Not asking anyone to try my app, genuinely just stuck on a problem. I’ve got real traffic, real installs happening, but almost no one sticks around long enough to tell me why they left or what confused them. I’ve reached out directly a couple times and just got silence.
For those of you further along than me, how did you actually get real, honest, detailed feedback out of early users? Did you pay someone to do a real session with you? Find them through a specific community? Just get lucky with one person who cared enough to be blunt? Genuinely trying to learn the mechanism, not just “ask nicely and hope.”
Hey, @Debnix
Hope you are doing well!
One thing that worked for me was making the feedback request extremely specific. Instead of asking “Why did you leave?”, ask them to walk through the exact point where they got stuck. Even a 10 to 15 minute call can reveal more than dozens of survey responses.
Really useful distinction, thank you. “Why did you leave” does invite a polite, vague answer, “walk me through where you got stuck” actually forces specifics. I don’t currently have a way to get anyone on even a 10-15 minute call, no real users to ask yet, but I’m going to hold onto this for the next real install, ask for that walkthrough directly instead of just hoping they leave a note or reply to an email.
Yes — make the request specific. “Why did you leave?” invites a polite answer; “At which step did you hesitate, and what did you expect to see next?” gives you something actionable. One 10-minute walkthrough paired with the user’s actual behaviour usually teaches more than 20 vague ratings. Ask about one task, not the whole app.
That’s the third time now someone’s pointed me toward this same idea, specific question, one task, not the whole app. Starting to feel like a real pattern rather than one person’s preference. “At which step did you hesitate, and what did you expect to see next” is a genuinely better question than anything I would’ve thought to ask. Appreciate it.
The exit-survey idea is new and something I don’t have yet, a single required question right at the moment of uninstall, rather than depending on someone replying to an email afterward. That’s worth building. Appreciate you laying it out clearly.
Good point on the structure, multi-choice with an optional “Other” is a much better ask than a blank text box for someone who’s already leaving. Makes sense that lower effort means more responses.
The advice above about asking one narrow question is right. I want to add a data point about where you should not look for honest feedback, because it is where most app founders look first.
I crawl the Shopify App Store daily and hold full review histories for 912 apps, which is 60,278 reviews. The distribution: 90.7% of all reviews are 5 stars, 4.7% are 1 or 2 stars, and 25.7% of the 5 star reviews are written within one day of install.
A quarter of your positive reviews come from someone who has had the app less than a day. They are not telling you the product works. They are telling you the install did not hurt. That is the feedback channel with the most volume and the least signal in it.
Negative reviews have the opposite shape. The median 1 star reviewer had the app for 90 days, and for 2 star it is 120 days. Real judgement shows up around the three month mark, which is far too late to be useful to you now.
What that means for your situation: the honest feedback you want is structurally invisible in every public channel. You have to go and take it before someone forms a verdict. The exit survey idea further up this thread is the right instinct, one required multiple-choice question at uninstall, because it catches people at the only moment they are motivated to say anything at all.
One more thing worth knowing while you are small. Across those reviews, developers reply to 51.9% of 1 star reviews and only 27.1% of 5 star ones. Everyone is doing damage control and almost nobody asks the happy users what actually worked. Your happy users are the cheapest interviews you will ever get, and they are the ones being ignored.
(Disclosure: the numbers come from BestAppify, which I build.)