As an academic scientist interested in entrepreneurship, Eric Ries’s The Lean Startup was one of the first books I read to better understand and begin crossing the seemingly large divide between academia and the world of startups. What I quickly came to realize and appreciate however, were not the contrasts between the two, but rather the parallels between the methods of scientific research and product/business development.
The Lean Startup’s central thesis is product development—for startups and large corporations alike—must incorporate hypothesis-driven testing and “validated learning” to achieve an ideal product-market fit. This is driven by Ries’s Build-Measure-Learn Loop, an iterative process in which an idea or hypothesis is formed about the customer or market and a product created to test the hypothesis. Data for the new product (e.g. changes in specific user metrics) is measured and the validity of the hypothesis evaluated. This process is repeated a number of times, with the overall goal of closing in on the best possible product-market fit. Whether or not an iteration of the loop results in a correct hypothesis does not matter as much as whether validated learning has occurred: do we now better understand our customer or market, and do we stay the course or pivot?
I (and likely you) immediately noted the clear similarities between the Lean Startup approach and the scientific method—the Lean Startup approach is simply the scientific method applied to product development! In academic research, a hypothesis is formulated and an experiment conducted to test if a hypothesis is correct or not. Validated learning from the experiment is used to hone in on a product-model¬ fit, that is, to gain an understanding and develop a model of the system being studied (e.g. cells, populations, disease). With each iteration of the loop, a hypothesis is kept or discarded and the model further refined and updated.
Having my roots in academia, I assumed the divide between the academic and startup worlds was vast, but in reality much of the script is the same—it’s just the characters (and terminology) that are different. Just as VCs and other investors will call out a startup for presenting vanity metrics (user or other statistics without real meaning), reviewers of a scientific publication will demand convincing experiments and the right data to accept what a scientist is proposing. For both scientific research and startups, success depends upon asking the right questions, conducting the right experiments, and continual validated learning.
For those in academia hoping to make the leap to the startup world: do it! The divide is not only traversable, but the entrepreneurial world offers a great environment to apply your experience and knowledge to building great companies.
Douglas Deutsch is a Graduate Fellow at Rockefeller University and an InSITE Fellow. Prior to Rockefeller, he conducted natural products and medicinal chemistry research at Cornell University.