The future of the lean startup from a science perspective


Tldr: The Lean Startup is much closer to science than you think. Lean Startup can be improved through a stronger connection of learnings and a critical discussion of experiments and results with a bigger community.

The Lean Startup approach is built on the scientific method and the idea of testing an assumption with the help of an experiment. This method goes back to 1000 CE and the scientist Alhazen. It shows the systematic and rigorous approach to generate learning on a particular domain, when there are many white spots and a high level of uncertainty.

A lean entrepreneur is also confronted with a high level of uncertainty. He continuously tries to decrease the risk that comes with decision making under high uncertainty. A lot of entrepreneurs would never want to be compared with scientists. This article will do this job and looks at the commonalities/differences of Lean Startup and science.

To tackle the question of how much science is in Lean Startup, its essential to identify the essence of science. There have been many discussions what the essence of science is. Many researcher have debated, if inductive or deductive reasoning is truly scientific. Plus, there were many discussions between positivists and constructivists on the question if a hypothesis can be validated or if they can get only falsified. Paul Hoyningen-Huene a German professor in philosophy, developed the concept of systematicity. Which doesn’t try to distinguish science from pseudoscience, but the scientific process of knowledge generation vs. the everyday process of knowledge generation. Many previous attempts to define science where focused so far on distinguishing science from pseudoscience (e.g. Astronomy and Astrology).

He identified using different dimensions, how the scientific knowledge generation differs from the everyday knowledge. These dimensions include the following.

  • Descriptions – science provides a very systematic description of phenomena using e.g. classifications, typologies and taxonomies.
  • Explanations – sciences develop theories and their explaining and predicting power.
  • Predictions - various sciences use predictions that are generated through statistical methods e.g. correlations and regression analysis
  • Generation of new knowledge – Scientists systematically search for new or improved empirical data. They systematically try to eliminate chance through e.g. ‘brute force’ approaches or use explorative and creative methods, so that science becomes and autocatalytic process.
  • Defence of knowledge claims – sciences systematically try to eliminate the risk of wrong statements.
  • Ideal of completeness – Scientists are trying to get to a compelling knowledge base instead of fragmented knowledge on a specific domain.
  • Critical disclosure – Scientists actively engage in the critical discussions. They attend conferences, present their results and are also eager to use platforms like researchgate to discuss their progress.
  • Epistemic connectedness – Researchers actively try to connect knowledge. This is particularly evident when looking at footnotes. Scientists use footnotes to refer to other sources results and thoughts.

Within these domains scientific knowledge generation shows a higher systematicity than everyday knowledge generation. This means also, since the Lean Startup follows the scientific method, it should somehow consist of a higher systematicity than everyday knowledge creation. I want to highlight in what dimensions the lean startup approach currently focuses on, where it differs for good reasons from science and I want to highlight dimensions where I see potential for the Lean Startup to develop a greater systematicity to optimize the effectiveness of a Lean Entrepreneur.

Descriptions - Lean Entrepreneurs do definitely use various tools for making profound descriptions of their business theories. They outline Personas or line out the Lean Canvas or the Business Model Canvas. These descriptions are trying to comprehensively describe the business model in order to validate and optimize the different parts. Such tools and process are not used in everyday learning and knowledge accumulation and are thus very close to science.

Explanations - Something that is at heart of the Lean Startup is to validate or invalidate cause and effect relations, so that the entrepreneurs are able to explain what happend in a particular circumstance. The Lean Entrepreneur is consequently looking for explanations - Why do customers buy? Why does the company grow? Why is channel A better than channel B? Why did the value proposition not work? Furthermore, Lean Entrepreneurs systematically map out their assumptions, build hypothesis and seek to (in-)validate the hypothesis. This very systematic approach to get explanations is very close to science.

Predictions - Lean Entrepreneurs don’t just look for explanations, but they want to turn the explanations into predictions. This is really one of the basic tasks of the Lean Entrepreneur - to make better predictions on the business outcome. Entrepreneurial Management differs strongly at this point from Traditional Management. Since it’s not about delivering against predictions (this is impossible in a startup situation), but to become make better predictions. There is only rigorous learning possible, if you’re predicting, run an experiment and you’re validating or invalidating your predicted outcome.

Generation of knowledge - Scientists and Lean Entrepreneurs objective is to generate knowledge. The scientist in order to explain the world better and thus with a societal impact, the Lean Entrepreneur wants to generate knowledge in order to make better business decisions. Both scientists and lean entrepreneurs use a very structured approach in the generation of knowledge, where they actively and systematically engage in the generation of new assumptions (e.g. through observations or ideation) and the testing of these assumptions through a deductive approach (deduct assumption from theory, deduct hypothesis from assumption etc.). The experiment board is nothing else than deduction.


Defence of knowledge claims - It’s pretty clear what the Defence of knowledge claims is about in science - is the target to increase objectivity, validity, and reliability. The Lean Entrepreneur also tries to reduce the influence of biases in order to increase the objectivity. There is also often the question if one method is more valid than another to validate or invalidate the experiment. Many advanced Lean Entrepreneurs use statistical metrics - just like scientists - in order to guarantee objectivity, validity and reliability. Although the factor of cost plays a role in science as well, it is by far more important for a Lean Entrepreneur. That’s why the time and the cost will always come into play, when he/she evaluates a particular method.

Ideal of completeness - The ideal of completeness is for sure something that is not as important for Lean Entrepreneurs as it is for science. A scientist tries to discover a phenomena holistically and wants to identify the underlying cause-and-effect relations in order to create better theories of a phenomena. He strives for a completion of his picture. A Lean Entrepreneur on the other side is not looking for perfect information. He acknowledges the high level of uncertainty and tries to gain as much from his methods as necessary in order to make a decision. Information comes with diminishing returns. A scientist would try to make arguments and statements bulletproof. The Lean Entrepreneur wants to get to validated learning as soon as possible.



Representation of knowledge - Scientists are systematically trying to represent the generated knowledge through publishing it in academic papers or presenting it at conferences. The generated knowledge from Lean Entrepreneurs is most of the time represented internally in the company and should be accessible for every member of the staff.

Critical disclosure - One of the foundations of science is the open discussion of experiments results. There are peer-reviewed papers and there are specific awards for making the best progress in the particular domain. Critical disclosure is a point, where there is not so much emphasis on in the Lean Startup, although it can provide essential information. There is no really systematic approach in talking about experiments, the used methods and the results. In an open society as Karl Popper points it out, entrepreneurs shouldn’t be afraid of sharing their experiments and their results. The additional knowledge that can be generated through sharing learnings and opening the Experiments and methods for critical discussions can only result in progress. But wait, am I really suggesting, that you share your experiments with the world. Yes, I do think so. There are for sure different degrees of openness in this discussion, but I think it’s essential to put more emphasis on discussing experiments, how they were conducted and also, about the result. By doing this a Lean Entrepreneur is able to dramatically increase his/her learning. So share your experiments, your methods and your results, with others. I would also advocate for an open discussion culture, where employees are actively trying to invalidate the critical business assumptions in order to get to a sustainable competitive advantage.

Epistemic connectedness - Epistemic connectedness is not that well-developed in Lean Startup compared to science. There is a huge value in the connectedness of knowledge in science. You don’t reinvent the wheel again and again. You can actually build on the shoulders of giants. I don’t see much of the systematic connection of knowledge in the Lean Startup space. Many times in companies, you run the same experiments multiple times. I’m advocating for a stronger connectedness of the different knowledge that is generated by a Lean Entrepreneur. This requires for sure some discipline and tooling, but it will not only help to avoid making the same learnings again and again, but it will also help to evaluate disconfirming evidence that you generate. With the help of connected knowledge domains (e.g. Customer, Problem, Solution, Product, Channel, Revenue stream), a Lean Entrepreneur is able to trace the effect on a particular element of his/her business.

Comparing science and Lean Startup allows us to identify gaps in the current conception of the approach. The comparison shows that the Lean Startup is probably more scientific than we think. It is very clear though, that a scientist has a totally different goal than a lean entrepreneur. Nevertheless, being more systematic in critical disclosure and epistemological connectedness holds great values for companies that apply the Build-Measure-Learn Loop. It is especially fruitful to think about these two points, when you are not thinking about one venture only, but multiple and if you are not thinking about utilizing the Build-Measure Learn Loop for months, but when you want to build all your running activities (e.g. A/B Testing) on it.

This article has been originally published on LinkedIn on the 18.08.2017



  • Paul Hoyningen-Huene (2013), Systematicity
  • Eric Ries (2011), Lean Startup
  • Eric Ries (2015), Leaders Guide
  • Douglas Hubbard (2014) How to mesaure anything
  • Hug/Poscheschnik (2010), Empirisch Forschen
Lutz Göcke