How many small businesses use AI, and where to start?

AI adoption among U.S. small businesses rose from 48% in mid-2024 to 68% in April 2025 (QuickBooks/Intuit), and around 89% already use some AI tool (ICIC/Intuit, February 2025); in Mexico the trend is clearly rising, though robust data is still lacking. Translation for you: AI is no longer just for big corporations—today it’s an everyday tool for businesses the size of yours.

If you’ve felt that “everyone talks about AI except you,” the numbers give a useful hint: it isn’t a passing fad among large companies, but a massive adoption that has already reached small ones. And what matters isn’t jumping in for the trend, but understanding where to start so it delivers real results.

How widespread is AI use among small businesses?

Much more than most imagine, at least in the U.S. market, where there’s solid data. According to QuickBooks/Intuit, AI adoption among U.S. small businesses rose from 48% in mid-2024 to 68% in April 2025: two out of three already use it in some way. And an ICIC/Intuit study from February 2025 found that around 89% use at least some AI tool (counting things as common as writing assistants or smart features inside programs they already used).

It’s worth being precise about the framing: these figures are from the United States. In Mexico the trend is also clearly rising—more and more small businesses try AI assistants to write, serve customers, or summarize information—but we don’t have data as robust or as recent as the U.S. numbers. Even so, the direction is the same, and the gap with those who don’t adopt keeps widening.

“AI isn’t for my business”: is that true?

Almost never. It’s the most common objection among small-business owners, and in the vast majority of cases it stems from a mistaken idea of what AI is today. We’re not talking about robots or million-dollar projects, but everyday tasks that almost any business has:

• Do you write emails, quotes, or messages to customers? AI drafts them in seconds for you to just review and adjust.
• Do you get the same questions over and over? An assistant can answer the common doubts and leave the important cases to you.
• Do you receive long documents, PDFs, or reports? AI summarizes them and tells you the essentials in a minute.
• Do you work in two languages for the Latino and U.S. market? It translates and adapts content with reasonable quality.

If your business does any of these—and almost all do—then AI is for you. The question isn’t “whether,” but “where.”

Where do I start without getting lost in the hype?

Start with a single use case with clear impact, not with “implementing AI” in the abstract. The most common mistake is wanting to transform everything at once or buying tools because they’re trendy. The sensible route is scoped and measurable:

• 1. Choose a repetitive, high-volume task. Look for something your team does many times that consumes time: answering common doubts, drafting texts, summarizing documents, sorting emails or leads. Start with just one.
• 2. Define clear data rules. Before feeding information into an AI tool, decide which data is okay and which isn’t (don’t upload sensitive customer data into uncontrolled services). This rule protects your business and your customers.
• 3. Measure before and after. Note how much time or money that task takes you today. After applying AI, measure again. Without numbers, you won’t know if it worked.
• 4. Keep human review. AI proposes, a person approves. Especially if the result reaches the customer, someone on your team should review it before it goes out.
• 5. Scale what works. If the first case delivered results, add a second. Growing step by step avoids wasted spending and frustration.

It’s not about “using AI” just to say you use it, but about solving a concrete problem in your business and measuring whether it was worth it.

And when does something custom make sense?

When the use case touches your own data and processes, generic tools fall short. Starting with ready-to-use tools is the right thing to do to validate. But there comes a point where you want AI to answer with your business’s real information, connect to your system, or follow your specific rules: that’s when a custom solution, built on your data and your operation, makes sense.

That leap isn’t the first step, but the natural evolution once you know which use case gives you a return. That’s why it’s best to start simple, measure, and—when the value is clear—invest in something built for your business.

In summary

The U.S. data is compelling: AI adoption among small businesses rose from 48% to 68% in less than a year (QuickBooks/Intuit) and around 89% already use some AI tool (ICIC/Intuit). In Mexico the trend is rising, even if data isn’t yet as solid. For your business, this means AI is already small-business territory, and the advantage lies in starting well: one use case at a time, with data rules, measuring before and after, and scaling what works.

If you’d like to identify the first AI use case that would give your business the most return—and, when the time comes, build a custom solution on your own data—at Normandia Web we can help you take that first step with your feet on the ground.