We were solving the wrong problem until we started reading support threads.
When we started working on All Cloud Hub, most of our growth ideas looked pretty standard.
SEO.
Social media.
Product directories.
Launch platforms.
Then we tried something different.
Instead of asking where to market the product, we asked where people were already frustrated.
That led us to cloud storage support forums, Reddit threads, app reviews, and community discussions.
A pattern showed up almost immediately.
People rarely complained about storage limits.
Instead they were asking things like:
"Which account is this file in?"
"Why can I only find it on one device?"
"Who owns this shared folder now?"
"Why do I have three copies of the same document?"
Those conversations completely changed how we talk about the product.
Instead of describing features, we started describing the workflow problems people were already trying to solve.
I'm still early in the process, but it made me realize something.
Support forums might be one of the most underrated places for customer research. People explain their problems in their own words, without trying to sound like product managers.
I'm curious where everyone else finds the most useful product insights.
Support forums?
Reddit?
Sales calls?
Something else?
Strong point. Involuntary churn is one of the few growth problems where the customer has already been acquired and still intends to stay. Beyond smart retries, improving payment-method coverage and making card updates frictionless can turn retention into a very efficient revenue channel.
the underrated part: use the exact words people use, not just the problems they have. "which account is this file in" as a headline will out-convert anything you write yourself, because it matches the sentence already running in their head. support threads and app reviews are the cheapest voice-of-customer research there is, and they also tell you the moment the pain hits, which is when your marketing should show up.
I like that framing. One thing I've noticed is that support threads often describe the problem after multiple workarounds have already failed. By the time someone writes "Which account is this file in?", they've usually spent several minutes searching. That makes those conversations especially useful because they're capturing the point where the workflow actually breaks.
Support forums are underrated for this exactly. We found the same pattern listening to how founders describe their bottlenecks on Indie Hackers and HN. The way people frame the problem in their own words is different from any product brief or survey response. The cloud storage examples you shared illustrate this well. Nobody writes a feature request for multi-account search, but the frustration is obvious. Do you automate the signal gathering or manually read through threads?
Manual-first is the right call for discovery. Dashboards measure what you already know to look for, and the early pattern is often something you would not think to measure. The insight about the original issue often being different from the real problem is exactly why context matters more than frequency counts. Do you capture the discovery manually in a structured way, or is it still loose notes?
Mostly manual right now. I wanted to avoid jumping straight to dashboards or keyword counts before understanding the context behind each complaint. Reading the full thread often reveals that the original issue isn't the real problem. Once patterns become consistent enough, I think automation becomes much more valuable for monitoring changes than for discovering the initial insight.