There's a number in Goldman Sachs' new small business survey that I can't stop thinking about.
Not the good one. The weird one.
93% of small business owners using AI say it has had a positive impact on their business.
That's not a typo. Ninety-three percent. That's closer to unanimous than any business survey has a right to be.
And yet.
14% say AI is fully embedded in their core operations.
So here's the math problem Goldman Sachs handed us: nearly all of the small business owners who use AI think it's working. And almost none of them have made it a real part of how their business runs.
Those two numbers do not add up. And that gap is the most important story in small business right now.
What the Survey Actually Found
Goldman Sachs 10,000 Small Businesses Voices - the firm's long-running initiative tracking Main Street sentiment - published these findings this week. The sample size is substantial and the respondents are actual operating small business owners, not aspirational entrepreneurs or corporate buyers. This is a credible dataset.
Here's the full picture:
- 76% of small business owners report currently using AI
- 93% of those users say AI has had a positive business impact
- 84% cite increased efficiency and productivity as the primary benefit
- 67% expect AI to increase their revenue
- 87% say AI augments rather than replaces employees
- 14% say AI is fully embedded in their core operations
- 73% say they would benefit from more training and implementation resources
- 85% support the proposed "AI for Main Street Act," which would direct the Small Business Administration to expand AI education
Let's sit with a few of those numbers.
The 93% Number Is Suspicious (In a Good Way)
When 93% of any group agrees on anything, you should be curious about what you're actually measuring.
In this case, I think the number is real - but it's measuring something more modest than "AI transformed my business." It's measuring: did this make something easier than it was before?
And the answer is yes, almost universally. Because that bar is genuinely not that hard to clear. ChatGPT helps you write a first draft faster. Canva helps you make something look professional. An AI phone assistant means you stop missing calls at dinner. These are real benefits. They're also entry-level benefits.
The 93% number tells us that AI tools, used even lightly, tend to deliver some positive value. That's meaningful. It's also not the same as saying AI is a strategic asset for these businesses.
The 14% Number Is the Honest One
Now look at the other end.
Only 14% say AI is fully embedded in their core operations.
This is the number that matters if you're trying to understand where small businesses actually stand in 2026 - not where they want to stand, not how they feel about AI, but how deeply it's actually changed how they work.
14% is not a majority. It's not even close. It means that among small business owners who are using AI and feeling good about it, 86% are still in a version of "I use it sometimes for some things."
That's not failure. But it's also not transformation.
There's a term for this stage in technology adoption research: surface integration. You've added the tool. You haven't changed the workflow. The tool helps when you remember to use it. When you're busy or stressed, you go back to what you always did. The tool never becomes infrastructure.
Surface integration produces the 93% satisfaction number. Full integration is what produces compounding returns. Most small businesses are still at step one.
Why the Gap Exists
The survey asked about this directly. The barriers small business owners cite for deeper AI integration:
- Lack of technical expertise - they don't know how to set tools up properly
- Difficulty choosing tools - the market is overwhelming and the pitches all sound identical
- Data privacy concerns - they're not sure what's safe to put into AI systems
- Time to learn - they're already running a business; adding "figure out AI" to the task list is hard
None of these are shocking. They're the same barriers that slowed adoption of every previous business technology - from accounting software to cloud storage to credit card readers. The pattern is familiar.
What's different this time is the speed. AI tools are evolving faster than any previous business technology category. By the time a small business owner finishes learning one tool, that tool has been updated and three new competitors have launched. The learning curve doesn't flatten because the target keeps moving.
The Training Gap Is an Opportunity
73% of small business owners say they'd benefit from more training and implementation resources.
This is a large number that mostly goes unnoticed. People focus on the adoption numbers - how many businesses are using AI, which categories are growing fastest. The training gap is less interesting to cover, so it gets less coverage.
But if you're a small business owner trying to understand where you are in this landscape, the training number is the one that's actually relevant to you.
Because here's what it tells you: you're not behind because you lack ambition or resources or sophistication. You're behind - if you are - because the support system for small business AI adoption hasn't been built yet. The tools came first. The education is still catching up.
The "AI for Main Street Act" - which 85% of survey respondents support - would direct the Small Business Administration and Small Business Development Centers to provide structured AI education specifically for small business owners. It hasn't passed as of this writing, but it has the kind of bipartisan small business appeal that tends to move through Congress eventually.
What This Means Practically
If you're a small business owner looking at these numbers and trying to figure out what to do with them, here's my honest read:
The 93% is permission to experiment without pressure. The tools work. You don't need to do extensive research before trying something. Pick one thing you find annoying in your week and try an AI tool on it. The odds that you'll find it helpful are very high.
The 14% is a realistic goal, not a rebuke. Full AI integration takes months, not weeks. It requires choosing specific workflows, building habits, and going through the awkward middle period where the tool sort of works but you're still adjusting. Almost nobody is there yet. You're not supposed to be there in a few months of trying.
The 73% training gap is your actual to-do item. The biggest thing holding small business owners back isn't skepticism - they're overwhelmingly positive about AI. It's implementation knowledge. Find one good resource (a course, a consultant, a community) that focuses on practical implementation for businesses your size. That's more valuable than adding another tool.
The Goldman Sachs survey is useful because it's honest about where we actually are. Not the hype version, where AI has already transformed Main Street. Not the fear version, where small businesses are getting left behind.
The realistic version: AI is helping, broadly and meaningfully. Most businesses haven't figured out how to make it foundational yet. That's okay. That's where we are in April 2026.
The gap between 93% and 14% isn't a failure story. It's a roadmap.
Source: Goldman Sachs 10,000 Small Businesses Voices survey, published March 2026. Goldman Sachs pressroom.
Priya Kapoor covers the numbers behind small business decisions for The Useful Daily.