The AI-jobs conversation is loud, and almost all of it is about big companies. Enterprise layoffs. Public walk-backs. Predictions about entire professions.

Payroll records from actual small businesses tell a different story. Last week, Gusto, a payroll company serving several hundred thousand small businesses, published a study of 2,262 of its customers. Roughly 1,600 had adopted AI. About 670 knew about AI and chose not to use it. Then the researchers watched what happened to everyone's payrolls over the following year.

The adopters came out ahead. In the year after adopting AI, those businesses grew headcount about 7% more than comparable non-adopters. The gap wasn't a one-time hiring spike. It built steadily through the year, from about 4.5% in the first quarter to nearly 7% by the fourth.

The smallest businesses gained the most. Companies with fewer than 10 employees grew about 10% more than comparable businesses that didn't adopt. For a typical nine-person shop, that works out to one extra employee within the year.

This matters beyond any single storefront. Small businesses employ roughly half of the American private-sector workforce, and the prevailing storyline says automation is coming for them first. This data says the opposite, at least for the businesses that actually use the tools.

Who Got Hired

Not technologists, for the most part. Restaurants, manufacturers, and accounting firms all showed hiring growth after AI adoption. The new roles were hands-on and customer-facing: cooks, technicians, teachers, front-desk and admin staff. Gusto's researchers described them as do-ers, people who produce more of the work the business actually sells.

The strongest sector was health care and social assistance. Small practices with fewer than 10 employees (dental offices, mental health clinics) grew nearly 20% more than comparable non-adopters. They hired more care providers: dental hygienists, therapists. A practice that spends less time on patient paperwork can see more patients. That shows up as revenue, and then as staff.

Why It Works Out This Way

The mechanism is unglamorous. In a business with a handful of employees, the owner's hours are the scarcest resource, and most of them go to overhead: scheduling, paperwork, invoices, follow-up. AI handles a surprising amount of that. Hours come back. Hours back means capacity to serve more customers, and serving more customers means hiring people who deliver the work.

Nich Tremper, the senior economist who leads Gusto's research group, put it this way: "The smallest businesses are the ones who are most used to doing more with fewer resources... AI may be shifting their work back to their expertise, and supporting a business's productivity. Small business owners seem to be investing that back into their businesses through increased hiring."

The Fine Print

This is one study of one payroll platform's customers, and it shows association rather than proof. Businesses that adopt AI may already be better run or growing faster for other reasons. The comparison group helps: Gusto compared adopters against businesses that knew about AI and declined to use it, which controls for awareness. It can't control for everything.

One finding deserves attention from anyone above ten employees. Businesses with 10 to 24 employees showed no measurable headcount change after adoption. The researchers' read is that companies that size already have processes in place, so the gains likely show up somewhere other than payroll.

The broader survey data points the same direction. McKinsey asked the question twice. In 2025, 32% of the businesses it surveyed expected AI to reduce headcount over the following year. When it checked back a year later, only 14% reported that it had. The expectations outran the results.

What to Do With This

The finding isn't that AI magically creates jobs. It's that the businesses in this study treated AI as a way to reclaim hours from overhead and put them back into the work customers pay for, and their payrolls show it.

The other path, using AI to run the same business with fewer people, is the one big companies tried and, in a growing number of cases, publicly reversed. We covered that in July: service teams replaced by automation, then rehired when quality slipped and customers pushed back.

You don't need an AI strategy for this. You need one process that eats too much of your week, and an honest attempt to give those hours back to the actual work. Then watch where they go.

If you want help picking the first process, and an honest read on whether the tooling is worth it, that's a conversation worth having with someone who does this for a living. A practical look at where your hours actually go, not a sales pitch.