68% of Real Demand Has No Product Behind It
Of 322 market gaps GapMine tracks, 219 have no product serving them. What that number means, and the 16-day window it opens.
Founders keep asking the wrong question. Not "is this idea taken?" but "is anyone actually asking for it?"
We can now answer the second one with numbers.
As of August 2026, GapMine tracks 322 live market gaps — each one a cluster of real people, on Reddit or Hacker News, describing the same thing they cannot buy. We check every gap against what already ships.
219 of those 322 gaps have no product serving them. That is 68%.
Not "underserved." Not "room for a better one." Nothing.
The gap that has been open since May
The clearest case in the dataset is Shopify bot protection.
Twelve different Shopify sellers, across two platforms, have described the same problem. One of them put it plainly:
"I get around 100 new bot customers daily…"
These are store owners watching fake checkouts eat their conversion data and their ad spend. The complaint first surfaced on 2026-05-01. It is still open today.
Twelve people. Two platforms. Three and a half months. Zero products built for it.
That is not a niche too small to serve. That is a queue nobody is standing in.
Demand shows up before anyone builds — and you can measure the lag
Here is the part most idea-hunting misses. Demand does not appear the day a product launches. It appears in complaints, months earlier.
We measure that lag directly. For 297 gaps in our dataset, we know both when the first person complained and when the market started responding.
The median lag is 16 days. For 105 of those gaps it exceeded 30 days. For 8, it exceeded 90.
One note on honesty: the average lag is 39.7 days, but that number is dragged up by a handful of extreme cases (the longest is 1,776 days). The median is the number to trust. We publish both.
Sixteen days is a small window. It is also a real one — and it is the entire reason listening beats brainstorming.
What 68% actually means
A missing product usually means one of three things:
- Nobody has noticed yet. The complaints are scattered across subreddits nobody aggregates.
- It is hard. Shopify bot filtering means fighting an adversary who adapts.
- It is genuinely too small. Some gaps are twelve people who will never pay.
That is why we grade by evidence, not by an invented opportunity score. A gap backed by twelve named complaints across two platforms is a different animal from one mentioned twice on the same thread.
The other 32%
103 gaps do have products shipping. Those are not worthless — for 130 of our gaps we also record what users specifically complain about in the existing tools.
"Competitors exist" and "the problem is solved" are not the same sentence. The second-best place to build is where a product exists and everyone using it is annoyed.
How to read this dataset
If you are hunting for something to build:
- Start from a complaint you can quote, not an idea you invented. Every gap here links to the original public posts.
- Count distinct people, not upvotes. One angry thread with 200 upvotes is one data point. Twelve people in twelve threads is a pattern.
- Check the age. A gap open since May, with new complaints arriving, is a stronger signal than one that spiked once and went quiet.
Limits of this data
We track four sources: Reddit, Hacker News, Product Hunt and GitHub. Demand that lives in Slack groups, sales calls or enterprise procurement is invisible to us. Our 68% figure describes gaps we can see, not the whole software market.
We also only count a product as "shipping" if we found it. A tool with no web presence and no discussion will look like absence to us.
We would rather state that plainly than round the number up.
Every figure here was computed on 2026-08-16 from GapMine's live dataset. Each gap links to the public posts it was built from. Nothing on this page was generated by an AI writing about a market it cannot see.