AI for Small Businesses: It's Not Irrelevant, You Just Haven't Found the Use Yet
Many business owners tell me, "AI is impressive, but it doesn't really have much to do with our line of work." That one sentence is exactly the first hurdle where small businesses tend to get stuck when getting started with AI.
This year's Federal Reserve small business survey, the OECD's report on AI in SMEs, and a Forbes column published in early November all point to the same thing: what holds small businesses back usually isn't the technology. It's that they haven't yet figured out where AI should be used.
Today's Highlights: Small Businesses Get Stuck at "Finding the Use Case"
The author of the Forbes column, Natalie Madeira Cofield, is CEO of the Association for Enterprise Opportunity (AEO) in the U.S. Drawing on several studies, she points out a paradox: the small businesses that stand to gain the most from AI aren't necessarily the first to adopt it. According to the association's report released in September, only 8.4% of solo businesses use generative AI, compared with 20% of small businesses with employees. Yet among those who have tried it, 92% of solo businesses report a positive impact, higher than the roughly 70% of businesses with employees.
Even among businesses already using AI, adoption tends to be shallow. The Federal Reserve's 2026 Report on Employer Firms (based on the 2025 Small Business Credit Survey) shows that 46% of businesses already use AI and another 15% plan to start within a year. But about half of those users are still experimenting, 44% have only partially integrated it, and just 7% have fully woven it into day-to-day operations. The most common uses are writing and marketing (83%), personal productivity (61%), and planning and analysis (51%), mostly low-risk work that doesn't touch accounting or core operations.
As for results, 71% of users say productivity went up, 39% say product or service quality improved, and 31% say revenue increased. These are all self-reported, though, and most say their labor costs haven't changed. Current users' biggest concerns are accuracy (46%) and how hard it is to tailor tools to their own needs (43%); those planning to adopt worry about finding the right tool (54%) and the time it takes to implement and train (37%).
Most striking of all: among the one-third of businesses with no plans to use AI at all, more than half say the reason is that "AI doesn't apply to my business." The columnist also argues that what many owners lack isn't technical skill, but the ability to see how AI connects to their own business.
Digital foundations matter, too. The AEO report found that only 2.8% of businesses with no online presence use AI, versus 19.3% of those with a website or online operations. An April OECD report surveyed more than 2,000 SMEs across 12 countries (the sample isn't nationally representative) and concluded that most businesses rely on off-the-shelf tools, with lack of time, maintenance costs, and skills gaps as the main barriers, and cybersecurity risk as a major concern. In other words, the question is no longer "Are there tools available?" It's that the tools still aren't connected to everyday work.
One caveat, though: the columnist herself notes that much of this data reflects 2025 or earlier, and owners' views may have shifted since. The self-reported results are also best read conservatively.
Jesse's Take: Start With One Thing You Repeat Every Week
After going through these numbers, the phrase that struck me most was "doesn't apply." Helping SMEs in Taiwan adopt AI, I hear something similar all the time. A common scenario: the owner of a hardware store that's been open for twenty years thinks AI is something only internet companies use. Yet every day he answers customers on LINE (Taiwan's go-to messaging app) asking "Is this in stock? How much is it?", and at the end of every month he reconciles the books line by line in Excel. These are exactly the tasks where AI can help the most. The problem isn't that AI doesn't apply; it's that no one has helped him match it to a real use.
Tip 1: Look for uses in your repetitive grunt work, not in the tools.
Don't start by comparing which AI tool is the most powerful. First, list the tasks you repeat three or more times a week that wear you down, such as answering the same price inquiries, putting together quotes, or logging incoming stock. Pick one, try it with AI for two weeks, and see how much time you actually save. A common scenario: a salesperson at a small trading company answers a dozen or so nearly identical inquiry emails every day. Just having AI draft those replies and a person edit them could save several hours a week.
Tip 2: Get the information out of the owner's head first.
Businesses that operate online use AI at much higher rates. My reading is this: AI only has something to work with once a company's information is digitized. If your product list, prices, and FAQs are still scattered across LINE chats, paper notes, and your own memory, spend an afternoon organizing them into a spreadsheet or document. It will come in handy no matter which tool you end up using.
Tip 3: Make accuracy your acceptance standard.
Nearly half of the users in the survey worry about AI making mistakes, and they're right to. For anything involving quotes, accounting, or promises to customers, let AI write the first draft, then have a person check it before it goes out. I'll also be candid: most of these results are self-reported, and success stories compiled by associations or vendors inevitably carry some selection bias. Don't expect revenue to jump the moment you adopt AI; aim first for reliably saving a few hours each week.
These three tips may look basic, but I've seen too many companies do it backwards: they buy a tool first, then try to figure out what to use it for, and three months later the owner is the only one who still opens it now and then. Get the order right and the tool becomes the least important piece. Even a free chatbot is enough to get one small task running smoothly.
A first step you can take today: Before you leave work, grab a sheet of paper and write down "Things I did three or more times this week": at least five of them. Then circle one to hand over to AI and try out tomorrow.
Further Reading
Related posts on JCXO
Original sources
Comments