Most people find the task of starting a startup to be very daunting. I was one of those people. Over my years of learning through various incubators, programs, books, and diving deep down into the ecosystem, I've come to fully understand the solution to that problem.
Drawing an analogy, it's to think like a scientist. (Actually half-analogy, because the scientific method is more methodology than analogy, but it's an analogy all the same)
A common pitfall lies in thinking like a philosopher. Very often, whenever faced with a convicting startup idea and the personal attachments that come with it, one starts to over-plan and over-hypothesise. This is an all too common trap that amateur founders fall into. They debate over every side of the issue, start writing comprehensive business plans and plan for all possible scenarios. This is methodology designed for questioning the fabric of reality itself, not for starting businesses.
Don't think like a philosopher, think like a scientist. When a scientist what comes up against a daunting problem, what does he do?
Reduce it to that which can be experimented upon
How does a scientist face the daunting task of decoding human consciousness? Of discovering our origins? Of understanding the material universe?
When faced with a daunting problem, the scientist first reduces the polarising problem to something he can comprehend fully with his knowledge and resources at hand. This is the concept of scientific reductionism.
In a startup world, the word "experimentation" is less used. Rather, it's usually words like "MVP", "prototyping" and "iteration". But they are basically equivalents.
Your first thought whenever faced with a daunting problem, should be to reduce your startup idea to its minimum viable form → minimum viable product (MVP) or [prototype](https://en.wikipedia.org/wiki/Prototype#:~:text=A prototype is an early,%2C electronics%2C and software programming.) in startup/software engineering lingo.
If your idea is to build X, start a minimum viable product, testing the core concept, and see if your hypothesis has any merit. Here are 10 great examples of MVPs (see how startups like Uber, Amazon and Facebook started).
Perform the experiment
The scientist assigns specific, core assumptions to be tested → called a [scientific hypothesis](https://www.livescience.com/21490-what-is-a-scientific-hypothesis-definition-of-hypothesis.html#:~:text=A scientific hypothesis is the,the definition can be expanded.). He gathers his resources needed to perform the tests and jots down notes to record down his findings. He details his observations as objectively as possible, and analyses them to see if his hypothesis is validated.
Onto his startup equivalent, he gathers his resources to build his minimum viable product and assigns his specific, core assumptions to be tested → also called a hypothesis. At this stage, it would be irrational to start thinking about investing a large sum of money on building it. Rather, he should focus on testing to see if his core assumptions are accurate, resourcefully. Here are 18 types of cost-friendly MVPs to test your hypothesis with.
In chemistry, the scientist tests to see if particles react according to his hypothesis. In startups, the founder tests to see if people react according to his hypothesis. He keeps his assumptions in mind and analyses the outcomes objectively to see if he achieved his intended reaction (pun intended). In order to test reactions, the founder needs to get out there and meet where buyers are. Whether they be online conferences, Reddit, Twitter, or email. The founder, particularly the CEO, must spearhead this approach.
Often, the scientist discovers new problems when testing his hypothesis. Interestingly, while this prospect excites scientists, this usually seems to frustrate founders - who get increasingly perplexed by the complexity of their endeavours.
Constantly modify the hypothesis
Once he has performed his test, he will be able to tell if the observed reactions were consistent with his hypothesis or not. Read in detail on the MVP process.
According to his findings, he would either continue with his experiment, adjust his hypothesis, or scrap the experiment entirely. Scientists know full well that it's very normal to be wrong. In the same manner, founders should be quick to evaluate their projects too and scrap them if it seems to be futile. Lest they waste their time, energy and resources trying to prove a futile hypothesis, and eventually end up creating a solution in search of a problem. Founders need to be wary of becoming too attached to one idea.
Most experiments unsurprisingly turn out wrong. After all, that is the nature of experimentation itself. Which is why it is of absolutely no surprise that 90% of all startups fail. It's like saying 9 out of 10 experiments fail, which suddenly feels a bit less shocking or interesting to say.
But for every 90%, there is a 10%. The measure of a successful experiment for a founder is when his hypothesis is validated. In other words, he finds that the problem he identified is real and his solution fits → problem/solution fit, in startup lingo.
Upon finding problem/solution fit, his new goal is to get his solution adopted by his target customers → product/market fit, in startup lingo. This is a whole new world, with countless resources, talks and books on this one topic. Achieving product/market fit is a startup founder's utopia, and this is where most startups either stay alive or go to die (34% of all startups fail due to lack of product/market fit).
Limitations of applying the scientific method to starting a startup
I wish I could go further into how to start startups the right way, but I am not qualified to do so. So here are some recommended readings from those who are.