Use better data

Founding a company is kind of a crazy thing to do. There are rare cases where weekend projects spin out into massive opportunities, but most companies are built with stress and perseverance. Oh, and a heap of faith.

That's the dirty secret that we shy away from. If the decisions that create great companies were wholly data-driven then they wouldn't hold any risk. You would only ever be a few data points away from the next Uber . Yet the industry is so all-in on data right now that we're completely ignoring all the other traits that form great products.

Many companies start because of problems that the founders themselves have experienced. They have a determination to solve those problems, and often a clear vision for where they want the company to go. That trajectory is what creates a leading product, and pushes competition aside with sheer momentum. These elements, when nurtured by a founder that still remembers the frustration that caused them to start the company, create something special. This is one of the reasons companies that are led by their original founders do better than professionally managed companies.

I'm not saying that founders create products with zero data, but a heap of founder stories start with "I experienced problem X so I created solution Y". In fact, I would wager most MVPs are made with the founders being the sole data points.

Obviously, good product decisions use data. What I'm saying is we should use better data. Look at usage metrics on the website, but also sit down and watch someone use it. Talk to a bunch of customers, but also work alongside them for a day. Get passionate about the problems your solving and the market you're operating in. Product Management may not be an art, but it definitely isn't a science either. So sit between those two worlds.

There are subtle things that don't have data points. Industry changes that are felt before they can be measured. Sentiment on products or solutions that shift. Sometimes markets change abruptly and waiting for the right data points will mean reacting too slow. Companies that have only practiced making product decisions with neat data sets won't be prepared for how to make decisions in a rapidly changing environment.

Data is a fantastic de-risking tool. Presenting product decisions as a set of logical steps is the easiest way to get buy-in from stakeholders. Making a feature or a product an "easy sell" shouldn't be the sole driver though. That's how you get a great list of features with no heart.

Back your product decisions with data, but don't refuse to build anything without it.