How to find people describing your problem in their own words
A practical playbook for finding buyers on Reddit and YouTube: the search operators that beat native search, the phrases that signal intent, and where the real signal hides.
Almost nobody searches for your product by name. They haven’t heard of it. What they do instead, constantly and in public, is describe the problem it solves in their own words. “Anyone know a tool that does X.” “Moving off [big competitor], what are people using.” “Is there anything cheaper than [competitor] that doesn’t fall over?”
Each of those is a person mid-decision, saying out loud exactly what they want. Find that sentence and you’ve found a buyer who is warmer than any cold list will ever be. So the job comes down to two questions: where do those sentences get written, and how do you spot one when you see it?
Here’s how to do it by hand. No tool required for most of it.
Google finds Reddit better than Reddit does
Reddit’s own search is famously bad. Google, which pays Reddit a small fortune to index it, is much better. The move is to search Reddit through Google using operators.
Start with site:reddit.com and a phrase in quotes, so Google matches it exactly:
site:reddit.com "looking for a tool that"
site:reddit.com "alternative to [competitor]"
site:reddit.com "what do you all use for" [your category]
Then tighten it. A few operators do most of the work:
site:reddit.com/r/SaaS ("recommend" OR "recommendations") pricing
site:reddit.com intitle:"vs" [your category]
site:reddit.com "still using" [competitor]
intitle: limits the match to the thread title, which is where comparison shopping (“X vs Y”) usually lives. The parentheses let you OR a few phrasings together. And if you only want fresh threads, add a date filter in Google’s own tools, or append after:2026-01-01 right in the query. Stale threads are dead leads. Someone who asked two years ago has already bought something.
Swap the bracketed bits for your own category and competitors, run a handful of these, and you’ll have a page of real conversations in about five minutes.
The phrases worth searching for
The operators are only as good as the phrases you feed them. After you read enough of these threads you start to recognize the language of someone who’s ready to move. Here’s a starter set. Search them literally, in quotes.
Shopping right now:
- “looking for a tool that”
- “is there anything that”
- “what do you all use for”
- “recommend a” [category]
Actively switching (the warmest of all):
- “alternative to” [competitor]
- [competitor] “but cheaper”
- “moving off” [competitor]
- “about to cancel”
- [competitor] “keeps” (raising prices, breaking, logging me out)
Frustrated, not yet shopping (get in early):
- “I’ve tried” (three, four, several) “of these”
- “why is every” [category tool] “so”
- “does anyone else hate”
That last group is the sleeper. Someone venting about the category hasn’t started comparing options yet, which means you’re the first reasonable voice in the thread instead of the tenth. Those convert better than they have any right to.
The signal is three replies deep
Here’s the mistake most people make: they read the post title, decide it’s not a fit, and move on. But the title is almost never where the intent is. The title is the setup. The buying signal is down in the comments, where someone replies to the original poster with “honestly I switched from X last month and here’s what I wish I’d known.”
That person, the commenter, is often the better lead than the one who started the thread. They’ve used the competitor, they have an opinion, and they just announced it to everyone. You have to actually read the thread to find them, which is exactly why so few people do.
So when a thread looks even loosely relevant, open it and read the replies. That’s the work. That’s also the edge, because it doesn’t scale for your competitors any better than it does for you.
Set an alert for the ones you’d miss
You can’t run searches all day. For the ongoing version, F5Bot is free and does one thing well: you give it keywords, and it emails you whenever they show up in a new Reddit post or comment. Point it at your competitors’ names and the two or three problem phrases your customers actually use. Now the threads come to you, the day they’re posted, while the person is still in the market.
It won’t tell you whether a mention is a real lead or just chatter. But it makes sure you see the conversation while it’s fresh, which is more than half the battle.
YouTube is the half nobody searches
Everyone’s fighting over Reddit. Meanwhile the comment sections under YouTube videos are full of the same buying language, and hardly anyone mines them.
The catch is that YouTube has no built-in comment search, which is probably why it stays underused. Two ways around it:
- For a smaller video, load the comments (scroll until they stop appearing), then hit Cmd+F or Ctrl+F and search the page for words like “still”, “switched”, “worth it”, or a competitor’s name.
- For a video with thousands of comments, a free comment finder like CommentShark will pull them all and let you filter by keyword in a few seconds.
The best hunting ground is a competitor’s review or tutorial video. The comments under “[Competitor] review 2026” are full of people asking “is this still the best one?” and “anyone found something cheaper?” That’s a room full of buyers raising their hands, and the competitor put them all in one place for you.
The honest catch, and where a tool helps
Everything above works. I’ve described it in enough detail that you can go do it today without buying anything, and you should.
But you’ve probably noticed the theme: it’s a lot of reading. Running operators, opening threads, going three comments deep, checking the phrasing, doing it again tomorrow because the good threads are the fresh ones. It’s effective and it’s genuinely tedious, and tedious is the first thing people quit.
That reading is the part I build a tool for, so weigh this accordingly. The desktop app I work on runs exactly these passes for you, at the same depth: it opens the full thread and the comments under it, then scores each candidate against a short description of who you sell to and says, in plain English, why it’s a fit. You still read the shortlist and reach out yourself. If the manual version above sounds right but you don’t have an hour a day for it, that’s what this is for.
Either way, the lesson holds. People are describing your product without knowing your name, in public, right now. Learn the words they actually use, and get to those threads before they go cold.
Two more that go with this one: what to do when you find the thread, without getting banned, and why the reading part keeps breaking the tools that try to automate it.
Want to try it on your own market?
Leadiver finds warm leads on Reddit and YouTube, scores them against your customer, and drafts the first reply. It browses like a real person, not an API call.
Get early access