The effect of firm size and firm experience on the success of an app development firm

(1)

The effect of firm size and firm

experience on the success of an app

development firm

Master Thesis – MSc BA Small Business & Entrepreneurship

University of Groningen

Faculty of Economics and Business

By: Patrick Keizer - s1771027 Supervisor: Dr. F. Noseleit Co-assessor: Dr. Ir. H. Zhou Date: 16 June 2016

(2)

This study examines whether there is a relationship between firm characteristics and the success of an app development firm in app store markets. This research is a first step in identifying firm characteristics that influence the success of app development firms in app store markets. An analysis is provided that shows the relationship between firm size and app development firm success. Also an analysis is provided that shows the relationship between firm experience and the success of an app development firm. The analyses were based on a hand-‐collected dataset of top 100 grossing ranking charts of 34 days in four countries. The grossing top 100 ranking charts show the most profitable apps. Not much research could be found about the influence of firm size and firm experience on the successful deployment of software in the current literature base. What could be found though in the literature is that firm size has a significant negative relationship on the innovativeness of a software development firms. Furthermore it might be expected that firm size is also of relevance for the success of apps, as general research on small firms shows small firms are more successful in new markets. Literature also shows that firm experience is not or less helpful for software firms than for manufacturing firms for the success of new products developed by the same firm. Researchers identified that this is caused by the fact that software firms do not enhance their work processes based on past experience. Previous research did not identify whether there can be a difference identified between the level firm experience enhances work processes in software firms located in different countries. Interestingly this research finds that countries showing a significant relationship between firm size and the success of an app development firm do not show a significant relationship between firm experience and the success of an app development firm and vice versa. Further research is needed to analyse why the differences between countries appear. Also more research is needed to identify a more clear view about the relationship between firm size and firm experience for software development firms.

(3)

Acknowledgements

Since the age of thirteen I am running a business in the information technology industry. Therefore during my university studies I was always trying to perform research about subjects that combine business developments and information technology developments. I feel that new trends in information technology change the way of doing business. I also believe that without running a business into the right direction new information technologies will not be deployed on the market.

New technologies could make new business opportunities possible. This is one of the reasons why I support the view that the research fields business management and IT need to combine their knowledge and skills. With this master thesis I provide new insights that enhance (new) business opportunities and adds knowledge to the current business & economics literature base.

One of the current trends is that people use more and more mobile devices. This makes business investors realize that they have to invest in new mobile technologies. Beside devices (hardware) that are deployed in the market people run software on those devices. Business people say most often that time is money. As it takes a lot of time to develop successful mobile applications, research could help businesses to get a better understanding of app markets by providing more insights about what kind of (new) mobile projects/firms are worth to invest in. By enhancing the knowledge base of stakeholders that act in app markets the business and economic developments could also be enhanced.

This research provides a start in collecting insights about (new) mobile app market(s). The outcomes could help investors and stakeholders of firms to make better business and investment decisions so that (new) opportunities can be utilized as efficient as possible.

I would like to thank my supervisor Dr. Florian Noseleit for his patience and the way he was able to direct me into new directions. He gave me the inspiration to discover new insights.

(4)

listen to my stories about new developments concerning my thesis. Especially my brother Maarten and my sister Michèle reflected on my thoughts. This was very helpful.

(5)

Table of tables

TABLE 1: European Commission firm size classification TABLE 2: Firm success measures

TABLE 3: Types of top-‐ranking lists TABLE 4: Firm sizes LinkedIN TABLE 5: Data sources

TABLE 6: Significance compsizelevel TABLE 7: Significance totalappsdeveloped TABLE 8: Significance ln_totalappsdeveloped

(7)

1. Introduction

Nowadays apps have taken the world of most smartphone and tablet users (Peabody, 2012). Google followed soon after the launch of the Apple App Store in Mid-‐2008 with their own app store. Currently there are more than 2 million apps available in the Apple App Store (‘Number of apps available in leading app stores as of June 2016’, 2016). Also in the Google Play store more than 2 million apps are available nowadays.

The global app economy is growing at a compound annual growth rate of 28% (2012 to 2016), and it is forecast to be worth $143bn in 2016 (Hubbard, 2015). In 2014 Gartner published a report that states that it is important for development firm stakeholders to recognize that most applications are not generating any profits and that many mobile apps are not designed to generate revenue. Gartner’s results show that of the paid applications about 90% are downloaded less than 500 times per day and make less than $1,250 a day (‘Less than 1% of apps to be financial successes: Gartner’, 2014). Therefore he concluded that most apps are developed to build brand recognition and product awareness or are just for fun. With increasing competition in the app market this is going to get even worse, especially in successful markets.

Lee and Raghu (2014) analysed which attributes are influencing the success of apps. The authors found that free apps offer, high initial ranks. Continuous quality updates as well as high-‐volume and high-‐user review scores are attributes that influence the success of apps. Jung, Baek and Lee (2012) found that customer ratings, content size and early entrants affect the success of apps. Many vendors are deploying apps nowadays which makes the risks of deploying an app high, also when the previously mentioned attributes are taken into consideration. There is much competition so most likely more marketing, financial and other resources are needed to make a single app a success.

(8)

provide stakeholders and investors a better understanding of the type of firms it is worth to invest in. The identification of these characteristics could also provide an inside in what kind of firms to research when trying to find an answer on the question which kind of strategies of app development firms are more successful than other strategies.

Tsvetkova, Thill and Strumsky (2014) state that firm size is the most studied determinant of business survival. According to Tarus and Sitienei (2015) no relationship can be identified between firm size and the successful development of software products. It can be questioned whether this is also the case for software firms that develop apps for new online markets.

Firm experience is also a firm characteristic that is researched a lot in the current literature base. The authors Lyytinen and Robey (1999) show that no relationship can be found between past firm experience in software development and the successful deployment of (new) software projects. As stated by the authors the reason for this is that software development firms are not able to learn from experience.

The aim of this research is to provide some first insights concerning the influence of firm characteristics on the success of an app development firm. This research will be limited to the relationship between the success of an app development firm and the firm characteristics firm size and firm experience. The outcomes will provide stakeholders with a better understanding of the firm characteristics that are needed to become successful as app development firm on app markets.

(9)

2. Theoretical framework

This research tries to find evidence for the influence of firm size and firm experience on the success of an app development firm. The central question to be discussed is “What is the effect of firm experience and firm size on the success of an app development firm?”. First the definitions used will be described in the theoretical background section in paragraph 2.1. Secondly the hypothesis will be developed based on the literature mentioned in paragraph 2.2. And in paragraph 2.3. the hypothesis will be summarized. 2.1. Theoretical background

2.1.1. Firm size

Tsvetkova et al. (2014) state firm size as the most studied determinant of business survival. It is a firm specific characteristic frequently used to analyse issues dealing with structure, behaviour and strategies of corporate enterprises (de Brentani, 1995). Firm size is measured in terms of various metrics such as number of employees, assets and sales volume (Tsvetkova et al., 2014). The current number of employees is preferred as a measure of firm size in the business survival literature (Tsvetkova et al., 2014). The European Commission uses the firm size classification shown in table one which is also based on the number of employees.

TABLE 1: European Commission firm size classification

Size Description

Micro Fewer than 10 employees

Small Fewer than 50 employees

Medium Fewer than 250 employees

Large More than 250 employees

Source: Tarus et al., 2015

(10)

products. Both small and large firms can be found that have grown as the result of being successful innovators (Ettlie and Rubenstein, 1987). Yin and Zuscovitch (1996) explain that small firms are more likely to be successful in new product markets than large firms and large firms are the ones who dominate the post-‐innovation market. As described by Mainiero and Tromley (1968) firm size influences the issues a firm is facing. Firms who retain the same number of employees show the same problems over lengthy periods. When size increases firms show other problems such as coordination and communication problems.

2.1.1.1. Firm size and software development firms

Up to my best knowledge Tarus et al. (2015) are the only authors that analysed the effect of firm size on the output of software companies. As shown by the authors a significant negative relationship occurs between firm size and innovativeness of software firms. According to Tarus et al. (2015) small and medium sized software firms are better innovators than larger firms because of their informality and fewer intra-‐firm hierarchy levels.

2.1.2. Experience

Firms knowledge stocks are created by past experience (Buelens et al, 2006). Two types of knowledge stocks are available: the internal knowledge stock (own experience) (paragraph 2.1.2.1.) and the external knowledge stock (other’s experience) (paragraph 2.1.2.1.2.).

2.1.2.1. Internal knowledge stock

(11)

learning change the internal knowledge stock and the level of experience of a firm. Measures such as the number of patents (Antonelli et al., 2015) and the number of development projects performed (Lyytinen et al., 1999) are used to calculate the level of internal knowledge available in the internal knowledge stock. These measures can be used as measures for the internal knowledge stock because working skills improvements are facilitated by past production experience (Bhandari, 2010).

2.1.2.2. External knowledge stock

External experience can be accessed via knowledge interactions and transactions with external sources such as suppliers and customers (Antonelli et al., 2015). When internal knowledge is unavailable, external knowledge is often the only means for most organizations to learn (Lyytinen et al, 1999). Business cooperations can be an important route to access and transmit knowledge and experience (Urbancová, 2013). The validity, transferability and relevancy of external knowledge is questionable. According to Urbancová (2013) small firms have a reduced innovative autonomy and less collaborations with technological centers, this is why it is even more important for small firms to collaborate to access and transmit knowledge.

2.1.2.3. Mutually exclusive

As explained by Antonelli et al. (2015) the generation of knowledge based on internal and/or external experience are mutually complementary. This means that no firm can generate knowledge without the access to external experience and no firm can generate knowledge without appropriate internal experience.

2.1.2.4. Firm experience and software development firms

(12)

software development tasks enhances next project work-‐processes. However according to Lyytinen et al. (1999) software development organizations fail to learn from experience, because these firms show limits of organizational intelligence, disincentives for learning, organizational designs and educational barriers. Lyytinen et al. (1999) explain that software development organizations could learn from experience if they use the experience developed during their own internal projects for modifying their theories/practices in use. But the authors found that modification of theories/practices based on experience is not done by software development organizations so software firms do not learn from experience.

2.1.3. Success

Wright, Gardner and Moynihan (2003) describe six major measures for performance used by the firm headquarters. According to Wright et al. (2003) these measures are an indicator for business success. The metrics shown in table 2 are used to measure business performance.

TABLE 2: Firm success measures

Metric Description

Workers compensation / Sales Worker’s compensation divided by sales.

Quality 100,000 pieces per error.

Shrinkage % inventory loss.

Productivity Payroll expenses for all employees divided by the number of pieces.

Operating expenses All relevant business operating expenses.

Profitability Pre-‐tax profit of the business unit as a percentage of sales.

Source: Wright et al., 2003

Other authors such as Mithas and Rust (2016) use profitability as only firm success variable measured as revenue minus costs.

2.1.4. App store

(13)

data. At the moment multiple app stores are available for example the app store of Apple, Google Play/Android and Windows. The app markets are growing which is a good reason to analyse important questions about topics such as software innovation, firm entry and exit strategy e.g. (Garg et al., 2013). However it is difficult to analyse app store trends because Apple, Google and Windows do not provide app sales information (Lee et al., 2014). Most app stores only provide the three ranking list types that are shown in table 3.

TABLE 3: Types of top-‐ranking lists

Ranking list Description

Top-‐free Most downloaded applications without upfront purchase price.

Top-‐paid Most-‐downloaded applications that have a non-‐ zero price.

Top-‐grossing Most revenue generated applications.

Source: Garg et al., 2013

The top-‐grossing ranking list combines the most revenue generating free and paid apps in a single ranking chart (Lee et al., 2014). Revenue includes the amount paid for downloading the app plus the revenue generated by in-‐app purchases. In-‐app purchases create additional revenue generated by purchases of content, functionality, services, or subscriptions bought via free or paid apps. According to Garg et al. (2013) an app listed in the top 200 of the iPad rankings would generate about 100 downloads per day. Garg et al. (2013) calculated that an app ranked at position 1000 would generate only about 25 downloads per day. Mid-‐2015 more than 1.5 million iPhone apps where downloadable for apple store users. This is why Garg et al. (2013) conclude that most apps generate little to no demand. For most app stores the top ranking lists are available per country in multiple categories. Apple provides the top ranking lists in 24 different categories such as games, books, education and business per country. An overview of all categories is provided in appendix A. Apple also provides an overall top 100 ranking list which includes the top scoring apps covering all categories.

(14)

2.2.1. The relation between firm size and app success

Firm size is one of the most used determinants of business survival (Tsvetkova et al., 2014). As described by Ettlie et al. (1987) it does not always mean that larger firms have more change to be innovative or successful. In literature there are significant differences found between small and medium firms and large firms (Mainiero et al., 1969). Tarus et al. (2015) are one of the first researchers providing significant results between the relationship of firm size and innovativeness of software firms. The authors found that small firms are more successful innovators than their larger counterparts, but this result does not imply that relatively large software development firms are less successful in deploying apps. App development firms are software firms that produce software for new markets named app markets. As shown in the research of Yin et al. (1998) small firms are more successful in producing products for new product markets. This is why it can be expected that small firms deploy more successful apps for app markets. This leads to the following hypothesis:

Hypothesis 1: Large sized firms do not deploy more successful apps than their smaller counterparts.

2.2.2. The relation between firm experience and app success

(15)

will be no correlation between past experience of a firm and the success of a deployed app. This leads to the following hypothesis:

Hypothesis 2: There is no effect between firm experience and the success of an app.

Hypothesis two tests for the effect between firm experience and the success of an app. If a development firm learns from experience it is most likely the case that not all developed apps will perform as well. In order to get a better understanding of how firm experience affects the successful deployment of apps it also needs to be tested if the overall success of all deployed apps on app development firm level is higher for a more experienced firm. As Lyytinen et al. (1999) describe software firms do not learn from experience so it is expected that the overall success of app development firm’s portfolio is also not affected by past firm experience. This leads to the following hypothesis:

Hypothesis 3: There is no effect between firm experience and the success of developer’s app portfolio.

2.3. Developed hypothesis

Summarizing the hypotheses that will be analysed in paragraph 4 are the following:

(16)

Authors of empirical articles that follow the hypothetic deductive model use theory to formulate hypotheses before testing those hypotheses with observations (Colquitt and Zapata-‐phelan, 2007). The hypotheses in this thesis are theory-‐deduced hypotheses and are tested during the research. The data collection methods used are described in section 3.1. In section 3.2. an analysis plan is provided. The controllability, validity and reliability of the data used is described in section 3.3.

3.1. Data collection methods

(17)

automatic PHP script is created that automatically reads and stores the amount of apps developed by a firm. The total number of developed apps per developer is added to the database one day after finishing the collecting process of the top 100 ranking charts. The RSS feeds of Apple do not include the developer firm sizes. LinkedIN provides an external source including the firm sizes of most firms worldwide. This LinkedIN source is accessible via an API for LinkedIN developers. After registration of a LinkedIN developer account it was possible to gather firm sizes via the LinkedIN API using a self-‐ written PHP script. The dataset is enhanced with the firm sizes of all firms provided by the LinkedIN API. LinkedIN uses a standard classification for firm sizes. This classification is shown in table 4.

TABLE 4: Firm sizes LinkedIN

Type URL A Self-‐employed B 1-‐10 employees C 11-‐50 employees D 51-‐200 employees E 201-‐500 employees F 501-‐1000 employees G 1001-‐5000 employees H 5001-‐10,000 employees I 10,001+ employees Source:https://developer.linkedin.com/docs/reference/company-‐size-‐codes

(18)

TABLE 5: Data sources

Source URL Free/Paid

Apple’s RSS feed http://www.apple.com/rss/ Free Apple Developer page https://itunes.apple.com/nl/developer/apple/id<developerID> Example: https://itunes.apple.com/nl/developer/apple/id284417353 Free

LinkedIN API https://developer.linkedin.com/ Free after

registration 3.2. Variables and measures

In this research as many as possible measures are based on existing validated scales from literature. Below the measures are listed per variable type.

Independent variables

Firm size. Firm size is measured as the amount of employees provided by LinkedIN and re-‐grouped in the categories used by the European Commission shown in table 1. 1 = Micro, 2 = Small, 3 = Medium and 4 = Large. The variable is stored in the dataset using the name compsizelevel.

Firm experience. Firm experience is measured as the total number of apps deployed by an app development firm collected one day after collecting the last top 100 ranking chart data. The variable is labelled with the name totalappsdeveloped in the dataset. The variable totalappsdeveloped is also available in a log-‐transformed version called ln_totalnumberofapps. For analyses reasons an extra variable is generated called totalappsdev_minus_current that shows the totalappsdeveloped minus the total apps that are developed by a firm and appeared in the top 100 grossing ranking charts.

Dependent variable

(19)

top 100 ranking charts show the apps that generated the most revenue. The grossing top 100 ranking chart shows indirectly which app development firms are the most profitable. The success of an app is measured as the number of days an app survived on the grossing top 100 ranking charts in the category “all”. The number of days an app survived is used because Apple does not release actual sales figures to the public (Lee et al., 2014). As noted by Garg et al. (2012) it has become common in academic research to use data top ranks as measure for actual sales or in lieu of sales. The total number of grossing days per app is available in the dataset and labelled as grossingdays.

Success of an app development firm. Most app development firms develop multiple apps. The sum of all grossing days of all apps that appeared on the grossing top 100 ranking charts per firm is measured and stored in a variable called sum_grossingdays. This variable makes it possible to analyse which firms have the most profitable app portfolio.

3.3. Analysis plan

(20)

3.4. Controllability, validity and reliability

Van Aken et al. (2012) describe three quality criteria namely: controllability, validity and reliability. The quality criteria of Van Aken, Berends and Van der Bij (2012) are applied during the research.

In this research all required instruments and methods are described in detail to make sure that other researchers can repeat this research ending up with exactly the same results. Being able to repeat this research based on the explanation provided makes this research controllable according to Van Aken et al. (2012).

Validity consists of three constructs namely construct validity, internal validity and external validity (Van Aken et al., 2012). A necessary condition for theory development and testing is the validity of constructs (Steenkamp and Van Trijp, 1991). Construct validity is met when the concept is covered completely and when measurement is performed without components that do not fit the concept. The supervisor of this research is asked as an expert to check whether the applied components are valid and complete. Also the current literature base is used to check for construct validity. To enhance the internal validity of the research outcomes the top 100 ranking chart data is collected and added to the dataset for as many days as possible. This research is limited to the top 100 grossing ranking chart datasets of four countries of the Apple App Store and 34 days. The group of app users that use other stores such as the Windows and Android store are neglected in this research. This could affect the external validity of the research outcomes.

As noted by Steenkamp et al. (1991) reliability is the level to which measures are free from random errors. If random errors do not occur the same outcomes can be conduced if the research is repeated in the same way. The methods used for establishing the results are described as objective as possible to make sure that it is possible to repeat the research in the same way and to avoid random errors.

3.5. Dataset statistics

(21)

ranking charts are shown in table 6. A dataset summary/description per country is provided in Appendix C. Please note that the complete dataset collected contains more than 34.000 apps and 20.000 developers and is filtered by the category “general” to avoid duplicates that could affect the results. After filtering the data the dataset consists of the amount of observations shown in table 6.

TABLE 6: Dataset characteristics

Top Chart 100 Category Observations / Apps Days measured

(22)

In this research it is tested whether firm size and experience have influence on the success of an app development company. In the below three sub-‐paragraphs the results of the hypothesis tests are provided.

4.1. Hypothesis 1

Hypothesis one states that larger firms do not deploy more successful apps than their smaller counterparts. The dependent variable grossingdays and the independent variable compsizelevel are used to reject the hypothesis. The Poisson distribution seems to be impropriate to use for analysis because the dependent variable shows signs of over-‐dispersion. However the Poisson distribution results are shown in Appendix G. The Poisson goodness-‐of-‐fit results (gof) are also shown in Appendix G. The gof results show very large chi-‐square values, which is also not a good indicator to proceed the analysis using the Poisson distribution. To proceed the analysis the negative binomial distribution is used which allows the variance to be greater than the mean. The results of the negative binominal distribution test are provided in Appendix H and the level of significance is shown in table 6. The negative binominal distribution analysis results show a likelihood ratio higher than 800 for each country. This confirms that the Poisson distribution results are indeed in-‐appropriate.

TABLE 6: significance compsizelevel

Country P>|z| DE 0,039 NL 0,023 UK 0,065 US 0,096

(23)

successful for the top 100 of UK and US apps. This means that the hypothesis can be partly rejected. For the top 100 ranking charts of NL and DE the hypothesis can be rejected. For the top 100 ranking charts of the UK and the US the hypothesis can be confirmed.

Hypothesis two states that firms that have more experience in app developing do not deploy more successful apps. The dependent variable grossingdays and the independent variable totalappsdeveloped are used to reject the hypothesis. The Poisson distribution test results are shown in Appendix I. The gof results show very large chi-‐square values, which is an indicator that the Poisson test is not reliable to test the hypothesized relationship. As alternative test for the Poisson distribution test a negative binominal distribution test is performed. The results of the negative binomial distribution are provided in Appendix J. If the results are significant it can be stated that firms that deployed more apps are more successful than firms that are less experienced in app developing. The measured significance levels are shown in table 7 per country.

TABLE 7: significance totalappsdeveloped

As shown in table 7 no significant results are found (all P>|z| < 0,05). The results confirm the stated hypothesis. As shown in Appendix K one or two developers created more than 600 apps. In order to avoid that outliers affect the results the log-‐transformed variable ln_ totalappsdeveloped is used and the data is re-‐analysed using the negative binominal distribution test. The results are shown in table 8 and in appendix L.

TABLE 8: significance ln_totalappsdeveloped

(24)

NL 0,390

UK 0,947

US 0,771

The outcomes of the negative binomial distribution test using the log-‐transfermed variable ln_totalappsdeveloped do not show significant results (all P>|z| > 0,05). As stated in the hypothesis it cannot be confirmed that firms with more experience in app development deploy more successful apps. The hypothesis is confirmed.

Hypothesis three states that there is no effect between firm experience and the overall success of software developer’s app portfolio. In order to test this hypothesis a new variable is introduced which is the sum of the total number of grossingdays of all apps developed by an app developing firm. This variable is called sum_grossingdays and is used as dependent variable. Some firms deployed a lot of apps and just appear with one app in the top 100 grossing days ranking charts. However other development firms deployed just one app and appear for a very long period in the top 100 grossing days ranking charts with this one app. In order to rectify for this a new variable is introduced which shows the total apps developed minus the apps that appeared in the top 100 grossing days ranking charts during the measurement. The variable is called totalappsdev_minus_current and is used as independent variable. The Poisson distribution test results are shown in Appendix M and the binominal regression tests are shown in Appendix N. The measured significance levels provided by applying a negative binominal distribution test are shown in table 9 per country. The outcomes are significant for all countries.

TABLE 9: significance sum_totalappsdev_minus_current

Country P>|z|

DE 0,003

(25)

UK 0,006

US 0,006

Because outliers are found in the dependent variable sum_totalappsdev_minus_current the test is re-‐done using a new log-‐transformed variable ln_sum_totalappsdev_minus_current. The outcomes of the negative binominal distribution test are provided in Appendix O and the significance levels are shown in table 10. The United States and the United Kingdom top 100 grossing ranking chart data show significant results for the relationship between the independent variable sum_totalappsdev_minus_current and the dependent variable sum_grossingdays. This indicates that development firms are found that are able to learn from experience of previously build apps in the UK and the US top 100 grossing ranking chart dataset. However firms in the German and Dutch dataset do not show significant results for this relationship.

TABLE 10: significance ln_sum_totalappsdev_minus_current

As stated in the hypothesis there is no effect between firm experience and the success of developers app portfolio, but this result is only found in the Dutch and German dataset. The hypothesis can be confirmed for the German and Dutch dataset. However the

(26)

The aim of this research is to answer the following research question: What is the effect of firm experience and firm size on the success of an app development firm?

In order to answer this question it is measured whether firm size is important to become successful as an app development firm and it is measured whether firm experience in app development has an effect on the success of app development firms. In this research the most successful firms are identified as the firms that earned the most profits via the Apple app store by selling apps and/or in-‐app content during 34 days of measurement.

The study provides two main outcomes that can help stakeholders to understand what kind of software firms are more successful in app development.

The first outcome of this research shows that firm size does have an effect on the success of deployed apps. However this relationship is only found in the German and Dutch dataset. No significant relationship between firm size and the success of deployed apps is found for the dataset of the United Kingdom and the United States. Based on the research of Yin et al. (1998) it was expected that small size firms are more successful than their larger counterparts in the deployment of new products for relatively new markets such as app markets. As this research shows the relationship between firm size and app success differs between countries.

(27)

software development firms are able to learn from experience, but show a lower learning rate than manufacturing firms. It could be the case that app software development firms in countries such as the United States and the United Kingdom where learning seems to appear found ways – perhaps based on constructing intellectual schemas as described by Sacks et al. (1999) – to enhance their work processes based on past experience.

At the time of writing (June 2016) not much is known in the current literature base about learning behaviour within software development firms. This is why it is not possible to provide a clear reason for the differences in results between the measured countries in this report.

TABLE 11: Outcomes

Country H1: significant H2: significant H3: significant

DE Yes No No

NL Yes No No

UK No No Yes

US No No Yes

(28)

etcetera. These sources could make smaller firms with fewer resources also experienced. In this research the relationship between firm size and firm experience and the correlation and influence of both variables on the success of an app development firm is not analysed. Further analysis needs to be done to check for multicollinearity and the correlations between the previously mentioned variables.

To summarize the research question of this paper is “What is the effect of firm experience and firm size on the success of an app development firm?”. And the main outcome of this research and answer of the research question is: the effect of firm size as well as the effect of firm experience on the success of an app development firm in app stores differs per country. Not enough literature was available to provide a proven explanation for the differences shown between countries.

5.1. Theoretical and managerial implications

This research is a first step in the understanding of the relationship between firm characteristics and the success of app development firms. The outcomes of this research show that for some countries there is a significant influence of firm size on the success of a development firm in app stores. Other countries show a significant effect of firm experience on the success of development firms in app stores. This means that it is proven that firm characteristics have a significant influence on the success of a development firm in app stores. For stakeholders it is important to know what kinds of firm characteristics do influence the success of app development firms. This helps stakeholders to make better (investment) decisions. The author asks for further research to analyse the relationship between the variables firm size, firm experience and the success of app development firms in more detail.

5.2. Limitations and further research

(29)

(30)

References

Abdul-‐Talib, A. N., Salleh, M. F. M., Shamsuddin, F. M., & Ashari, H. (2011). The effects of firm size and international business experience on export attitudes. Journal of

Competitiveness Studies, 19(1/2), 4.

Antonelli, C., & Colombelli, A. (2015). External and internal knowledge in the knowledge generation function. Industry and Innovation, 22(4), 273-‐298.

Bhandari, R. (2010). Role of Grids for Electricity and Water Supply with Decreasing Costs for Photovoltaics (Vol. 15). kassel university press GmbH.

Buelens, M. (2006). Organisational behaviour. New York: McGraw-‐Hill.

Colquitt, J.A. & Zapata-‐phelan, C.P. (2007). Trends in theory building and theory testing: a five-‐decade study of the academy of management journal. Academy of Management Journal 50 (6), 1281-‐1303.

de Brentani, U. (1995). Firm size: implications for achieving success in new industrial services. Journal of Marketing Management, 11(1-‐3), 207-‐225.

Eldring, J. (2009). Porter’s (1980) Generic Strategies, Performance And Risk: An Empirical Investigation with German Data, Diplomica Verslag GmbH, Hamburg

Ettlie, J. E., & Rubenstein, A. H. (1987). Firm size and product innovation. Journal of Product Innovation Management, 4(2), 89-‐108.

(31)

Garg, R., & Telang, R. (2012). Inferring app demand from publicly available data. MIS Quarterly, Forthcoming.

Hubbard, J. (2015). Why businesses need (beautiful) apps. Finweek 36-‐37

Jung, E. Y., Baek, C., & Lee, J. D. (2012). Product survival analysis for the App Store. Marketing Letters, 23(4), 929-‐941.

Lee, G., & Raghu, T. S. (2014). Determinants of mobile apps' success: evidence from the App Store market. Journal of Management Information Systems, 31(2), 133-‐170.

Lyytinen, K., & Robey, D. (1999). Learning failure in information systems development. Information Systems Journal, 9(2), 85-‐101.

Mainiero, L., & Tromley, C. (1968). Evolution and Revolution as Organizations Grow. Management, 113.

Mithas, S., & Rust, R. T. (2016). How information technology strategy and investments influence firm performance: Conjectures and empirical evidence.MIS Quarterly, 40(1), 223-‐245.

Paleti, R. (2016). Generalized Extreme Value models for count data: Application to worker telecommuting frequency choices. Transportation Research Part B:

Methodological, 83, 104-‐120.

Peabody, A. (2012). Invation of the App’s. www.iicensemag.com, 134-‐144

Sacks, M. (1994). On-‐the-‐Job Learning in the Software Industry. Westport: Quorum Books.

Steenkamp, J. B. E., & Van Trijp, H. C. (1991). The use of LISREL in validating marketing constructs. International Journal of Research in marketing, 8(4), 283-‐299.

(32)

development firms: the moderating role of firm size. Journal of African Business, 16(1-‐2), 48-‐65.

Typologies, F. B. (2008). The Influence of Family Business Size on Management Activities, Styles and Characteristics. New England Journal of Entrepreneurship, Fall 2008, 47.

Tsvetkova, A., Thill, J. C., & Strumsky, D. (2014). Metropolitan innovation, firm size, and business survival in a high-‐tech industry. Small Business Economics, 43(3), 661-‐676.

Urbancová, H. (2013). Competitive advantage achievement through innovation and knowledge. Journal of Competitiveness, 5(1).

van Aken, J., Berends, H., & Van der Bij, H. (2012). Problem solving in organizations: A methodological handbook for business and management students. Cambridge University Press.

Wright, P. M., Gardner, T. M., & Moynihan, L. M. (2003). The impact of HR practices on the performance of business units. Human Resource Management Journal, 13(3), 21-‐36.

Yin, X., & Zuscovitch, E. (1998). Is firm size conducive to R&D choice? A strategic analysis of product and process innovations. Journal of Economic Behavior & Organization, 35(2), 243-‐262.

Websites

Number of apps available in leading app stores as of June 2016 (2016, June) Retrieved from http://www.statista.com/statistics/276623/number-‐of-‐apps-‐available-‐in-‐leading-‐

app-‐stores/

Less than 1% of apps to be financial successes: Gartner (2014, January 13) Retrieved from

http://timesofindia.indiatimes.com/tech/tech-‐news/Less-‐than-‐1-‐of-‐apps-‐to-‐be-‐

(33)

Appendix A: App Store categories

Category ID Category title

Leave this field empty during the request All categories

6018 Books 6000 Business 6022 Catalogs 6017 Education 6016 Entertainment 6015 Finance 6023 Food&Drinks 6014 Games

6013 Health & Fitness

(34)

Total days measured: 34 days

Date Status Free Cat

P1 Free Cat P100 Paid Cat P1 Paid Cat P100 Grossing Cat P1 Grossing Cat P100

The effect of firm size and firm experience on the success of an app development firm