The Scientific Method (applied to entrepreneurship)

This is the Chapter 10 of my e-book Silicon Valley for Foreigners, that can be downloaded for free on www.siliconvalleybook.com or purchased for $2.99 on the iBookStore and Kindle. A new chapter will be posted on this blog every week.

========================================

The modern scientific method was responsible for groundbreaking discoveries and inventions in the last 400 years such as electricity, Penicillin, the camera, the combustion engine, the plane, nuclear energy, the microprocessor, and the Internet.

The scientific method consists of systematic observation, measurement, and experiment, and the formulation, testing, and modification of hypotheses. The good thing about science is that it works whether or not people believe in it.

In business, the scientific method has been used over time to create new technologies but has rarely been directly applied to product development. The reason is simple. When you are selling atoms, the scientific method approach generally takes too long, is costly and thus is incompatible with business plans and quarterly results.

Nonetheless, as I explained in the last chapter, the exponential growth of Internet users changed the way in which startups develop digital products. Long gone are the business plans, forecasts, and market research pre-launch. The always on nature of mobile devices allowed and forced entrepreneurs to devise a new methodology, based on the scientific method, when building their software products.

This methodology consists of releasing an incomplete version of a product as soon as it is ready, instead of waiting months or years to release a “final” build. The rationale is simple. With the advent of the Internet, digital products became live entities that are constantly being improved and iterated.

Thus, the trial and error methodology is now used by all kinds of companies ranging from tiny game developers to tech giants such as Uber or Facebook. It allows companies to better understand their customers, calibrate the messaging, look for signs of traction, and test the scalability of the backend systems before reaching a broader market.

If you follow the story of the most successful startups, the patterns leading to world domination looks eerily similar. They launch a bare-bones version of their product for a small audience (early adopters), collect feedback, experiment, conduct A/B testing, analyze the data, iterate dozens or hundreds of times and add features little by little in a quick but controlled process based on real metrics and feedback coming from users.

Facebook, for instance, initially released their site only with Harvard students, then opened it for students from all universities and finally, two years later, Facebook was available to everyone else.

Uber followed a similar pattern. The service was tested first with the founders’ friends. A few months later, the app was launched in San Francisco followed by other large cities in the United States. Three years down the road, the startup raised enough money to expand internationally and today is present in more than 600 cities in 80 countries.

Game developers have an even more interesting course of action. They first launch their games in a small market such as New Zealand to get feedback from users, measure the numbers so they can calibrate game mechanics and monetization loops, test the backend infrastructure, and fix critical bugs. This process, where the app goes through dozens of iterations, may take months. If the game does not reach certain metrics by then, it is killed mercilessly. If it passes, it goes live across the rest of the world.

Over the course of this decade, apps have become much less reliant on gut feeling, offline research, and creative decisions by a company committee. Now, product features are driven mostly by analysis of collected data from real customers, and the decision making is done by product managers to speed up the release of new versions. Actually, many startups and sophisticated tech companies update their apps every single week.

After 400 years helping the humanity to thrive, a variation of the good old scientific method is now being used by entrepreneurs and companies to achieve success in the digital world. The Internet not only created new distribution channels and changed the way we consume content, but it also made this trial an error approach an irreversible trend. Entrepreneurs outside Silicon Valley should take notice.