Small business owners are cutting outside spending faster than at any point in 2026, and feeling better about the economy because of it. New research says the two are connected. Whether that connection holds depends on what gets cut.

Citizens' Q4 2026 Business Pulse survey, fielded September 1 to 17, 2026 across 500 US business decision-makers, found revenue expectations climbing steadily through the year: 43 percent of owners expected revenue increases in Q1, rising to 48 percent in Q2, 50 percent in Q3, and 54 percent heading into Q4, the highest level recorded all year. Hiring plans are just as strong: 83 percent intend to maintain or increase full-time headcount over the next three months, with only 3 percent planning cuts.

The sharpest split in the data is around AI. Thirty percent of owners plan to increase technology spending in Q4, including AI, up from 23 percent the quarter before. Among businesses using AI regularly across multiple functions, the effect on confidence is dramatic: 74 percent expect revenue increases, against 54 percent overall. Fifty-one percent report extreme or very high confidence in the economy, against 29 percent of everyone else. And among those extensive AI users, 94 percent have reduced or eliminated spending on outside vendors, led by marketing services at 60 percent, data analysis and reporting at 39 percent, and bookkeeping and accounting at 38 percent.

Here is the trap hiding inside a genuinely good trend. AI is very good at tasks. It can write the ad copy, build the report, reconcile the transactions, draft the campaign brief. What it cannot do is tell you whether the ad is aimed at the right customer, whether the number in the report should change tomorrow's pricing decision, or whether the campaign fits where the business is actually trying to go next year.

Those are not tasks. They are decisions. And the problem with cutting a vendor is that the invoice and the judgment usually left the building together, you do not notice the judgment is gone until a decision goes wrong with nobody accountable for having made it.

Kevin has coached 200+ businesses one-on-one across 30+ industries over more than 30 years in business, including his years inside DDB, Ogilvy and Saatchi and Saatchi working with 60+ global brands. The pattern repeats at every size: the businesses that get hurt by a new efficiency tool are rarely the ones using it well. They are the ones who let the tool quietly take over a decision nobody assigned it.

Before you cut, or keep, a line of external spend this quarter, run it through three questions.

1. The Task Check. Is this pure execution, writing, formatting, calculating, reporting, with a clear right answer once the inputs are correct? If yes, AI can very likely do it as well or better, for a fraction of the cost.

2. The Decision Check. Does doing this well require judging what matters more than something else, which customer to prioritise, which number is the real problem, which campaign fits the brand? If yes, that is not safe to hand over silently. Someone still has to own the call.

3. The Steering Check. Once you have cut the spend, who is now checking whether the result is actually working? Every vendor relationship came with an implicit second pair of eyes. If AI replaces the doing, make sure a person still replaces the checking.

Run every line item through those three, and the picture usually splits cleanly: the marketing execution can go to AI, but someone still needs to own the marketing strategy. The bookkeeping entries can go to AI, but someone still needs to look at the numbers and decide what to do about them.

The categories being cut fastest, marketing, data analysis, bookkeeping, are exactly the categories where a small business owner is least likely to notice the judgment gap opening, because the AI output looks finished. A report that renders cleanly feels authoritative whether or not anyone checked its conclusion against reality.

That is the quiet risk behind an otherwise healthy trend. Confidence is up. Costs are down. But confidence built on execution you have automated is not the same as confidence built on decisions someone is actually making well.

“Kevin had a no-nonsense approach and tackled the issues head on. Accountability was the key.”

— Ashley Berrysmith, Icehouse Coaching Programme

That is the piece AI does not replace. Accountability for the calls that actually move a business, what to cut, what to keep, what to bet on next quarter, still needs a person who can be wrong and answer for it.

If you have cut the vendor but you are not sure who is making the calls that used to be theirs, that is exactly where your conversation with Kevin starts.

...and you become the owner who knows who is making every call the vendor used to make.