Any scientist worth their salt knows that when preparing to study a concept/phenomena you have to use the good old scientific method. What’s that you say? Essentially, it involves using empirical (e.g. observable/measurable) data and a bit of reasoning to ensure that what you think you know, is real. Oxford’s English dictionary put’s it another way; they call it “a method or procedure that has characterized natural science since the 17th century, consisting in systematic observation, measurement, and experiment, and the formulation, testing, and modification of hypotheses.”
The last word there will be the focus of this post. Eric Ries’ Lean Startup Methodology is firmly based on the scientific method and it’s critical that startups understand the basics of hypotheses. If you look at a process flow for the Lean Startup (based on the chart in Reis’ HBR case), you’ll see that the decisions are based on data gleaned from experiments. It’s impossible to learn from the experiments if you don’t establish hypotheses you can actually test. What does that mean? It means you need to set affirmative predictions that can be measured and proven false.
To remain objective, you should use quantitative data whenever possible so you can ensure that your enthusiasm doesn’t color your perception of the results. Your working hypothesis must be capable of being proven false. My good buddy Karl Popper espoused the concept of falsificationism and demarcating that which is science and that which is non-science. If you don’t set the right hypotheses, you’ll end up just “seeing what happens” and more likely than not, you’ll succeed in seeing what you hoped to see. We don’t want to be considered non-scientific, so we better establish falsifiable hypotheses.
When considering your hypotheses, ensure that you’re testing those that are most central to the business model. It’s unlikely that you’ll have just one hypothesis as a business model is made of up numerous components (e.g. value proposition, profit formula). It may be necessary to study a flow of hypotheses if there is serial dependence of business model elements. Luckily, in the iterative world of the Lean Startup, you can expect to test and revise hypotheses over time. Keep in mind that you may not test everything at once, but you should test the most critical elements first and plan the downstream topics into future tests.