Signals

Frustration Is a Better Signal Than Fit

Someone annoyed at a competitor in public is warmer than anyone who matched your filters.

Tim Burnham, Founder & CEO

Tim Burnham

Founder & CEO

August 20, 2026

Years ago I cold called for a finance software company.

The good calls all started the same way. Somebody said what we have now is pretty bad, and I would look at something else.

The simple idea

A filter tells you someone could buy.

Frustration tells you someone wants to.

Those are not close to the same thing, and only one of them is on a list you can purchase.

How it applies

People complain in public now. Under a competitor's post. Under a review. Under an announcement about a price change or a new feature nobody asked for.

They are not just annoyed. They are annoyed and they are typing about it with their real name attached, in a place you can read.

That person has already done the hardest part of your sales process for you. They have admitted the status quo is not working.

You do not have to create the problem. It is sitting right there.

What I found

In August I sat down with a group of people who all build this kind of thing and we listed signals nobody was using yet.

Negative sentiment on a competitor's content was the one that nobody in the room had built.

Everyone agreed it was the strongest. Everyone had skipped it, because it does not come in a data feed and it does not produce big volume.

That combination is usually a good sign. Strong, obvious, and unbuilt is a decent place to spend a week.

Do not quote someone's comment back at them and do not name the tool they were complaining about. It reads like you were watching, and it puts them on the defensive about a choice they already made. Talk about the problem underneath the complaint instead. They will connect the two on their own.

How it works

Pick 5 to 10 accounts. Your direct competitors, plus any big vendor whose customers overlap with your buyer.

Scrape the comments on their posts. I use an Apify comment scraper for this. It runs about $5 per 1,000 results, so the whole thing costs less than lunch.

Run each comment through a sentiment step. An AI step is fine here. You are sorting for three things. Open frustration. Questions like has anyone found an alternative. And the quiet ones, where someone answers a customer's complaint and a second customer replies me too.

Check the commenter against your ICP. This is where most of them fall out. Plenty of the angry people are competitors of the competitor, or students, or nobody in particular.

Reach out about the problem. Not about the post.

There is a faster version of this, which is timing rather than sentiment. When a competitor has an outage, their customers gather in one place and complain with timestamps. That is a few hours where every one of those people is thinking hard about switching. Being in their inbox that afternoon is worth more than being in it any other week of the year.

The limit

The volume is small.

You might get a handful of real names a week from this. It will never fill a pipeline on its own and I would not build a whole motion on it.

But run it next to a normal filtered list and compare. In my experience the person who complained converts at a rate that makes the filtered list look like guessing, because it mostly is.

There is also a taste question. Sitting in a competitor's comment section waiting for someone to be unhappy feels a bit ghoulish, and I understand people who will not do it.

My honest take is that the person complained in public because they want the problem solved. Solving it is not the rude part. Being weird about how you found them is.

If you build this and it works, or it does not, I would genuinely like to hear which.

Share

Working on something like this? Tell us what's slow and we'll tell you what we'd build.