Innovation processes usually start with brainstorming sessions, whose participants are expected to come up with the one brilliant idea (or preferably several) that will change the company forever. However, big ideas are neither created at the push of a button nor in uninspired conference rooms.

Every idea, no matter how great, has one thing in common at the beginning: It is incomplete, unpolished and as such indistinguishable from bad ideas. So let’s take the pressure off and call them what they are at best: hunches of a problem.

It started with business plans, today they are called Business Model Canvas. Those concepts suggest that the framework of an innovation can be defined even before the first step is taken. If your product, however, is nothing more than a vague hunch, how can you dare to know your customer.

Focusing on the user means (re)building an existing product. The truth is that true innovation involves risk and uncertainty. The only way to counter this is to take it one step at a time. The rest – including the user – will fall into place.

Most innovation processes recommend launching products as Minimum Viable Versions with just enough features, and then developing them further based on user feedback. This implicitly assumes that the user has both the knowledge and the capability to develop the product. He doesn’t.

MVPs involve confirmation biases. Instead of wanting to have one’s own hypothesis confirmed, assertions should be verified not validated—and in the best case proven wrong. The learning effect is all the greater engaging in trial and error. Investigating what went wrong, can be fixed.