Bruce T. Dugan

Bruce T. Dugan

Abstract thinker · Linear process · Forty-four years of building things

Eighty-four percent already use it. The gap in AI marketing for small businesses is training, not tools.

Small business owner evaluating AI marketing tools while an AI assistant highlights analytics, audience targeting, and marketing strategy training opportunities.

On October 14, 2026, Meta published survey results from 3,983 small business owners. Eighty-four percent already use AI. Eighty-six percent said they need more guidance to get the benefit. Together, those two numbers tell the whole story.

Read the methodology before the findings. YouGov ran the survey for Meta between August 6 and September 2, 2026, across nine Asia-Pacific markets — Australia, Singapore, Vietnam, Indonesia, Taiwan, India, Thailand, Japan and Hong Kong. The company set quotas by size, from 1 to 49 employees. Meta states plainly that the results are not nationally representative. Meta also sells advertising tools to the people surveyed. None of that makes the data worthless. It makes it directional, and it means the U.S. figures may differ.

What are owners actually doing with it?

Marketing, mostly.

Among owners who use AI, the leading uses were content creation or editing (51 percent), local advertising and marketing (43 percent), translation (38 percent), and customer support and communications (37 percent). Three of the top four are marketing functions. This is reported as an AI adoption survey. It reads as a marketing adoption survey.

The reported returns are strong: 76 percent said AI improved productivity, and 71 percent said it improved revenue or sales over the past twelve months. Eighty percent agreed AI helps a small business compete with larger companies, and 83 percent said it lets them do things that used to require significantly more money.

What do they do with the time it frees up?

They reinvest it, which I find most encouraging.

Of the owners reporting productivity gains, 58 percent used the time to improve existing work, 50 percent on growth and business planning, and 46 percent to get more done with the team they already have. Nobody in the top three answers used it to cut staff. The hours went back into the business.

Then why does 86 percent need more guidance?

Because the tool is cheap and the judgment is not, and almost nobody is buying the second one.

Fifty-one percent named cost as the leading barrier to training. That is worth sitting with. A subscription to a frontier model costs $20 or $30 a month per seat. A day of structured training on how to use it on your actual workflows costs more than a year of the subscription. So owners buy the subscription, skip the training, and end up with staff that uses a powerful tool the same way they used the one before.

The result is the pattern I see in nearly every engagement. The business is using AI for the easiest third of the job — drafting copy, writing the ad, translating the page — and not for the part that compounds, which is connecting those outputs to each other so the work runs without a person starting it each time. Content creation at 51 percent and workflow automation nowhere in the top four is exactly that shape. But remember: AI is a powerful tool, but you can’t verify what you don’t know.

If marketing is where the adoption already is, then marketing is where the discipline has to arrive first: a brand narrative the generated copy is held against, one set of claims, one voice, and a measurement framework so you can tell which of the fifty assets you produced this month did anything. Incognito Worldwide starts there, with a single narrative, because volume without a position just produces more of the same noise faster.

What is the practical conclusion?

The question stopped being whether to adopt AI some time ago. These numbers say the market has answered it. The open question is whether you can tell what it is earning you — and 86 percent saying they need guidance is a polite way of saying most owners cannot.

Start with the hours, not the tools. Find the hours in your week that follow a rule, measure what they cost, automate those, and then check the number again in ninety days. That sequence — where AI fits, what it costs, how to tell whether it paid off — is the whole argument of AI Today.