like a red thread through the definitions. These are the presence of incumbents, costs not borne by incumbents or start-up costs, capital requirements and time (McAfee, et al, 2004). The presence of incumbents is nearly always a fact when entering into a market. Unless the actor is first in a newly emerged market, there will most likely be one or more incumbents acting in the market (Gilbert, 1989). Incumbents usually have an advantage over new entrants due to them already overcoming barriers and having cemented their position in the market (von Weizsacker, 1980; Bain, 1956). The presence of incumbents allows for the second barrier, which is costs not borne by said incumbents, this is not costs that the incumbent had to pay in order to enter, but costs which have emerged afterwards, e.g. certifications and patents, often lobbied for by the incumbents (Fisher, 1979; Gilbert, 1989; von Weizsacker, 1980). In many markets, certain capital requirements act as barriers, as this study aims to focus on the construction industry; it is one such market that has heavy capital requirements. These requirements can come in the shape of heavy-duty machinery, without having enough capital goods, a new entrant cannot enter such a market (Bain, 1956; Carlton & Perloff, 1994; Stigler, 1968). Lastly, time can be a barrier for entry as a new entrant can enter too early or too late timing-wise or that it can take a long time to enter a market (Carlton & Perloff, 1994).
These four barriers mentioned above are present in each definition of what constitutes a barrier for entry. How these barriers affect new entrants differ from case to case and depends a lot on which industry the new entrant is targeting. It would seem that entering into a market for pharmaceuticals differs a lot from creating a digital e-commerce store with regards to how high the barriers for entry are (Gasser & Palfrey, 2007). The emergence of ICT and the pervasiveness of digitalisation throughout many industries and markets worldwide has seen a lowering or removal of these barriers (Gasser & Palfrey, 2007). Research suggests that because of some inherent characteristics of ICT, e.g. interoperability, modularity, self-reference and reprogrammability, some barriers hindering market entry can and has been lowered (Yoo et al, 2010; Gasser & Palfrey, 2007).
The identified barriers are all present in the construction industry. We regard this as an exciting area for research and would like to add to the body of research through this study. As there is no one definition for barriers for entry to go by, we would utilise the identified and established barriers as a concept for analysis and build our discussion upon it.
3 Methodology
11
different stages of the research. Lastly, we will clearly state the ethics guiding our research and argue for our choices in relation to observed criticisms of such an approach.
3.1 Research context
For our study, we have chosen to focus on the Nordic region mostly based on a choice of comfort as it is the region we are living in. The Nordic region is generally acknowledged as an innovative region that prides itself in taking the lead in many digitalisation efforts, wherein the construction industry is no exception (Randall & Berlina, 2019; Navrestad, 2019). Most large construction companies in the region are currently developing and trying out many novel solutions and tools that aim to further the digitalisation of the industry (Navrestad, 2019). They are following the global trends of moving towards paperless construction sites, cyber-physical spaces and more (Alaloul, et al, 2018). However, the region has come further than many others, thanks to the innovative and digitally mature nature of the region (Randall & Berlina, 2019). The region's market is dominated by a few large actors who are spearheading the digitalisation efforts and who, through their sheer size, acquire the government contracts (Siemens, 2018). However, even though the regional construction incumbents are making efforts and investing in digitalisation, the industry is lagging in the region, as has been observed globally (Siemens, 2018).
special-12
competencies. The long-term goal is to disrupt the incumbents and draw them into the marketplace when enough users are active, and MarkPro is the go-to place for contracting.
Most research to date has focused on the larger incumbents in the industry and the digitisation of standards and formats (Agarwal, et al, 2016; Alaloul, et al, 2018; Krämer & Besenyői, 2018). As this study focuses on SME’s and the collaborative aspects of quick contracting and the easing of networking. We argue that our study is motivated and relevant for the current state of the industry and can hopefully add to the body of knowledge within the industry itself in how SME’s can leverage the opportunities afforded by digital platforms.
3.2 Case study approach
We choose to utilise a qualitative case study since the phenomena of digital platforms is relatively new, especially within the business domain, making it a good fit for the method (Yin, 2018). To succeed in formulating a rigorous study, we turned to the definition outlined by Robert Yin (2018), which is an empirical method that “investigates a contemporary phenomenon (the “case”) in depth and within its real-world context, especially when the boundaries between phenomenon and context may not be clearly evident” (Yin, 2018, p. 45). We deemed that our study matched the definition of a case study and then looked deeper into the guidelines outlined by Yin (2018) for what constitutes a good case study and how to structure it. Furthermore, settling on using the definition by Robert Yin (2018) is due to his credibility in the field of social science case studies and the pervasive use of his theories in our field.
Our research questions are defined as what and how questions, which goes in line with good questions as a basis for a case study (Yin, 2018). Even though a what question can be argued to fit other forms of study, our is exploratory due to the novel setting and therefore fits a case study
(Yin, 2018)
. Secondly, our study does not require control over behaviouralevents; on the contrary, we relinquish any claims to control behaviour. This is a good foundational argument for a case study as it allows events occurring within the confines of the case to occur naturally and spontaneous (Yin, 2018). Third, we focus on contemporary events and are not entirely devoted to historical events, which is also a good argument for conducting a case study (Yin, 2018). However, our case was afforded to us through a local EU-funded project. This meant that the founder of the start-up that is our case had some demands as to the framing of the case. This is increasingly the case for social science studies as we often emerge ourselves in the organisations studied and therefore are often given requirements by the organisation at hand
(Van de Ven, 2007)
. The realities of such an13
Mathiassen, et al (2012). An inductive style allows inferences to be drawn from the findings present in the context while minimising preconceptions, which becomes more critical in situations where researchers have to emerge themselves in the case setting. This means that our inference will be grounded in findings from our problem-solving cycle and embedded in the theoretical concepts from our literature review (Mathiassen, et al, 2012). Through our interviews and subsequent data analysis, we will draw out themes which we can ground in the research we have done prior. This was also a result of the founders' requirement, as we built upon his perception of the problem, we had to let the data speak to us and then relate it to the research we read. Lastly, our intended contributions can be both be characterised as experience report, i.e. arguments from insights for IS practice (Mathiassen, et al, 2012, p. 6) or a field study, i.e. real-world insights for primary contribution to the area of concern (Mathiassen, et al, 2012, p. 6).
3.3 Data collection
Primary data for this case was collected from a sample of 12 actors who participate in what we refer to as the regional construction business community. These are categorized as either being active participants or supportive participants in order to distinguish between actors whose structure is being investigated. We define active participants as SME’s offering products and services in various segments of the construction industry. Supportive participants are actors who in various ways, interact with the active participants but do not directly offer construction-related services. Examples of active participants are a painting firm, an engineering firm, a carpentry firm. Examples of supportive participants are a digital platform provider, a marketing firm, a public business support hub, see table 1. In total, 13 semi-structured interviews with 12 different actors were conducted in person at various places during two days in February 2020, with the average length of each interview being 50 minutes. The interview guides (see appendix 1) were structured in a manner to gauge the general conditions in the problem areas identified by MarkPro; this structure is intended to provide as general answers as possible relating to the issues MarkPro is attempting to alleviate.
Actor
Position
Sector
Participant type
Founder CEO IT Owner
Firm #2 CEO Carpentry Active
Firm #3 Project leader Painting Active
Firm #4 CEO Excavation Active
Firm #5 Project leader Carpentry Active
Firm #6 Consultant Business support Supportive
Firm #7 Advisor Business support Supportive
Firm #8 CEO Demolition Active
Firm #9 CEO Marketing Supportive
Firm #10 CEO Sales Supportive
Firm #11 CEO Engineering Active
Firm #12 Supervisor Aggregate Active
14
3.4 Data analysis
Our approach to analysing the collected data is mainly based in the research philosophy of interpretivism and its epistemological belief that in order to generate knowledge of social processes a researcher needs to get inside the world of the actors generating it (Orlikowski & Baroudi, 1990). This is consistent with the overarching approach of this study, where we mainly study the phenomena in its natural setting and interpret the data from a variety of perspectives. This also contrasts to a more positivistic approach which utilises hypothesis deduction and fixed constructs as a mean to understand the reality of social processes (Orlikowski & Baroudi, 1990).
The main method for analysing our data is through thematic analysis. Thematic analysis has been argued to be a valuable method for capturing complexities of meaning within textual data sets (Guest, MacQueen & Namey 2011), as well as being useful when discovering knowledge through interpretation (Alhojailan, 2012). We argue that it provides us with a systematic framework to analyse the frequency of emerging areas of concern and how they relate to the different actors active on the platform. It will also aid us in contextualising the results within the broader research domain, as the emerging themes facilitate comparison between our findings and related research.
Our philosophical approach to thematic analysis could broadly be categorised as experientially oriented. It is based on the assumption that language reflects reality and focuses on what participants think, feel and do in regards to the phenomena being investigated (Willig & Stainton Rogers, 2017). The practical approach is derived from the six-phase outlined by Braun & Clarke (2012), which asserts that the first step, shared with other forms of qualitative analysis, is to become familiar with the data. Familiarisation, in this sense, not only consists of listening to the recorded audio and transcribing it but highlighting observations and areas of interest as this process is being undertaken (ibid). During the first phase of our data analysis we conducted daily meetings where we, during and after the interview had been transcribed, discussed the content and made collective notes from our meetings in our shared project space. Our notes at this stage were mainly general observations of what type of problems that were addressed by the interviewee’s multiple times. The second phase, consisting of generating codes as a form of labelling the data and serving as the basis for further analysis, was performed after all of the interviews had been transcribed. The codes produced in this phase were the result of an inductive approach and mostly of descriptive character to stay close to the content of the data (ibid). Examples of the codes produced are: current or past usage of technology, networking, communication channels, industry-specific problems. As the process continued, more codes emerged from the data which were more specific to our research context, examples being: barriers (specific problems in contrast to general problems), region size, trust issues and seasonal fluctuations (of available work).
15
problematic aspects, specific problematic aspects related to contracting, and perspectives on the launch of MarkPro. An example of how this process was performed is the formulation of the themes related to the topic of “the current state of digitalisation”, which among others, were formed through synthesizing the output from the codes of “usage of technology” and “communication channels”. From the data detailed by the participants, we determined how they had adopted digital technology and which processes that were most supported it. This was then thematised as “utilization of smart devices increases”, “administrative tools are highly prioritized”, and “the choice of communication technology correlates to the private technology adoption”. Through themes such as these, we generated a guideline for how to present the data in a way that emphasize the re-occurring patterns and tell a coherent story; which is consistent with the final steps of the approach prescribed by Braun & Clarke (2012).
Based on these themes, we turned to the literature in order to crystallise and clarify what our theoretical foundation needs to contain and what theoretical constructs that are valid for our analysis. Since we already had a general idea of what theoretical constructs could help in our research, the job of mapping the results from our thematic analysis to our theoretical foundation was more about details than building the foundation from scratch. In practice, we took the results from our thematic analysis and applied the theoretical constructs from our foundations to ground our findings on the constructs deemed relevant for this thesis. This could, for example, be taking a theme like “the current state of digitalisation” and analysing the codes and findings underlying that theme. Then through the analysis, find what correlates to our theoretical constructs, for example, analysing how the actors' current state of digitalisation is attempting to minimise transaction costs already. This is a substantial part of the data analysis process, to take what actors said during our interviews and understand what they are saying through the application of relevant research and to abstract their descriptions to match the constructs.
3.5 Research Ethics
Ethical principals are a cornerstone for good and moral research, and in this thesis, we have taken into account and applied established principles for conducting an ethically correct study. For this thesis, we have decided to apply the principles for ethical social science research defined by Israel & Hay (2006). In broad terms these can be divided into four principles all aimed at minimising the risk of harm to the subject at hand, these four are obtaining informed consent, protecting anonymity and confidentiality, avoiding deceptive practices and providing the right to withdraw (Israel & Hay, 2006).
16
make sure that no personal connections to the individual could be derived from the transcripts. Lastly, before the interviews began we made sure that all our participants understood that they could abort the interview at any stage and if they so wished they could have the recordings deleted, which also included the transcripts should they call for that after the interviews were done.
During our data analysis and in writing this thesis, we have tried to maintain strict anonymity for our participants and have strived to keep our descriptions as minimalistic as possible in order to safeguard their identity.
3.6 Methodological discussion
Case studies as a research method have been a controversial subject in the scientific community for some time due to the difficulties of generalising the contributions and the context-dependent knowledge generated (Flyvbjerg, 2006). Campbell & Stanley (1966) describe the flaws of case study research as “[..] such studies have such a total absence of control as to be of almost no scientific value” (Campbell & Stanley, 1966, p. 6). This highlights a disconnect between the natural science way of researching with hard facts, repeatable and generalisable findings to the more soft and subjective social sciences (Campbell, 1975). Still, case studies have cemented their place in the scientific community thanks to its inherent subjectivity which allows it to find human nuances and highlight the contextual conditions that often have substantial impacts on the outcome (Flyvbjerg, 2006). It is harder to draw truly generalisable conclusions from studying context-dependent cases as the findings are intrinsically linked to the context in which they exist (Flyvbjerg, 2006). However, this is also where the strength of a case study lies, by contextualising findings it is often easier to convey the knowledge to the reader rather than the stating of hard facts found in the more “traditional” scientific methods (Flyvbjerg, 2006).
17
4 Findings
The findings presented in this section will be presented based on the themes from our data analysis. In order to highlight different perspectives between the platform owner and platform users, we begin by presenting the results stemming from our initial interviews with the platform owner (4.1), focusing on market research and problem analysis made by Markpro. This is followed by results from our interviews with SME’s active in the business community. Next, section (4.2) explores the SME’s past and present adoption of digital tools. We do so as a way of analysing which business areas the participants have identified as more valuable to digitalise and adopt in their organisations. We continue with past experiences of stimulating networking (4.3), that details earlier attempts within the industry of addressing some of the identified problem areas. After that, we highlight the respondents’ general views on what they consider to be areas of frictions (4.4). Furthering their perspective, we present the areas expressed being most problematic (4.5). We finish the findings section by presenting the outcome from the launch of the digital platform, as viewed by both MarkPro and its users (4.6).
4.1 An idea is born
The idea for the platform was born through the founder discussing conditions in the market for SME’s with his wife. At her workplace, she had observed how seasonal fluctuations affect the demand for short-term contracting within the region. Her workplace had a demand which could fluctuate rapidly, affecting suppliers who fundamentally depend on their orders. Therefore, suppliers strived to have just enough employees and not to risk recruiting and employing people during the spikes in demand. Based on her observation, the founder reasoned:
“So they rent labour from staffing agencies, especially if its for the factory floor. But they can’t cover everything. So we thought that this could be an alternative to the recruitments and layoffs. Instead of renting labour, buy the service from hungry entrepreneurs in the region.” - MarkPro Founder
MarkPro’s founder is an established actor in the local business community since around ten years back and has a good reputation with many local firms, partly due to being a successful entrepreneur in another venture. Therefore, he could informally discuss and verify the observation made by his wife with several of them. Through these discussions, he could identify three problematic areas which he believed could be helped by the introduction of a digital platform for services procurement and sales.
“We solve the firm’s problems with… there are three categories really, but say that they need an extra resource, one category is a temporary need that could stem from a peak in demand […] The second is if they need a special competence, say like an engineer or excavator […] The third is if they need help for an undetermined period of time, if they can’t or don’t want to employ at the moment” - MarkPro
Founder
18
verified his identified problem-areas and were instantly excited, offering him the support. The founder then set up an advisory board for MarkPro consisting of representatives from these business support functions. The advisory board’s primary task was to support the founder in strategising and conceptualising user experience and business model design. Through his previous entrepreneurial ventures, he had managed to accumulate some capital. Thanks to this capital, he could fund the initial development, and start MarkPro relatively quickly with his funds.
The founder perceived that the SME’s were suffering from insufficient networks and inefficient matching mechanisms. He deemed that the MarkPro platform would alleviate these frictions. He wanted MarkPro to become the central network for regional SME’s to find extra resources, a place to showcase their company profiles and access an extensive, growing network of potential collaborators and projects.
“Then, of course, you have all these SME’s in the service sector, who simply need
to market their services […] from the little one-man entrepreneur to the bigger firms.
Everyone has a need for greater visibility. So, we can see that we fill a function both
ways here. So yeah, we definitely solve those problems. Later, as we see it, when we
can teach the SME’s to collaborate, we’ll create totally different synergies and it’ll
be win-win all the way.”- MarkPro Founder
MarkPro’s current functionalities are search and filtering, job-posting and in-app messaging, which they believe to be the basics needed to establish MarkPro and enable this new possibility for networking and contracting. MarkPro also strives to create a community on their platform, creating newsletters and doing interviews with users to highlight individual interesting firms, trends or needs. Following its launch, MarkPro had some initial troubles with attracting and acquiring users. To address this problem, it enlisted the help of a local marketing firm who in turn managed to increase the number of users quickly. This resulted in MarkPro currently having around 260 registered users.
4.2 Realities of digitalisation
We now inquire into what digital tools the actors are utilising today and how they are using them in order to understand the level of digitalisation in the regions SME’s. A majority of the active firms wish to focus on the operative side of the business instead of having to spend time on administrative duties. Therefore, there has been a significant move toward implementing digital accounting, HR and salary systems to automate and streamline administrative processes. These efforts have, generally, been adopted and implemented with ease as there was such a drive from the actors themselves to be rid of administrative duties. Digitalising administrative processes meant that, for most, there needed to be an investment in smart devices to provide the necessary input for the different systems serving their new processes.
19
commonly used communication application in the SME’s was WhatsApp, which could explain the more grassroots approach as many of the younger workers are very keen smartphone users and could, therefore, implement WhatsApp as they were already using it. Many of our participants expressed the satisfaction adopting apps like WhatsApp due to it being free and still allowing for a much more efficient internal communication process, with a tolerated side-effect of non-work related chat in the groups.
“It works really well actually, we’re a kind of young crew […] 90% are between 20-27 years old so it’s natural for them to use digital tools and many of them used it at home already. It’s a pretty active group, there’s daily activity, fast messages and fast answers if something is needed. Then there’s about 50% funny greetings and stuff” – Firm #8
These smart devices are now an integral part of all participant firms, and some describe how carrying a smart device has replaced the need to bring along laptops in their cars and onto sites.
Furthermore, apart from the streamlining of internal processes through the adoption of digital tools, most actors advertise their services online either through a personal website, social media accounts or a combination of the two. Most cite the fact that it is an absolute must to have an online presence to reach potential customers browsing for services. Other more localised efforts in specific firms reflect the needs of their specific market niche. For example, a critical business process for firm 11 is to build 3D-models, make digital drawings and render videos, resulting in a need to acquire and procure the latest software. Other examples are a project management system which was discarded due to low and sporadic usage. While Firm #2 tried a B2C platform that allows customers to send inquires to firms connected to the platform, it terminated its membership due to unsatisfactory interactions and matchings.
“There are a number of these platforms that send out inquires, where people can
inquire firms […] It ended up with around 20 inquires every day ranging from all
kinds of projects, finally you just give up, you can’t even be bothered reading more
than the header and then delete” – Firm #2
The focus of this section was to through inquire about the general trends in the SME’s digitalisation efforts to understand if there have been any such initiatives like MarkPro before and how they have functioned. However, none of the participants mentioned networking as a focus area for digitalisation. Concerning digitalisation in general, many highlighted the intricacies and soft values in short-term contracting as barriers hindering firms from investing in tools to handle these intricacies internally.
4.3 The breakfast club
20
various firms would gather for a typical breakfast, exchange contact information and details as well as engage with guest lecturers. Entry to these meetings was sometimes paid for through membership fees to networking organisations, which hosted these meetings.
According to firm 2, the stated purpose of arranging these meetings, and the inherent value of creating a space that facilitates network growth bears several similarities to the intent and means proposed by MarkPros. The respondent further reflects that the meetings probably contributed to growing his firm’s network, as well as generating a fair number of projects and new business contacts across the local construction industry. He partly credits the output of these meetings to building their current network and emphasises that he believes these forms of initiatives may be more valuable to younger firms that have not generated a network of sufficient size for their needs. Firm 2 is not active in any similar initiatives anymore since it is deemed too time-consuming to spend several dedicated hours for networking every month. A withstanding experience also seemed to be the lacking ability of the hosts to sustain long-term interest from participants.
“The goal was to always invite more people. But at [some point] you hit the end of your network, and there’s not any additional actors interested in joining. As long as there are new people coming in, it stays alive.” – Firm #2
4.4 Fighting frictions
A majority of the participants highlighted the influence of seasonal fluctuations on the amount of available work, and its implications on what type of needs the firms have during the different seasons. During high-season, the firms are more vulnerable to the loss of the workforce since the whole business community is bustling. In general, this fact increases the amount of time a firm needs to spend searching their network for replacements. During low-season, the opposite conditions lead to firms approaching each other and offering their services to a higher amount.
“I usually say like this. If you by the 24th of december don’t know what you are going
to do in January or February… at least your’re not going to be working.”- Firm #4
21
Another area of friction which MarkPro had identified is to what degree firms perceive a lack of specialised or less common, but desirable services. The dominant view among the participants was that most services they desire to contract were covered within their networks since they are active in the same business areas. Competencies that were highlighted as missing in their current networks were mostly related to administration, legal and marketing services.
“This whole paper war… It would be good to have someone. The prospect of starting new companies, deeds, basically everything. Every year. Everything is new. Both from the insurance companies and… We’ve been trying to save it and make copies, but then it doesn’t work… and it starts all over again.” - Firm #11
Furthermore, as MarkPro had the perception that if firms could collaborate more smoothly, they could take on larger contracts, e.g. municipal contracts. Some of our firms explained that it is more complicated than that as they lack the internal administrative resources which is present in incumbents. Firm 1 explained that there is so much coordination and documentation that needs to be handled that they are not interested in taking on large projects. Firm 2 also felt that the added paperwork and demands from municipalities are not worth the effort, even though there is much money involved. Most firms felt that they are happy with their current situation of catering to the private market and that the barriers for entry in the large contract market were too high to incentivise them.
4.5 Dealing with contracts
The issues highlighted as most problematic by the respondents were related to managing contracting on short notice and the difficulty in evaluating the quality of subcontractors and building trust.
“They will be calling like. Asking when we can come. “Maybe in a week?” then they answer: No, now!”-Firm #3
Circumstances, where firms must either buy or sell services on short notice, was a well-known aspect of the construction industry, according to several of the participants. This is re-occurring through the seasons and was considered to be a frustrating and problematic element. Illness, cancelled projects and peaks in demand were explained to be events that create these circumstances. Taken together, the respondents highlight several dimensions of this issue. Firm 3 stated a desire to have greater flexibility in case that a lucrative contract from larger construction firms would be offered on a short notice, which was described as a commonly occurring in the industry. At the same time, firm 3 argued that having idle workers in preparation for this scenario was unacceptable. Larger companies looking to buy services were described by other participants as well to generally employ shorter notices, in comparison to other SME’s. An example of possible reasons for the urgency involved was explained by one of the participants as the presence of tightly coupled service chains.
22
Under the unfortunate circumstances that a project would be stranded in the sense as described in the quote, the same respondent expressed frustration over the prospect of calling his network in the role of a buyer searching for an extra resource. This process of searching can be drawn-out, calling several contacts, receiving declines before finding a suitable resource amount to hours spent over the year. In the case of a service chain at a stop, the hours of phone calls can prove to be very expensive. On the other hand, there was a participant who witnessed that sometimes there is more flexibility available in the time frame. However, buying actors at first tactically insists on a shorter frame. According to their experience there often is room for negotiation, which makes it easier for the selling side to be able to provide support.
“Everybody keeps their eyes and ears open. You, you shouldn’t mess up. It’s important. One big mistake, and it could all be over.” -Firm #3
The ability to accommodate resources on short notice is valued by actors looking to contract resources in the business community. However, a critical aspect raised by most participants is the importance of work quality and the element of trust involved in the assertion of what constitutes a high-quality resource. Different views on important qualities were expressed by the participants, partly influenced by their specific business context. Qualities such as high willingness to work, the ability to deliver on time, social skills among others were named. Throughout the descriptions of how the contracting searching process within a firms network plays out, the pattern that all of the participants firstly utilize their closest assets in their network resonated. Another common aspect was for firms to approach friends and family before turning to channels that are less known.
The same participant argued that there is a rift in the market between SME’s and larger staffing agencies, in regards to the overall qualities of the service. Staffing companies refer to firms solely focused on leasing resources on-demand. His perception of the market is that staffing agencies provide a cheaper service, but that the resources provided by these companies many times can show up at a construction site without tools or lack in terms of work management. His main argument is that utilizing staffing companies as a way to contract resources is a risky endeavour since, in his experience, the best staff reside in the smaller companies active in the business community. There has, however, been an increase of staffing companies in the last years which they view problematic from a competitive standpoint since they are influential in lowering the price of construction services through, among other measures, contracting resources from countries with lower salary levels.
23
4.6 Two sides of the same platform
MarkPro managed to gain some traction thanks to the outsourcing of sales and customer acquisition. As its user base expanded up to its current number of around 260 users, MarkPros’ founder noted that even though their userbase had increased, and that the reports from the sales firm were positive in their feedback regarding the sales process and customer reactions, activity in the userbase was still quite low.
“[...] we began picking up good speed in recruitment last autumn, but then we realised that noone is using it, so what are we going to do with more members if the ones who are already in can’t see any value. That’s when we changed course”
– MarkPro Founder
This new course entailed halting new customer acquisition efforts and focusing on activating the existing customers through stimulus efforts by the sales firm. Ranging from the configuration of their initial profile for them to get up and running smoothly, to having feedback conversations probing users for any potential needs they could assist them in listing on the platform. MarkPro was not suffering from any substantial attrition of users during this period of low activity, and many users were appreciative of their efforts in helping them with initial profile creation. The feedback was positive enough that they decided to apply this tactic in their new customer recruitment strategy when MarkPro decided they would resume recruitment, meaning that they offered new customers to help with setting up their profiles.
“We changed the strategy to make it easier to get them on board, where we create the profile for them with the basic settings and then send the login to them with the intent that they’ll finalise it on their own.” – MarkPro Founder
However, while the strategy was increasing the number of profiles to browse and details to filter searches on, activity was still not rising. MarkPro perceived that there was still the chicken or egg problem to crack, i.e. which user-groups to focus on amassing first in order to draw others in. The CEO argued that ignition in the community was going to kickstart the users and create the network and community of contracting collaborators they strived for. They would, however, need more outside help to realise igniting the activity on the platform. Therefore, MarkPro enlisted the services of local marketing and communications bureau for help with part managing the social media and website content to increase their visibility and part conduct interviews and write entries on the platform.
“We want [users] coming in through the windows and doors by their own free will and they should just feel like Wow, I’d like to be on there” – MarkPro Founder
The marketing scheme seemingly helped spread the word about MarkPro, and they continued to receive good feedback from the community regarding their concept. However, activity was still stagnant, and MarkPro expressed that they did not really understand the misfit between the good feedback when presenting the idea and the slow adoption rate of the platform.
24
platform functionality as inadequate and awkward, that the searching, querying and waiting for responses actually added to their workload in an unacceptable manner.
“[…] most importantly it needs to be user-friendly and that you find what you need, because if it’s not there once and you go on one more time and still nothing you know that you cannot spend time searching there. Then you’ll just go back to calling like usual.” – Firm #2
Also, the choice to primarily develop the platform as a web application aimed to cater a market transitioning from desktops and laptops to mobile devices as their primary workstation further adds to the inconvenience of the users.
“Earlier I always had my laptop in the car with me, but I’ve left that at home the last couple of years because I always have my phone with me” – Firm #4
The functionality offered for conducting a short-term contract seemingly also missed the mark with the users. Many felt the querying involved too much work and the waiting times too long to be a lucrative alternative to calling. That the core functionality when it comes to querying involved sending an e-mail through the platform to sellers frustrated users they felt their inboxes were already cluttered and searching through one's inbox was already a time thief. The users also highlighted that the mind-set has transitioned to a more on-demand and fast information consumption pattern. They were feeling that the waiting for an e-mail in their inbox seemed out of date and they were expecting something along the lines of push-notifications present in many applications, e.g. Instagram.
“I feel that... yeah it needs to be faster, pure and simple... fast and easy... and just
faster haha. Queries need to come to my knowledge much faster and without me
having to log in” – Firm #4
The user-interface was sub-par according to many respondents, also feeling out-of-date and hard to grasp. They wanted a cleaner interface with intuitive symbols, e.g. a green light for availability and red for not available. They felt that the current layout was simply too muddy and cluttered for them to search with ease and motivate themselves to use the platform.
“Yeah I mean.. I believe that it’s just too much work to update and all, if you have someone that is free, just have a green light or something next to them” – Firm #2
Lastly, the decision to utilise a subscription-based payment model seemed to antagonise many users as they felt they did not get enough value from the platform. The CEO described that in their first round of invoicing, at least two-thirds of the customers refused to pay. This realisation led MarkPro to subsidise the users that did not want to pay for a time so they could realise the potential of the platform and become paying users afterwards.
5 Discussion
In this section, we expand our analysis and discuss our findings concerning our stated research questions. We begin the discussion by analysing one aspect of MarkPro’s value proposition: its ability to function as a contact point for scaling SME’s networks, comparing the digital platform to previous initiatives within the business community aimed at accomplishing the same objective. Further, we discuss possible long-term aspects
25
identify processes that influence market frictions, specifically transaction costs related to contracting and compare how the functionality of MarkPro supports these processes.
The headings on the discussion all relate to different aspects of the research questions, and the subsequent discussions are intended to answer the questions and give insights into how some of the highlighted issues theoretically could be solved.
5.1 Abilities of shared spaces to facilitate network growth
In this section, we discuss the value of using various forms of shared spaces (breakfast meetings and digital platforms) as a contact point to growing a firm’s network. We analyse what types of value that are being offered to the participants during these forms of events and how this compares to the functionality provided by MarkPro’s digital platform.
Our findings show that the main benefits experienced by SME’s participating in breakfast meetings were a particular ability to scale their networks which is beneficial when searching for resources to contract. The participants acknowledged that the hosts of these meetings, in general, invited a broad spectrum of actors. Although there was some initial social awkwardness involved in the start, it would occasionally lead to new contacts being formed between actors which may have projects of interest of working with each other. During these meetings, actors are provided by the host with an opportunity to meet, discuss and get a general feel for each other. This would then lead to the exchange of contact information, effectively expanding their network of possible resources and further communication proceeds through the regular channels of the actors, outside to the shared space where they have met. To summarize, through participation in a time-constrained event, the actors get connected with new actors, and at the same time evaluate them through conversational introductions and interaction.
MarkPro’s ambitions are similar in the sense that they have a stated goal of digitally connecting various actors which may have similar interests. The current functionality that corresponds to the purpose of facilitating network growth practised by MarkPro is a database of connected actors, search and filtering and a five-star rating system. For actors connected to the digital platform, this is implied to improve networking by providing actors access to an extensive network and the tools to sort out actors in order to improve match quality. The similarities with the physical spaces are that both provide a contact point with connected actors and a form of self-stated description of what they have to offer other actors. A difference between the physical shared spaces and MarkPro’s digital platform is the time limitation of the breakfast meetings and how this influences the probability of interaction failures between actors.
26
that would suggest that they actively orchestrated their ecosystem in a way that would attract actors to interact on their platform during a set timeframe. Instead, MarkPro’s approach has been to continuously market their users in the form of introductory posts in a newsfeed on their platform, under the implication that other actors will see the posts and initiate contact. In this sense, we argue that the defining difference between the orchestrating of the physical shared spaces and Markpro’s digital shared space is a matter of the degree of activeness in facilitating interactions on behalf of the matchmaker. Where physical shared spaces actively facilitate interactions by bringing together actors during limited time frames, Markpro’s strategy has been to facilitate interactions by marketing passively. In the time-limited scenario actors who participate are guaranteed to notice each other during the set timeframe, wherein the non-time-limited scenario actors will notice the marketed actor the next time they log in to the service. Due to this difference, we would argue that for interactions to take place in a digital platform, it would require what Van Alstyne, et al., (2016) refers to as a high level of engagement for this form of strategy to be effective. In its current form, we would argue that the data suggests that using Markpro’s digital platform to facilitate the initiation of contacts and grow the network of SME's
generates more interaction failures in comparison to earlier initiatives such as breakfast
meetings.
For Markpro to increase the number of interactions and match quality between actors, we argue that they would either have to manage time-limited events where actors would notice each other by design, a direct equivalent to the breakfast meetings would be to host online meetings in some form. Another measure to accomplish this would be to make contact on behalf of actors looking to grow their network. This could either be performed by the staff of Markpro contacting potential candidates for a match or, as commonly applied by platform owners (Van Alstyne, et al, 2016); utilizing technology which makes a relevant contact and returns an answer on whether the corresponding actors would be interested.
Furthermore, MarkPro as a platform has no extensible codebase available to third party developers with regards to its architecture, meaning that there is no novel functionality being implemented by external developers (Tiwana, et al, 2010). Therefore, MarkPro is a
closed platform as defined by Smedlund & Faghankani (2015) and can only be developed
