A Series A Infrastructure company put a Founding Solutions Engineer role live in March. One title, one compensation band, one process, open to anyone in the United States.

Four months later they had made no hire. Bay Area candidates walked at the offer stage. Candidates in Denver and Raleigh accepted first-round calls quickly and then lost interest somewhere around week three, because the process was built for people who expected six stages.

Nothing about the role was wrong. The role was fine.

What was wrong was the assumption that the United States is one hiring market.

Four Markets, Not One Country

For AI and infrastructure hiring, the US behaves as at least four distinct markets. They differ on price, on what motivates a move, and on how long a process can run before a candidate disengages.

• The Bay Area. Deepest pool for research and frontier engineering, and the most contested. Compensation expectations are set by companies with vastly more capital than an early-stage startup. Candidates here are the most sophisticated readers of equity in the country, and the most likely to be running three processes at once.
• New York. Where enterprise go-to-market depth actually lives, particularly for anyone selling into financial services, media and large enterprise. The most likely place to find a founding seller who has sold a category that did not exist yet.
• Seattle. Unusual density in cloud, distributed systems and platform engineering, largely a function of who is already there and what they have already built at scale.
• Everywhere else. Austin, Denver, Boston, Miami, Tampa, plus a large distributed population who left a hub during or after 2020 and have no intention of returning. This is now a serious market, not a fallback.

Those four are not tiers of quality. They are different economies with different rules.

Compensation Is the Obvious Difference. It Is Not the Expensive One.

Everybody already knows Bay Area compensation runs higher. That gap gets budgeted for, more or less.

The costlier differences are the ones nobody budgets for.

Equity literacy varies enormously. A candidate who has been through two Bay Area exits will ask about preference stack, option pool refresh and the last 409A within the first twenty minutes. A strong engineer in Boston with an enterprise software background may take the equity component at face value and then discover its actual shape during reference conversations, at which point trust is harder to rebuild than it would have been to establish.

Process tolerance varies too. Hub candidates expect a long, structured, multi-stage process and read a short one as a company that is not serious. Outside the hubs, a five-stage process with a take-home reads as a company that does not value people’s time, and the strongest candidates simply stop replying.

And risk appetite differs. Joining a fifteen-person company is a normal career move in San Francisco and an unusual one in most other places. That does not mean people elsewhere will not do it. It means the conversation has to be different, longer, and much more specific about what happens if the next round is hard.

What This Means for Each Function

The practical answer is not to pick one market. It is to pick a market per function, deliberately.

• Research and frontier engineering. Density still matters. These roles benefit from proximity to people solving similar problems, and hub premiums are usually worth paying because the alternative is a smaller pool and a longer search. Budget for the market, not for the national average.
• Infrastructure and platform engineering. The most portable technical function. Seattle carries real depth, but strong distributed systems people are genuinely everywhere, and this is the function where remote hiring works best with the least compromise.
• Founding go-to-market. Hire where the buyers are, not where the office is. If the first ten customers are large financial institutions, New York is not a preference, it is a requirement. If the product sells to engineering leaders at tech companies, the geography matters far less than the technical credibility.
• Solutions and sales engineering. Follows the customer base and the time zones it lives in. This is the role most often placed badly, because it gets hired against a sales spec and then judged against an engineering bar.

One plan, four markets, four sets of expectations. That is more work than a single national band, and it is the difference between an offer that lands and an offer that gets used as leverage somewhere else.

Three Things to Fix Before the Next Search Opens

Benchmark by market, not nationally. A single band for a role across the US will overpay in one place and lose in another. Carry a range with an explicit market attached to each end of it, and know which end you are willing to go to before the first conversation, not during the offer call.

Build the process to fit the market. The same five stages will read as rigour in one city and as disrespect in another. Decide up front what the shortest defensible process is for this role, then hold it. Speed is a compensation lever that costs nothing.

Explain the equity properly, every time. Not the percentage. The shape. What the company is currently worth, what the strike is, what the vesting looks like and what has to be true for it to be worth anything. Candidates who understand the offer make faster decisions, and the ones who say no say it earlier, which is a gift.

The Broader Point

Founders talk about the US AI talent market as a single thing because that is how the headlines describe it. Shortage, war for talent, compensation spiral.

None of that is much use when there is one specific role to fill.

The companies that hire well in the US are not the ones paying the most. They are the ones who decided, before the search opened, which market each role belongs in and what that market actually expects.

It is not about competing everywhere. It is about choosing where to compete, and then competing properly.

If you are planning US hires over the next couple of quarters and the roles span more than one market, I would be glad to look at the plan with you. You can see how we approach it on our AI talent agency page.