Small businesses that use AI expect stronger sales and hiring over the next year than those that don’t, according to a New York Fed survey published October 8.
If you run a store, that sentence will tempt you to buy more AI tools or hire fewer people. Hold off on both until you see what the survey measured. At Ecommerce Paradise we read these releases for one reason: to decide where a high-ticket store owner should put the next dollar and the next hour. This one measured owner expectations, not results, and 77% of the AI users in it reported no change in labor costs.
The researchers also found that the gap shrinks once you control for how healthy the business already was, and it is widest among firms in weaker financial shape. That detail matters if your store is lean. I’ll walk through the numbers, the competing surveys that put small-firm AI use at less than half the NY Fed’s figure, and the five moves I’d make before year-end. If you are still deciding what kind of store to build around AI, start with my guide to what high-ticket dropshipping is.
Form your LLC with Northwest Registered Agent before you scale anything else, including your AI stack. Northwest is the formation service I’d pick for an ecommerce owner because the address on an LLC filing is where plaintiffs and regulators look first, and a registered agent address keeps yours off the public record, unlike cheap formation sites that upsell you after checkout. Form your LLC with Northwest →
NY Fed: AI Users Expect +33 Net Hiring vs +15 for Non-Users
Will Aarons and Asani Sarkar of the Federal Reserve Bank of New York published the analysis on October 8 on the bank’s Liberty Street Economics blog. They used the 2025 Small Business Credit Survey, fielded September to November 2025. The AI module covers 5,248 employer firms, weighted to match the national population of employer firms.
Per the post, 46% of firms with at least one employee said they or their employees were using AI tools, and another 15% planned to adopt within twelve months. Among the users, 63% called AI somewhat or very important to production, and 51% had integrated it partially or fully into business processes. Then the two numbers that deserve the most attention: 77% reported no change in labor costs from AI, and 31% reported increased sales.
The headline comparison uses a net score, which is the share expecting an increase minus the share expecting a decrease over the next twelve months. The authors report the gaps below.
| Measure | AI users | Non-users | Gap |
|---|---|---|---|
| Employment, unadjusted | +33 | +15 | 18 points |
| Employment, with firm and owner controls | n/a | n/a | 13 points |
| Employment, controls plus past employment change | n/a | n/a | 10 points |
| Revenue, unadjusted | +48 | +21 | 27 points |
| Revenue, controls plus past revenue change | n/a | n/a | 14 points |
The gap is larger for firms in fair or poor financial condition. For them it is 21 points on employment and 24 on revenue, against 10 and 15 for firms in good condition, per the authors. Among AI users, firms that reported trouble using technology, a category the survey defines to include websites, social media, ecommerce and cybersecurity, had a net revenue expectation 16 points higher than users without that trouble.
The authors are careful about what this proves. They write that AI “may be helping some small businesses overcome specific operational or technical constraints,” and that the results do not reflect “broad-based optimism or expectations of expansion through hiring.” Whether adopters actually get the outcomes they expect, they say, “is a subject for future research.”
46% Versus 19.8%: Why Surveys Disagree on Small-Firm AI Use
The NY Fed’s 46% does not match the Census Bureau’s number, and the difference is large. The Census Bureau’s Business Trends and Outlook Survey put overall business AI use at 19.8% as of May 3, 2026, covering data collected December 14, 2025 to May 3, 2026. Firms with four or fewer employees were under 20%, firms with 250 or more employees were at 37%, and retail trade sat near 14%. Census also reworded its AI question in November 2025, so comparisons with older releases are not clean.
A Federal Reserve Board note from April 3, 2026 by Jeffrey Allen reached about 18% of firms at year-end 2025 and said “the BTOS is the best source for an estimate of the percentage of U.S. businesses that have adopted AI.” Allen’s FEDS Note attributes most of the spread between surveys to sampling, units of measurement and question wording. I can’t reconcile the two figures any better than that, and the sources don’t either, so treat both as ranges.
The same bank published a different cut on September 1. Its regional survey of New York and Northern New Jersey firms found 61% of service firms using AI, up from 40% in 2025 and 25% in 2024, and 51% of manufacturers, up from 26% and 16%. Layoffs tied to AI hit 4% of service firms, up from 1%. About 15% of service firms said they hired fewer workers than they otherwise would have, against 12% a year earlier, and about 13% hired more. The authors added that “these patterns could shift as adoption matures.”
There is precedent for the cautious reading. Bloomberg reported in September 2024 that AI adopters weren’t slashing jobs in an earlier NY Fed survey. Two years on, the pattern is still retraining and slower hiring more than mass cuts. I covered the same bank’s tariff research this week in my post on NY Fed tariff price effects, where the same discipline applies: read what the sample can and cannot say.
Counterpoint, stated plainly: the October 8 authors stress that these are expectations, not realized outcomes, and that the post does not establish causation. The earlier September data shows the opposite pressure is building at the margin, with more firms hiring fewer people. Both can be true at once. No source in this set measures ecommerce stores specifically, so any claim about your store is an inference.
What the NY Fed AI Gap Means for a High-Ticket Store Owner
This section is my read, not reported fact.
My read is that the survey describes confident firms picking up AI, and only partly describes AI creating confidence. Firms that were already growing adopt tools sooner, and the controls cut the employment gap from 18 points to 10 without erasing it. A gap that survives controls is worth taking seriously.
The part I’d act on is the finding about firms in weaker shape. A bigger gap among fair or poor firms fits the authors’ reading that AI relieves a specific constraint. For a solo high-ticket operator the constraint is usually hours: quote requests sitting overnight, product descriptions half written, supplier emails unanswered. A $200 buyer forgives a slow reply. A $6,000 buyer comparing three stores does not. My guide to AI tools for dropshipping sorts which of those jobs a tool handles well and which still need a person.
The 77% no-change-in-labor-cost number is the one I’d tape to the monitor. It says AI mostly added capacity instead of removing payroll. Here is hypothetical math, not a survey figure. Say a store sells $1.2M a year at a 20% net margin and spends $300 a month on AI tools, or $3,600 a year. At 20% margin the tools need to produce $18,000 in extra revenue to break even, which is 1.5% of sales. If the 31% of users who reported higher sales saw anything near that, the spend pays. If the tools only save you an hour a week and you spend it scrolling, they don’t.
Here are the scenarios I’d use, each tied to one constraint. If your constraint is response time, an assistant drafting first replies pays quickly. Claude does this job well. ChatGPT does too, and I’d pick one and stop shopping.
If your constraint is ticket volume, a helpdesk such as Gorgias is the layer to look at. Platforms will not fill that gap for you either, as the news that Amazon is cutting seller support staff shows.
If your constraint is supplier calls and meetings that never turn into action items, Otter solves the notes problem. My walkthrough on how to use Otter.ai covers setup.
The sales side is moving too. Meta is testing an AI ad agent, which I covered in Meta’s past-buyer exclusion rollout. Google is letting advertisers preview AI Max copy, and Shopify and Stripe signed onto Meta’s agent protocol, per my post on the Personal Agent Protocol.
If buyers start asking agents to shop, the store with faster, cleaner answers on a Shopify catalog wins the comparison. That is an inference, and nobody has published store-level data on it yet. Expect the first real numbers to come from the platforms, which have every reason to flatter their own tools, so read them with the same skepticism you’d bring to a vendor case study.
Employment is the other half of the headline, and I’d read it narrowly. The 10-point gap after controls says AI users feel better about hiring, not that they are hiring. The September regional data showed about 15% of service firms hiring fewer people because of AI and about 13% hiring more. For a high-ticket store with one to three people, the practical question is not whether AI replaces a hire. It is whether AI lets you delay the next hire until revenue justifies it. That is a real decision, and it is worth putting a date and a revenue threshold on before you make it.
One more point from the survey: the 16-point bump for AI users who struggle with website, social and ecommerce technology. That group is the new store owner. If that is you, the cheapest fix may not be another subscription. My team builds and runs the technical side so the owner can focus on sales, which is what the turnkey done-for-you service is for.
The Fed data says AI helps most where a constraint is binding. Want my team to build and run the store so your constraint is not the tech? See the turnkey done-for-you service →
Five AI Moves for High-Ticket Stores Before Year-End
I’d make these five moves, in this order, and I’d stop after any step that doesn’t move a number.
- Name one constraint before you buy a tool. Write down the slowest job in your week, whether that is quote replies, product copy or supplier follow-up. My list of best AI tools for small business is organized by job, so match the tool to that job. Test it with a month of Claude before you add a second paid assistant.
- Track one number that proves it. For most high-ticket stores that is time from quote request to first reply and the close rate on quotes. Log both before the change and after, in a HubSpot pipeline or a spreadsheet, so you are not relying on a feeling. My piece on summarizing supplier contracts and spec sheets with AI is a good first low-risk test.
- Put AI where buyers are already waiting. Post-quote follow-up, abandoned carts and win-back emails are repetitive and measurable. Audit your flows against my guide to the email flows every store needs. Then run them in Klaviyo, where a drafted subject line is easy to A/B test against your current one.
- Hire for hands, automate for drafts. If the constraint is physical work such as order tracking or supplier calls, a part-time assistant beats software. OnlineJobs.ph is where I’d start. My comparison of remote job sites covers the alternatives. Keep the hire small, because 77% of AI users in the survey did not cut labor costs, and you probably will not either.
- Know your margin by SKU. You can’t judge a $3,600 tool budget without clean books. Connect an ecommerce bookkeeping tool like Finaloop and review gross margin monthly. If you want a second set of eyes on the numbers, Explore Turnkey Builds and we’ll go through them together.
Frequently Asked Questions
Does this NY Fed survey prove AI raises sales?
No. The authors say it shows an association in owner expectations, and that whether adopters get the outcomes they expect is “a subject for future research.” Treat it as a reason to run a controlled test in your own store, not as a forecast.
Why do surveys show 19.8% and 46% of small firms using AI?
They measure different things. The Census Bureau survey covers all businesses and reworded its question in November 2025, while the NY Fed survey covers employer firms with at least one worker. A Federal Reserve Board note attributes most gaps to sampling, units of measurement and question wording.
Should I cut staff because AI users expect to hire less?
The data points the other way. In the survey AI users were more likely than non-users to expect higher employment, and 77% reported no change in labor costs. Retraining and slower hiring are showing up more than layoffs, per the NY Fed’s September post.
What is the cheapest AI test for a new store?
Draft replies and product copy with an assistant for 30 days and track reply time. If you have no niche yet, grab my free high-ticket niches list first. The full high-ticket niches breakdown adds context.
Is it worth hiring an agency instead?
Sometimes. If your constraint is build quality and not hours, read my breakdown of how to choose an ecommerce agency and when a done-for-you build beats one.
Where do I start if I have no store yet?
Pick a supplier model first, then add tools. A supplier marketplace such as Spocket helps you find products. My beginner guide walks you through the launch order.
Want to work through your AI stack with other store owners, and with me, instead of guessing? Join the Skool community →
That is the read on this one: useful signal, thin proof, cheap to test. Subscribe to the YouTube channel for daily breakdowns. More breaking news coming through the day.
Related Articles
If this was useful, these go deeper:
- Shopify Canvas Lets Sidekick Redesign Your Whole Store
- Shopify Now Writes and Updates Your Store Policies
- Stripe and FedEx Plan Shipping-Data Loans for SMBs
- ASOS Breach Hit Its Customer Messaging Platforms
- Claude for Ecommerce: How to Build a Leaner High-Ticket Business With AI
Trevor Fenner is an ecommerce entrepreneur and world traveler from Seattle, Washington, and the founder of Ecommerce Paradise, a platform focused on helping entrepreneurs build and scale profitable high-ticket ecommerce and dropshipping businesses. With over a decade of hands-on experience, Trevor specializes in high-ticket dropshipping strategy, niche and product selection, supplier recruiting and onboarding, Google & Bing Shopping ads, ecommerce SEO, and systems-driven automation and scaling. Through Ecommerce Paradise, he provides free education via in-depth guides like How to Start High-Ticket Dropshipping, advanced training through the High-Ticket Dropshipping Masterclass, and fully done-for-you turnkey ecommerce services for entrepreneurs who want a faster, more hands-off path to growth. Trevor is known for emphasizing sustainable, real-world ecommerce models over hype-driven tactics, helping store owners build scalable, sellable, and location-independent brands.
He maintains homes in Seattle and Bali, spends most of the year overseas, and usually returns to Seattle for about a month each summer.
