Stefan
Latest posts by Stefan (see all)
When I started in Product Management, I had little concept of the role. I researched the hell out of it, but ultimately there was no blueprint, no magic key, no source of answers.
During my early days as a PM, there were some episodes which stuck with me as examples of good product management. I am sure there were a thousand more (actually, I am certain of it), which I probably didn’t recognize as such. Maybe in the future, hindsight will be kind to me and help me see those as well. Until then, those are some of the stories that stuck.
The power of asking your users
At the time I worked on Google Analytics, the Web Analytics Tool. GA is a powerful tool, with a lot of configuration capabilities. Some of them can be very complex. One component which allows more complex config in GA is called ‘the condition builder’. A condition builder’ is a visual element which allows the users to specify logical conditions. Those conditions are applied to the data, and if the conditions are satisfied, some action happens, like a label gets applied or a event fires.
Condition builders are an interesting thing. They are powerful, but combine a couple of things. One main one is that they exploit the concept of logical operators, such as And and Or conditions, or ‘smaller than’ or ‘larger than’. Logical.operators aren’t the easiest thing to understand, and in addition to that they were used to be denoted in it’s formulaic version, i.e. ‘smaller than’ became ‘<‘, and ‘larger than’ was ‘>’. As you can imagine, to the technically minded folks this seems, well, logical, but to most folks curious about the business or marketing side these operators were hard to read and understand.
One day I went to our San Francisco office (I was based in Mountain View at the time) on a Friday. Most folks were living in SF and commuting to MTV, so they could enjoy city life in their free time, and also benefit from the many services of the Google main campus (such as that famous nap pods, or one of the many many cafes serving all kinds of food). As I looked for a desk to squat (hot real estate of a Friday in SF), i bumped into a coworker. He was sitting on one of the free desks and seemed deeply entrenched in thought. When he saw me he invited me to check out what he was working on.
Dan, who was one of our more recent additions, was awesome. He was always enthusiastic, and really seemed to deeply understand our users and their needs. So Dan looked at the condition builder’ feature and wondered how to make it more effective and helpful. What he did, which was remarkable, was that he offered a sort of families survey to our users. He showed them a randomized list of operators in their formulaic format, i.e. a list of ‘<‘, and ‘>’, and so on. And then he had users associate these operators with the text based meaning of them, i.e. march against a list of ‘smaller than’, and ‘larger than’. The approach was great. It allowed users to engage in a fun way, and expose their depth of knowledge (or lack thereof) without exposing themselves. Genius.
The results were pretty clear. Formulaic operators aren’t a wide spread concept that web analysts and marketers care too much for. The data indicated a big opportunity to make this part of the product more helpful, which is what Dan went ahead and proposed.
To this day this story stuck with me, and still makes me chuckle. It truly encapsulated how extracting the right data can be the best way to make your decisions easy and clear, but if you don’t know how to do that you can easily get suck as well. Big thanks to Dan for helping me become a better PM that day 🙏.
