Why the Lean Startup doesn't always work: two types of uncertainty
The Lean Startup approach has popularized rapid experimentation within new ventures. It breaks down your venture into risky assumptions. This allows for focused information gathering, which indeed can be very helpful.
However, one year ago, in my entrepreneurship course, I kicked off on day one with students defining their risky assumptions in a workshop. First, they had a sample case to work with, from which they seemed to grasp the concept. Then, when they applied it to their own startup, chaos ensued: most were not able to complete the exercise, risky assumptions were not formulated, it did not assist them in moving forward.
Post-its, a funky environment and ill-defined ideas. The usual recipe for a startup workshop.
It seemed to me that this deliberate decomposition in order to run experiments is not always the appropriate method to build a startup. Sure, when you want to know which of your value proposition pages works best, an A/B test might help you out: this is where Eric Ries' innovation accounting comes in place. However, as it seems, much earlier in the entrepreneurial process, defining appropriate actions linked to risky assumptions is much harder.
In practice, but also in most academic research, there seems to be a limited understanding of why this is the case. Yet, I encounter courses on other universities or incubators that kick-off with risky assumptions. In this article, I will show why it doesn't always work. I will introduce two types of uncertainty through cutting-edge academic research which I will sustain with a case study I observed in my entrepreneurship course.
Entrepreneurship research and its gap from practice
Although scientific literature has answers, I want to acknowledge the following: Most entrepreneurs don't care about scientific theory on entrepreneurship. Rightfully so, as there is a huge gap between most theory and practice at the moment.
While dating back to at least 1755, entrepreneurship research really got a boom in the 1980s and 1990s. Influenced by how economics was studied back then, it resulted in what most entrepreneurs perceive as dry abstract frameworks. Those frameworks don't offer much help when devising actions, most entrepreneurs indicate to me that podcasts or Medium posts by experts in the field are much more valuable to them.
"In theory, theory and practice are the same. In practice, they are not" - Jogi Berra.
Currently, a stream of scholars recognized this gap between theory and practice. This new stream is pushing towards practice, some even going as far as stating all those models and frameworks should not be the focus, but only the observable practices of entrepreneurs. This view on entrepreneurship stems from the social sciences: what can we observe that entrepreneurs do in order to be successful?
Lean Startup is not always the right answer
In order to answer that question, we need to define success for an entrepreneur. 'Normal' people like to call this product-market fit, after which scale mechanics (such as marketing investments & growth hacking) can be implemented to more predictably grow towards a unicorn. Scholars like to define success as when an opportunity is being exploited. Yeah, that doesn't sound all too sexy nor useful. However, the construct of opportunity is prevalent in literature and I believe it's relevant for everyone entrepreneuring, since it seems to be the end goal at the end of the entrepreneurial journey.
An opportunity is a looming potential to generate and capture value; it is a venture vision of what could be in the future, it's something the entrepreneur perceives and acts upon. It's not a well-defined construct, there is not one consensus—in 2015, scholar Per Davidsson compared over 200 definitions of it, a tedious task. We don't need to define it, luckily.
Two types of uncertainty you should know
If we take the journey of an entrepreneur working towards an opportunity, we are mainly interested in the following question: what helps an entrepreneur move forward towards the opportunity? As appeared from my workshop in a previous semester, the Lean Startup approach is not the right mode of transportation for all phases of said journey. Why is that the case? It is because the Lean Startup best deals with one type of uncertainty.
Simply put, the Lean Startup is a way of getting answers, i.e. dealing with uncertainty. Dealing with uncertainty is something any entrepreneur does: a lot of things need to be determined. In a to be published paper (in AMR, a renown journal), Henrik Berglund together with Marouane Bousfiha and Yashar Mansoori, identified two types of uncertainty that can help us to answer the question of why The Lean Startup doesn't always work.
Knowledge-based uncertainty is tackled by The Lean Startup
Let's start with what we know: the Lean Startup. It deals with uncertainty via risky assumptions. As risky assumptions are basically hypotheses, it allows us to have specific answers to these hypotheses (yes or no). That means we have specific questions. When we are able to formulate a specific question, we know what we are looking for. Berglund calls this epistemic uncertainty. Epistemic is a fancy word for knowledge or information-based. Something is uncertain because we lack certain, specific information. Which value proposition appeals best to our customers? What should be our pricing?
Essence-based uncertainty is tackled by shaping
The second type of uncertainty Berglund introduces is ontological uncertainty. Again, a fancy philosophical word—it seems silly that researchers wonder why there is a gap between theory and practice. Nevertheless, this type of uncertainty deals with uncertainty when the questions nor answers are clear. For this type of uncertainty, formulating risky assumptions is much harder, I would say impossible. What is it that our company does? What is our value proposition? These are questions concerning the essence of what the startup will be. These are not to be answered with yes/no or the price should be €25 euros. For this type of uncertainty, the Lean Startup falls short. Designers and entrepreneurs overcome this type of uncertainty with a creative, social process. For example by talking with many stakeholders, ranging from your team to customers to suppliers. In this way, theoretically, the uncertainty of the opportunity changes and it slowly becomes more concrete. Berglund calls the process of determining the essence transformation. Yet for us, a more intuitive word is shaping. In the next section, I will show you how this works in practice.
Shaping a venture
My entrepreneurship course is my lab. Students are guided to explore and launch their venture idea(s). This happens in the course 'Build Your Startup', a 2.5 day per week course at Delft University of Technology. The main focus is: get out of the building, start doing. Our motto is that you can't find the value of your idea on the drawing board. Each semester I see about 8 journeys unfold right in front of my eyes. This particular batch, there is a team called Yung Sprouts (no relationship to the philosopher). It houses four ambitious students that aim to enter the local and sustainable food market.
Finding your sense of direction by invalidation
From the start, they had multiple ill-defined ideas of what shape their startup could take, in other words, multiple opportunities they could pursue. They started with actual sprouts, which they soon steered away from as they realized this was not big enough of an opportunity. The name for the group stuck, however.
Their second idea was to make local farmers' vegetables more accessible with a local food box. An agricultural expert they interviewed revealed this problem, hinting at a possible opportunity. By desk research, they found a lot of (local) competition in this market. On top of that, they found that margins are very thin. The idea of fulfilling a logistical role in a competitive game with razor-thin margins seemed an unattractive opportunity if it was an opportunity, to begin with. The second idea was invalidated, only taking them a week or two. Well done.
During previously mentioned desk research they compiled a list of sustainable food products they encountered, finding many more they didn't know of. On this list were many non-local sustainable products that they believed should reach more people. An idea for this was an innovative sustainable food box. This was one among many ideas at this point in the startup, and they did not really have a clear sense of direction.
Problem owner makes an opportunity concrete
In order to get a sense of direction, they followed the mantra of 'get out of the building'. They talked to producers of sustainable products, continuing with shaping activities. All team members point out that one talk with a producer made things much more concrete. Let's call this company Shroomville, a firm that grows mushrooms out of coffee grounds.
As a researcher and educator, I was interested in how this helped them make their venture idea more concrete. We attempted to reconstruct this process step by step, I interviewed each team member for 10 - 20 minutes. Below is a reconstruction of a key moment in the shaping process of the startup.
One day, a Yung Sprout walked in Shroomville local facility, and just started chatting with the founder about their ambition in the sustainable food market. Then something magical happened. Enthusiastically engaged in the conversation, Shroomville told them there are many producers like them, but it's hard to get to the market. She mentioned that reaching the end-user, also in the supermarket, and bridging the last mile are challenges. They found a problem owner. The team members said this is a problem they thought of before during a brainstorm, but now it got substance. They were able to link it with the sustainable food box.
One of the Yung Sprouts holds hands wide in front of his webcam, signifying a broad area. He says: "At the start of that week this was our problem space. There were all these points where we could do something." He moves his hands close together in front of his webcam, signifying only a portion of the broad space.
"When we talked with Shroomville, she told us [..]: I've got very concrete input [a problem] for you to work with. This helped us with making the idea more concrete. We've considered this spot [problem], but she owned this problem, and could much more precisely tell us what her barriers were"
Before this meeting, they had this problem as a reasoned problem, 'among 80 others' he added and therefore didn't add much value to it. By talking to this problem owner, the perceived potential value grew immensely. It underpinned their sustainable food box idea. They found a problem for their solution. The contours of an opportunity started to grow, however intangible this is, the first-time entrepreneurs indicated in the interviews they sensed this potential. They wanted to keep their velocity up and settled on this idea for now.
Shaping the opportunity with stakeholders
What we see here, is a phase in which the team is making the opportunity more concrete through a social, creative process. The team is talking with potential stakeholders about the realm they operate in. They don't have distinct questions yet. The only thing they aimed to answer is: "What is it what our startup should do?" The uncertainty is predominantly ontological, meaning they are figuring out their essence. Then, when talking to Shroomville, the problem realm shrunk immensely. Something 'clicked'. They sensed an opportunity. Their idea got more concrete by the first-hand account of the problem owner. By means of shaping, they define the essence of the opportunity, decreasing the ontological uncertainty. They all of a sudden have a much more concrete venture idea.
When to start experimentation
Concrete enough to experiment
The next Tuesday the team arrives in my coaching session. Each team gets 30 minutes of coaching per week, aiming to assist their journey towards exploiting an opportunity. At that point, I was not aware of what is reconstructed above. They opened with telling: "We've settled on a direction." They pitched the sustainable food box idea. It felt concrete to me.
I didn't question their fixation on this idea. When teams struggle with directions at the start of the course, I'm always happy that they find a focus. I let them find out whether it's a good focus, furthermore, who am I to tell?
Intuitively I felt they should start doing this idea. Not just drawing it, but executing it, in a pilot experiment. That's what I suggested: "Try selling 10 of these boxes and learn."
The mockup they used to get their first customers via LinkedIn. "Sink your teeth into real good food"
In next week's coaching session, they inform my colleague (and ex-entrepreneur) that they are doing this pilot. And that something funny happened. They told him that while the decision to go for this pilot made their venture idea more concrete, uncertainty went up, too.
This seemed paradoxically to them and to me at first. They said that all of a sudden, a lot of things needed to be defined. One member says:
"If you are forced to make decisions, you find out what you don't know"
They needed to define, among other things:
All these questions popped up. It dawned on me this was a different type of uncertainty, it was knowledge-based uncertainty. I wanted to know how they experienced this new type of uncertainty that arose. I was specifically interested in how it compared to the uncertainty that evaporated once they settled on a venture idea. This question was quite easy to answer, without hesitation two members said:
"These things are so much more concrete and practical." - Member 1
"Before, everything was much more conceptual." - Member 2
It appears that making a decision will reduce conceptual uncertainty and will increase uncertainty about the details. You shift from essence based uncertainty to knowledge-based uncertainty. One member said:
"It's not that things get less concrete. It's only that you realize that some elements are relatively less concrete as certain elements advance" - Member 3
They saw their venture slowly take shape. As I finished the third interview on Thursday, they already sold 1 box for €25 euros to their first customer, one member proudly told me over Zoom. When I finished the fourth and last interview four days later, team member four said full of enthusiasm they've sold all 10 and have 4 customers on a waiting list for a second pilot. That is some early success.
Choosing appropriate actions
At this point, I worried about my interference in the process. Would they've done the pilot without my suggestion? And is that a bad thing? I checked this in interviews, asking if they considered the pilot before I mentioned it. The answers are a bit ambiguous, some members said yes, some said yes but not as clearly. But they were unanimous on that the coaching session accelerated everything and made them prioritize the pilot. I asked them why they thought it made sense to do a pilot. One member said:
"We've been speculating a lot about our customer, but we need to validate this. The pilot seemed a good way of doing this."
And here I had a sigh of relief and a lightbulb moment. This is not the first time I encounter their situation. However, these students are first-time entrepreneurs. They didn't have the experience yet to realize that their idea was concrete enough to do an experiment around. I have influenced them, nudged them in that direction. They might have realized to do this pilot themselves later, a team member points out; we will never know. But for the development of their startup, they are moving forward towards learning whether this indeed is an interesting avenue for them. And that is a win.
Shaping precedes experimentation
As it seems in the case of Yung Sprouts, experimentation did only happen after sufficient shaping has taken place. Recognizing sufficient shaping has taken place required at least a more experienced look from the outside. When to start experimenting? Does experimentation always follow shaping?
If we relate this back to the workshop from last semester, we might be able to get more data to this. We can also answer the question of why risky assumptions resulted in chaos. It was premature experimentation. At the start of the semester, most startups had no clue what the essence of their startup is. There was only one startup for whom the assignment worked. I learned that they've worked on their startup already over the summer.
Shaping seems to take place in the fuzzy-front end of entrepreneurship and is crucial to understand how new ventures emerge. This does not imply that after experimentation, shaping can't occur again. As Berglund suggests, we might find an oscillation between the two modes if we have longitudinal data for that. The hunch that I have is that at the start of any entrepreneurial journey, shaping takes place before experimentation.
The future ahead
For entrepreneurs, coaches and educators
There is a lot to unpack here, and this article is just the beginning. Experimentation, as defined by Berglund, happens but in this case only after shaping. It is very crucial to recognize where a startup is in the process. Premature experimentation can lead to being stuck as it gives a false sense of concreteness to something that is ill-defined. Recognize where you are in your journey, and pick appropriate actions. Read more on methods that help with shaping below.
For researchers
In his paper, Berglund suggests a potential dynamic between shaping (transformation) and experimentation, this could be the first sign of that dynamic. There is room for deepening a lot of the constructs above. Also, the characteristics of 'sufficient shaping' are intriguing to me. Furthermore, the role of artifacts in the experimentation such as the LinkedIn flyer and their MVP-box or the form of artifacts in the shaping stage are topics that I'm eager to explore further. The fuzzy front-end of entrepreneurship potentially can draw a lot from design theory and innovation research. In any way, the basic dynamics of the two types of uncertainty are very interesting for the future of my research and research on this topic.
Do you want a product-market fit goodie bag and more insights? Subscribe to my newsletter to receive the latest insights in your mailbox and your welcome gift.
Edit: First boxes delivered
On Friday, the course group WhatsApp got an exciting update: The first boxes are being delivered!
Box added for privacy reasons.
Academic papers used
Berglund, H., Bousfiha, M., Mansoori, Y., 2020. Opportunities as Artifacts and Entrepreneurship as Design. AMR amr.2018.0285. https://doi.org/10.5465/amr.2018.0285
Carlsson, B., Braunerhjelm, P., McKelvey, M., Olofsson, C., Persson, L., Ylinenp??, H., 2013. The evolving domain of entrepreneurship research. Small Bus Econ 41, 913–930. https://doi.org/10.1007/s11187-013-9503-y
Davidsson, P., 2015. Entrepreneurial opportunities and the entrepreneurship nexus: A re-conceptualization. Journal of Business Venturing 30, 674–695. https://doi.org/10.1016/j.jbusvent.2015.01.002
Startup Growth & Low-Code ??
1 年Wout Withagen deze! ??
Founder & CEO at Reframe Technologies, Inc.
3 年Jeroen Coelen - you raise a good point here and there are certainly limits to the lean startup process. To add some vocabulary to the discussion, I think it would be more precise to say that the lean startup techniques you refer to are based on the assumption of an already existing *purposeful* market attack strategy where clear hypothesis are attempting to be validated. However, if you are in an *emergent* strategy situation where there is not yet a strong market focused hypothesis, then we have to undertake an investigation to understand the 'Jobs to Be Done" and drive forward a process of collecting marketing information, progress through structured derivation of key *quality attributes* of doing the job well. From this framework, we can then put forth a clearer hypothesis and start to shift from *emergent* strategy to a *purposeful* one. I highly recommend Clayton Christensen's work and in particular his book 'Competing Against Luck' or the earlier work of Lanning and Phillips 'Market Focused Organization' papers that characterize how to systematically build the ground work that sets up the classical Lean Startup approach for success. Happy to discuss more anytime. Best, Jeff.
Design Scientist in all things Entrepreneurial
3 年Graham FellowsAngela YoungMarvin FoxAaron BirkbyChad RenandoTimothy HuiMarc OrchardSharon HunneybellDr Paul MathiesenVu Tran
Design Scientist in all things Entrepreneurial
3 年Jeroen Coelen great insights. Definitely useful not only for entrepreneurship scholars to rethink how we approach our research but also for the many incubators and startup ecosystems to improve the way they use and teach the 'Lean Canvas' in their boot camps and workshops.
Juridisch, bedrijfskundig & strategisch adviseur | Intervisie begeleider binnen de advocatuur | HBO docent Fontys HRM
4 年Interessant! ??????