I’m a digital product manager with a problem: I don't know what to build. This is mostly about features:
- What does the minimally viable product look like?
- What features are my customers likely to use?
- What does my roadmap look like?
- Do some customers want one thing and others want something else, and how do I make them all happy?
- Do my stakeholders want the same features as my customers?
That last problem is the toughest to solve. Most product managers who want to stay employed listen to the boss (and top stakeholders). Unfortunately, our best business ideas can be out of step with customer needs. As the UX folks say, "We are not our customers," and convincing everyone what the customer's pains and gains are is a challenge.
Maybe this happens to you, too. It's pretty common.
There are solutions. Surveys are great and have their place, but they're not perfect. Closed-ended surveys are easy to analyze, but can leave up to 60% of possible answers out of the results. Open-ended surveys can capture much more data but are a nightmare to process and are best used for qualitative studies. Even so, it takes a solid analytics specialist to get actionable data out of either.
When I worked for the New York Times, I hired a top market research firm who spent months and six figures of my budget to tell me what features my customers wanted in a new product I was developing. The data was fantastic (it was money well spent and led to a terrific design). Sadly, I can't spend that kind of money or time on every product that comes along.
I wanted a solution to this problem that was fast, accurate and cheap.
As time goes by I work out a solution in the form of an algorithm that orders ideas according to choice and desire. The algorithm accounts for what people think about an idea and how deeply they feel about it. It's gives a three-dimensional view of how people prioritize ideas. (There's a ton more to this, which I'll discuss in other blog posts.)
My algorithm is a new solution to a problem, but running it on paper is a bear. As a fun side project, I built pairLab to app-ify the algorithm I wrote. Along the way added features including real time results, a way to capture and feed ideas suggested by survey participants into pairLab so I can leave nothing unanswered and many others.
This annoying professional problem that led to a lot of reading about how people make choices and that became a fun project has changed the way I work. I no longer guesstimate what my customers need, they tell me through pairLab. I use it at Princeton University, where I work, to help:
- develop new digital products
- my stakeholders understand what will and will not work
- plan my product roadmap
- develop communication plans and social media messaging
Yeah, serendipity.