Quantitative data is great for PMs but don't trust it blindly. It can never tell you why, or why not, or only by inference. Here's a story about how data told me the wrong, but eventually right, story 🧐
We were looking to migrate functionality from a legacy app to a new one. This is always a good chance to spring clean, but also there are likely going to be things you need to lift and shift. We wanted to make sure we were concentrating on the right areas.
We had good analytics and ran site-wide reports to see what was used most often and by whom. There was one part of the product, a data visualisation and exploration widget, that was being hammered. By far the highest usage stats. People clearly LOVED this widget.
Now, the obvious pressure would be to move that widget straight away to the new platform. Lift and shift! Do some more deep dives on other stuff. This one's good to go. Make it happen!
But, I was suspicious...
But, I was suspicious...
Don't get me wrong, the widget was good and it passed the smell test. I could see why people would use it a lot. But that much? Something didn't make sense. I wanted to know more. Stats weren't enough. We installed @fullstory to get a view on what was happening
It soon became clear exactly what was happening. People were definitely using the widget, but every time they clicked an element to explore a different bit of data, the page reloaded. Not only that, but it collapsed all the accordions! Horrible UX!
It was still a key piece of functionality and we still moved it but we definitely didn't lift and shift. It was a perfect example of using tech debt mitigation to not just replatform but make users' lives better. But why hadn't anyone complained?
I think a lot of it is just "stuck drawer" syndrome. It was useful enough to still get the job done, kind of annoying, but you could still figuratively give it a shove. People have been brutalised by terrible B2B software for too long, and will put up with a lot!
This is why I remain convinced that some of the PLG enablers (slick onboarding, excellent UX, fast time to value, analytics) are essential no matter how sales-led your solution. We owe it to our users and ourselves!
Loading suggestions...