New software startups are born everyday around the world. Nonetheless, failure is the fate of most of them. The community already knows that several facts, such as market competition or lack of resources, can impact the destiny of a startup. However, little has been explored in term of the impact of software startup education on the success of failure of startups. In this sense, this study presents the initial steps that we are taking to understand how software startups are taught around the world. To do so, we design a qualitative survey aimed at software startup educators at Universities. Our goal is to gather enough information so we can help the academic community in improving their own courses. So far, we have gathered 10 responses from lecturers across the globe. This paper describes these ndings.
In the last decade, we have witnessed signi cant advances in technology, specially
after the popularization of the Internet. Today, any person with software
development skills is able to create applications that can be reached by millions (and
even billions) of people [
These technology endeavours, that are developed under high uncertainty, are
called startups [
From an education perspective, several software engineering/computer
science courses have focused on entrepreneurship in the last years ([
One of the biggest challenges reported on these studies is the lack of a realistic
environment for student to work on their startups [
In this sense, the goal of this paper is to understand how software startup is taught by lecturers/professor across the world. So far, we do not know how courses are carried out aside from papers describing individual experiences. Therefore, this study focuses on the following research questions: { RQ1: Aside from conventional lecture-based courses ending in an exam, how are software startups taught in universities? { RQ2: How do these courses deal with the multidisciplinary nature of a software startup (business, technology, design)?
The remainder of this paper is organized as follows. Section 2 presents the related work. Section 3 describes the research method used in this research. In Section 4 we show our preliminary results. Finally, we draw our conclusions in Section 5. 2
In this Section we de ne and explain in details what a software startup is. In addition, we depict how software startup education is being explored by the academic community. 2.1
Though software startups have recently had a large economic impact across
the globe, no clear consensus on what exactly a software startup is exists [
Startups operate under a lack of resources, both in terms of time, manpower,
and nances [
Perhaps the most important di erence between a conventional small
business and a startup is that startups are characterized by clear intentions for high
growth. While small companies generally wish to grow, and will usually do so if
presented with a clear opportunity, startups are founded with plans for high
potential growth from their inception. Indeed, startups typically seek a particularly
highly scalable business model [
Another important characteristic of a startup is that startups are temporary:
a startup does not want to keep being a startup. A startup will either fail
somewhere along the way or grow into a mature organization. Though it is unclear
when exactly a startup ceases to be a startup, drawing from the de nition of
Blank [
For the purpose of this study, we consider startups to be temporary organizations seeking a highly scalable business model. Software startups, on the other hand, we consider startups that deliver value primarily through software. For instance, though Uber is a taxi company, it nonetheless delivers its value to its customer through the software used to access the service; after all, it does not actually own a single taxi.
Software startups are typically associated with success stories such as that of
the aforementioned Uber. However, the majority of software startups fail [
As software startups have become more numerous and increasingly
impactful at an international economic level [
In terms of business, whereas founding a conventional company would see one write a detailed business plan for investors and perhaps take out a loan cover material costs as well, startups prefer one-page-long business model canvases over business plans and are far more focused on acquiring outside funding through short public talks referred to as pitches. Though startups are not completely unlike conventional small businesses at their core, startup entrepreneurship has grown into a sub-culture with its own community and jargon. Startup events across the globe (for instance, Slush 4) attract famous practitioner speakers and large, successful startups are motivational success stories for up-and-coming startup practitioners. Startup incubator organizations and various startup-related societies support startups during various stages of their lifecycles. As a result, startup companies use constructs that di er from conventional business vocabulary and have their own practices, for instance, in terms of searching for investments.
More speci cally in relation to software startups, software startups have been
shown to develop software di erently from SMEs and large corporations [
Just as organizations such as startup incubators and various startup event organizations have sprouted to support the high number of emerging software startups, some scholars have also begun to devise and carry out startup-related university courses. Whereas business and entrepreneurial education in general has a long-standing history in the academia, startup and software startup education as its subset is still in its infancy. In the following sub-section, we discuss the current state of practice on software startup-related education based on literature. 2.2
Three of the authors of this papers have worked on a systematic mapping
review on software startup education [
The authors broke down the research into two research questions. The rst one was related to tools, models, methodologies, and frameworks used in a software startup education context, whereas the second research question focused on best practices and lessons learned.
4 https://www.slush.org
Regarding the rst question, the authors found that the main components
used to teach software startups are:
{ Business Model Canvas - the Business Model Canvas [
In addition, some studies brought interesting insights for those who involved
in software startup courses. For instance, Genova and Gonzalez [
In another study, Bu ardi et al. [
The conclusion regarding this rst research question is that there are several di erent approaches being used. Since courses have di erent goals, objetives, and resources, each one ended up having a di erent focus. For instance, some courses just aim at inspiring students to further pursue an entrepreneurial career, while others focus more on technical aspects of a software startup.
In regards to the second research questions, we can break down the learnings into four categories: { Teaching: The journey is more important that the endpoint. In this sense, lecturers should assess students' progress. Exams are not a good option since concepts are easy in theory, but very hard to be applied in practice; { Real Projects: Whenever possible, courses should be connected to the market. When students work with real projects, their engagement and excitement rises. However, it is not always possible to do so. If that is the case, faculty should provide means for students to at least mimic a real world scenario; { Multidiscipline: Coordinating and combing courses from di erent colleges (in this cases, technology and business) is always challenging. However, good experiences have been reported. Students learn more when dealing with peers with di erent backgrounds; { Environment: The course should not be limited to the classroom. Connecting with the university ecosystems (such as technology parks) and even with other stakeholders always enriches the learning experience.
To sum up, several initiatives have been put in place in order to address software startup education. However, since it is a new topic not only for the academic community, but also to the industry, there is a lot or room for further research and development. 3
In order to study the current state of practice of software startup education in universities, we devised a qualitative, largely open-ended survey. The goal of the survey was to understand in detail how software startups are currently taught in universities world-wide. In creating the survey, papers discussing software startup courses in universities, alongside our own teaching experiences in the same area, were used to ensure that the questions covered all aspects of such courses, ranging from duration to group size where applicable. Though some questions were given multiple choice answer options, most of the survey consisted of open-ended questions. Open-ended questions were utilized to gather data as rich as possible with a survey while still consuming less resources from the responder than a qualitative interview would have. Similarly, a survey was selected as the method of data collection over interviews due to the nature of the phenomenon being studied. Though interviews would no doubt have achieved the same goal, we considered the resource-intensiveness of interviews to be a problem when interviewing other scholars. Furthermore, university education as an area of study and course-based university teaching is a well-understood phenomenon that can arguably be comprehensively covered with pre-determined questions.
The survey contained questions about both the course and the teacher(s). Aside from the way software startups are being taught, we were also interested in understanding which disciplines were concerned with them the most. In addition to focusing on teaching methods, the questions also covered the basic course information: course length, course name, which discipline the course is a part of, whether the course is mandatory or optional and other such generic university course information. Aside from asking how the course is held, we also aimed to nd out some of the reasoning behind the choices by asking some whybased questions. The survey in its entirety can be found in the following link: https://bepidpoa.typeform.com/to/kuh8bK.
The survey was sent out to individuals involved in teaching software startups in universities. Aside from contacting such individuals we knew beforehand, we searched for such courses online and contacted the teachers. However, this survey is still on-going, and we are in search of more responses from those involved with teaching software startups, as we will discuss further in the following sections. 4
Though the survey is still on-going, and we wish to gather more responses before
presenting further analysis on the subject, in this section we present preliminary
results based on the 10 responses gathered thus far. Perhaps due to the nature
of software startups, all of the responses so far have described courses either
involving a high degree of practical work or focusing entirely on practical project
work on a hypothetical or real software startup. As software startups operate
under a lack of resources, have little to no operating history, and typically consist
of inexperienced (developers or otherwise) individuals [
In relation to our second research question (RQ2), software startups are software companies operating in terms of academic disciplines, in an area combining business and information technology. This was also re ected in the responses. Eight of the ten courses were open to either a combination of IT and business students, or all students regardless of their major. Furthermore, all of the courses described in the surveys involved team-based work between students, and largely encouraged multidisciplinary teams including both business and IT students, as well as others if applicable. Student team sizes in the courses were varied but the common consensus was that at least three students would ideally be in a team as \2 is not a team, it is a pair", as one of the responses remarked. Conversely, ve students per team was generally considered to be a soft upper limit, with multiple responses arguing that more than ve students in a team would be likely to create problems in work distribution among the team.
Whereas all of the courses involved practical work, the nature of it was varied between responses. Some courses were more focused on software engineering with a secondary focus on the entrepreneurship aspect, whereas other courses were more focused on the entrepreneurship and innovation aspect with a secondary, if any, focus on practical software engineering. In two cases, the student teams would work on external commissions from real-world customers, although the trend seemed to be that the students were expected to develop their own ideas. These ideas, then, were worked on during the courses, and while they were never required to become real software startups, the students were typically encouraged to do so. In some cases, the students had indeed gone on to create successful real-world startups based on their ideas from the courses.
A clear line between a mock-up startup and a real startup in the courses described in the responses was not generally drawn. Even though the startups were not all intended to be real-world startups, or to become ones at a later point in time, all teams were expected to validate their ideas in some way, verifying that they would satisfy a real need. This typically meant carrying out surveys and interviewing potential customers, or even creating actual landing pages and social media pro les for the course startup. This was also the approach used for other work on the startups: for instance, in one of the courses everyone would pitch to real investors at a course end event, even if they had no plans of actually continuing to work on the idea after the course. In this fashion, teams that wanted to create a real startup based on the idea were free to do so without needing to take any actual steps, and the ones that were there purely for educational purposes nonetheless created a mock-up startup as if they had been working on a real one. Only one of the courses was described to be purely educational.
Past these similarities, however, the way the courses were carried out on the level of smaller details was highly varied. For example, in terms of deliverables or gradable tasks, some courses would require the students develop a working piece of software whereas other courses would focus more on honing the idea and then pitching the idea as the nal result of the course. In the cases where software development was to be carried out, agile methods, mostly ScrumBut, were typically followed, but on the other hand programming language and platform were typically not pre-determined. Seeing as the idea being carried out largely determines how it could (or should) be done, this is understandable unless the course is more focused on teaching, for instance, mobile application programming for Android while simultaneously teaching startup entrepreneurship. The way the students were supervised during the course also highly depended on the required deliverables of the course.
Though the courses focused on practical work, they featured weekly or
otherwise regular lectures. Aside from teaching relevant theories such as the Lean
Startup Methodology [
Based on the number of responses so far, we have outlined some of the general trends in the way software startups are taught in universities. However, this research is still on-going, and based on our current set of data we are as of yet unable to provide conclusive answers to our research questions. The more general trends in the way software startups are taught can already be seen in the data in order to provide a tentative answer to our, but our sample size is still too low to go into further detail. 5
To summarize our ndings, regarding que rst research questions - Aside from conventional lecture-based courses ending in an exam, how are software startups taught in universities? - the courses focus on carrying out practical work, either in the form of software engineering, creating a startup idea and developing it further, or both. The courses generally involve creating a mock-up startup in student teams and, at minimum, coming up with an idea and developing it into a business plan. No clear line is usually drawn between mock-up startups and real startups in that the student teams are expected to carry out the same tasks regardless of their own goal with their course startup or startup idea. The courses often encourage students to create a real startup with their idea but do not require them to do so.
In regard to the second research question - How do these courses deal with the multidisciplinary nature of a software startup (business, technology, design)? some of the courses focus primarily on one aspect of software startups such as software engineering and practical programming. These courses are typically only open to students of that subject such as software engineering. However, most courses seem to either involve students from di erent disciplines, typically from business and IT ones, in order to create multidisciplinary teams. Such multidisciplinary courses seem to be more common than those focused aimed at only students of software engineering, for instance.
Another important point is that startup-related concepts are seen as an integral part of entrepreneurship by now. Notably, the courses were not necessarily referred to as startup courses of any kind. In fact, only three out of the ten responses so far had the construct startup in the course title. The course titles were more associated with innovation, entrepreneurship, and software engineering practice.
Finally, it is very likely that courses described in academic papers present non-conventional educational ideas rather than tried-and-true methods for teaching. In our opinion, there is no reason to write a paper about a lecture-based course on software startups that ends in an exam about a book and the course content. Thus, in contacting the authors of various papers in relation to our survey, the data has become biased in this fashion. It is unlikely that all or even most courses on software startups would be so focused on practice, even though it would appear that the amount of practice-focused courses in the area could be higher than usual.