30 Things VCs Want Someone to Build

In The Atoms Gap I looked at what venture investors write about. Mostly about venture.

This time I only kept the part that matters if you build or invest. The posts where an investor says a specific new company should exist, and names what it would make. Out of 1,392 ideas from 841 blogs and newsletters, between August and November 2026, there are 194 of those.

Most ideas have one believer

87 percent of the 194 are argued by exactly one source. Nobody else in these four months makes the same claim.

The crowded ones are easy to spot. AI coworkers built at the application layer, 14 sources. Selling the work instead of the software, 7. An AI security operations center, 6. If you raise on one of those, every investor has already read three memos about it.

The other 169 sit alone. Some of them are wrong. But that is also where the next company comes from, before it is obvious.

Who is supposed to pay

84 of the 194 opportunities sell to large enterprises. 40 to consumers.

8 sell to small businesses. One sells to mid-sized companies. So if you want to build for the business down the street, almost no investor is writing about it. That's either a gap or a warning, and I think it is mostly a gap, because those businesses have the same problems as everybody else and less help.

When investors get concrete, atoms show up

In the first post atoms, energy, robots, biology, factories, mobility, were 9 percent of everything investors wrote. In the concrete opportunities they are 24 percent. When an investor actually names a company that should exist, it is much more often a physical one.

They are also the most ambitious. Atoms opportunities score 0.56 on techno-optimism, the rest 0.42. And they cost more. Half of them need a factory, a fleet or heavy assets. Across all 194, 63 percent are software a small team can ship, 19 percent need heavy assets, and 34 percent need approval from a regulator.

Nobody says when the money comes

For 163 of the 194, 84 percent, the post gives no basis to say when the first real revenue arrives. Investors are good at saying what should exist. They mostly leave the when to the founder.

The 30

I picked 30 that score high on both the venture score and on techno-optimism, that name a real product, and that cover different fields. The quote is the author's own sentence. The facts under each one are judgments, not guarantees. Filter them by what you can actually do.

What I take from it

If you build, the open ideas are the interesting ones. One believer means you get to be early, and you get the investor who already wrote your pitch. Start with the ones two founders can actually begin, and look hard at small and mid-sized businesses, because almost nobody else is.

If you invest, the crowded theses are already priced. The single-believer ideas that also need atoms are where the fewest people are looking.

Notes

• Writing is not money. This is what investors write, not what they fund.
• Scores and facts are judgments. "Two founders: yes" means a first paying customer within twelve months looks possible, not that it is easy.
• "VCs saying it" counts blogs and newsletters. A partner's newsletter and the firm's blog count as two.

Every idea links to the original. The writing belongs to each author.