Member-only story
The 10 principles of product discovery

1: Get the f*uck out of the building
Most product discoveries start with an idea. Ideally, it’s an idea about a problem worth solving, but often it begins with a solution in mind. Regardless, the idea consists of a bunch of guesses. To some extent, there might be a total absence of facts.
Even though sales, marketing, customer excellence, the CEO, or someone else has fantastic arguments that building X is a massive opportunity, it is not necessarily true (probably not). Listen to them, get their input to shape the idea, and understand the constraints. But facts live outside your building, with potential future customers. So, get the f*uck outside and start interacting with them.
2. Don’t just listen to your customers
While getting out of the building and talking to customers is an excellent start, more is needed. There is a massive difference between what people say and do, and an even greater difference between what they think they will do in the future and what they end up doing. So, speaking with customers is, at best, a very light form of validation. You need other means to validate that a significant group of customers share the same problem and, when appropriate, that customers want your solution and that you can build it in a way that works for your business.
3. Don’t make (useless) rigid plans
Please disregard principle number three (and all the rest, for that matter) if you can get predictable results with little or no risk. But, for most of us, that’s not the case. Discovery deals with guesswork and unknowns. A rigid and comprehensive plan does not account for uncertainty, so why waste time making one?

Instead of spending days making an inaccurate plan, adopt an (agile) approach that allows changes as one learns more. Plans are flexible, and tests are conducted to validate research, leading to rapid course corrections when needed. Keep an adaptable approach throughout the product’s entire lifetime. This implies continuously taking customer input and iterating…