Your Startup is NOT a Science Experiment. Or, a Scientific Endeavour.

Make evidence-based decisions. Make scientific decisions. Both are not the same. All evidence-based decisions are not scientific. Science is our search for truth where we need to follow a fixed set of rules and come up with a result. You form a hypothesis, you design a test, you test the hypothesis, you check for the significance of results and then iterate if you get new data. The new data might refute or support your scientific statement, you accept the results, whatever they may be, and move on.

This method, the method of science, the scientific method, is often cited when people make decisions or act, in their startups. And, that is a problem. What we do in startups, is not strictly scientific, we can’t establish controls, we may not check for significance, and we may not even have a fully testable hypothesis. But still, some people are hell-bent on calling what they do, scientific. I think this needs to be corrected.

I hear some of you asking, how does it matter? how does it make a difference? It does, for the ones working with you, who are independent-minded, and who do their work, and their reading before they follow your instructions blindly. Those people will go home and look up what the scientific method is, and then try to apply the method for conducting experiments in a business setting. And then they will be paralyzed because achieving scientific rigor is complicated, tough, and slow. As I said before, “You form a hypothesis, you design a test, you test the hypothesis, you check for the significance of results and then iterate if you get new data. The new data might refute or support your scientific statement, you accept the results, whatever they may be, and move on”.

What we do at startups is use pragmatism, not science. Pragmatism is similar to Science. So similar infact that you may confuse one with the other. Pragmatism, also asks you to consider the evidence, make a decision, and come to a conclusion about the truth. Pragmatism also recommends that you must be willing to change the truth in the light of new evidence, if and when it emerges.

The difference is that of drawing a line. When it comes to science, the line between truth and non-truth, science and non-science is drawn by general consensus, by what most people in the world believe. The method, the controls, and the settings of the experiment must adhere to those standards of science, or as one may call it - the scientific method. If it does not, whatever you are doing, is not scientific.

Compared to that, pragmatism states that drawing a line, that separates truth from non-truth, is subjective, and is decided by the one conducting the test, by the one making the decision. To be precise, if there is enough data for the result of an experiment to be useful, it is true. What is useful is true.

Pragmatism is an approach that works best when we don’t have the time, resources, or tools to conduct science. In the best-case scenario, we would all conduct science, and be absolutely sure about our decisions, about the results of our experiments. But in the real world, we don’t have that luxury. In the real world, there is a deadline, there is a burn rate, there is an expiry date for your business. In those situations, we have to work with what we have, and that is pragmatism.

Pragmatism also states that one must be willing to change one’s opinion of the truth or rightness of a decision, based on the new data. It states that all truths are fallible, they are temporary. But they are actionable. Once you have acted on a truth, on some data or evidence, the results will come, and it is your job to adjust your truth based on the results and act more. Again, very similar to science, but not science, pragmatism.

Here are a few examples that differentiate between science and pragmatism, which I generated with the help of GPT.

**Example 1: Launching a new product
**
Pragmatist

• Action: Launches a minimum viable product (MVP) to test the market quickly.

• Reasoning: Wants immediate feedback and is willing to iterate based on customer responses.

Scientist

• Action: Conduct extensive market research and A/B tests of different product features before launching.

• Reasoning: Seeks to minimize risk and uncertainty through empirical data.

Example 2: A/B Testing

• Pragmatist: May conduct A/B tests, but also willing to go to market sooner with less data if it solves an immediate problem.

• Scientist: Also conducts A/B tests, but typically waits for more comprehensive data before making a decision.

Example 3: Pricing Strategy

• Pragmatist: Sets a price that seems reasonable, and adjusts based on early sales data.

• Scientist: Conducts a detailed price elasticity study before setting the price.

I think is very obvious by this point that terms like lean startup, lean method, data-driven marketing, experiments in startups, etc, are based on pragmatism, not science. It doesn’t matter what people say about them being scientific, they are technically not. And for anyone dwelling in the depths of these domains, this can be confusing.

You might try to define science in a certain way, to call these approaches scientific. That is an error. You can’t make up a definition of science, that is the whole point of science, being objective, not subjective. If it is not scientific, don’t call it scientific. There is a perfectly good word for it, pragmatic.

One more thing, pragmatism entails more risk than science. A pragmatic decision has more chances of being wrong than a scientific decision. But that is the cost of action. You can be 95% or maybe even 99% risk-free by being scientific, but you may never act until you get to those percentages. Pragmatism has more variance, your action may in retrospect turn out to be scientific, but it may also turn out to be an utter disaster. The reward is that you keep moving, for better or for worse. It is very obvious from this, why you wouldn’t want to be pragmatic while building a biotech, health-tech, or nuclear startup.