Using It Isn't Earning From It
Henry Kim ·
This week, several surveys landed on the same number. Somewhere between 60 and 80 percent of small business owners use AI. But fewer than a third say it has measurably made them money. Why? The answer is inside the surveys. "The gap isn't about which chatbot or app a business picked. It's about what happened after the signup." The pattern is always the same. The owner or one employee starts using a free AI tool. Drafting emails, summarizing reviews. It works well enough to keep. But the business itself changes nothing. No rules for when to use it. No review step for accuracy. No plan to train anyone else. The tool becomes one person's habit — never the company's way of working. Picture it: one person in the corner with AI open. The person next to them has no idea what it is. The owner says "we use AI too." That is what AI adoption looks like at many small businesses right now. One practitioner put it this way: the requests coming in have changed from "should we look at AI?" to "our employees already opened it — now help us do this properly." That is not a product-selection question. It's an operating-model question. This is not a new story. Cloud software in the 2010s did the same thing. Social media marketing before that. The lower the barrier to entry, the faster the adoption and the easier it is to skip the implementation. AI has pushed that barrier to nearly zero. You can start in minutes, so there's no reason left to think about how to embed it. A Canadian survey put it more directly. 32% of small business owners say AI has had little to no impact; 27% aren't sure how it even applies to what they do. The reason is on the page too: "They're the boss, the bookkeeper, the salesperson and the customer service department, sometimes all before lunch." What that person lacks is not another app. It's time to sit down and think. Waiting is understandable too — many say they'll move once they see it working at a business like theirs. But while they wait, the gap widens. The ones who use it change how they work; the ones who watch just watch. So the question being asked is wrong. Not "which AI should I use" but "where do I put it." The order of operations is simple. 1) Pick one thing that happens every day. 2) Have a human review what the AI did. 3) Write it down as a rule. Take quote follow-ups: quotes go out daily, the AI drafts, a human glances before sending, and it becomes the rule. Daily frequency means the effect shows within a week. Starting with grand strategy is what enterprises do when they have analysts. One analysis put it this way: "Value follows the surrounding work rather than the model itself." True. AI is not an app — it's a change in how work gets done. If the way of working doesn't change, changing the tool changes nothing. Adoption is not a decision; it's a day one. The day you hit signup is day one, not the finish line. Spend ten times the hours you spent choosing the app on deciding where to embed it. Reverse that order and the tools pile up while the money stays flat. On social, "AI millionaire" videos go viral. Views are not reality. Reality is the owner's day — four jobs before lunch. One thing embedded in that day beats a hundred tutorials. Throw away the adoption checklist. Write an embedding checklist instead. Today: pick one daily task. This week: put AI on it and review it yourself. Next week: write it down as a rule. Are you using AI — or have you embedded it? Tonight, pick just one thing. One thing you do every day.