One of my top philosophies in product management is that everything is a hypothesis. If your boss has an “idea”, you should consider that idea as being simply a hypothesis that needs validation.

One of my favorite tools in dealing with hypotheses is to put everything in a hypothesis bank.

The hypothesis bank

The hypothesis bank is a living document where everyone in the company is encouraged to get creative and come up with hypotheses to be validated. 

This amazing tool works on two levels. First, it could actually generate brilliant hypotheses. And you could be sitting on a gold mine without even knowing it. Your team (especially the employees front lining users) know best some of the biggest problems your customers are facing. It is only natural that they should be given the chance to imagine how to find proper solutions. 

And second, it involves people into the roadmap creation and decision-making process and gives them a voice. Ask any human resources person and she will tell you that an employee that has her voice heard is more likely to be a happy and productive one.

A hypothesis bank could indeed help you tap into the full potential of your organization and talent. 

When written down, a hypothesis should convey the following information:

• Hypothesis
• Problem 
• How do we know the problem is real
• Objective

For Ring.md, our global telemedicine platform, our NSM was “Daily Consultation”. We wanted to increase this metric by 20% in the following 3 months.

We encouraged everyone to contribute with hypotheses on how to achieve that. One of it (coming from a developer) looked like this:

“New patients have doubts about the efficiency of telemedicine. They prefer the brick and border setup (study attached). If we charge patients, not in advance but only after the consultation finishes it will add additional trust to the service and increase sign-ups.”

This hypothesis described the problem, the objective and provided evidence supporting that the problem was real. And it actually proved to be a winner when tested.

The hypotheses will usually come from multiple sources (brainstorming, competitor analysis, feature requests, revelations, management, etc) and all are fine as long as the problem and the desired goal are clearly defined.