One thing I’d like to point out: when a community admin says « people haven’t complained about feature x », I always wonder if they know about the people who didn’t complain but just left the community.
There is always survivor bias amongst the people you get feedback from: they are those that stuck around. What about those that didn’t?
For many years, I took the trouble to message users who left the Facebook group my community lives in. Most people don’t answer (selection bias) again, but quite a few did, and their feedback was the most precious I ever got. And it was not what my active members were telling me. It shone light on our blind spots.
Therefore, I would not take lack on complaints as an indication an onboarding strategy is the good choice — because you’re precisely not likely to hear from those who ran for the hills as soon as they cracked the door open.
This also makes me question whether « complaint-driven development » doesn’t make a platform (or community) run the risk of getting stuck in a reinforcing feedback loop, improving things for those for whom it already works, and not for those for whom it doesn’t.
As a community manager, I will at times make changes that were not requested by the community, and refuse changes they ask for, because from my vantage point I can see that beyond the desired consequences there will be others that are negative. We all know the road to hell is paved with good intentions.
Yeah these are good points! on our hosting when someone cancels their subscription we ask them for feedback in hopes of capturing this sort of thing, though unfortunately once someone has made up their mind about leaving they’re not always willing to spend more time explaining why.
Exactly! In my community I have the « advantage » that people are joining because of a sick pet, and (given Facebook and the community topic) one can assume their digital literacy is low. So I can reach out saying « hope everything is ok with your pet, did you leave by mistake, in case anything went wrong for you in our community do let me know » — but not all contexts have that kind of relational lever.
This is why, imho, user research with non-users or naive users is so crucial to gathering feedback on tool functionality — second to that, paying close attention to what very new users recount of their first experiences, because they might provide a glimpse into where it is failing for the silent quitters.
It feels this is just general feedback to the Discourse team of:
Just cause nobody complains, does not mean you don’t have a problem
It is legitimate feedback but should be discussed in its own room.
Note that proactively as a company we reach out regularly to customers and even open source users for interviews. I think we do 1-2 a week at this point. Hunting for feedback is one way we gather signal.