At a Glance
Some customers are using ChatGPT and other AI tools to find and evaluate businesses, but usage varies by industry, buying situation and audience. The right question for a small business isn't whether AI search is growing; it's whether prospective customers use it during decisions that could produce qualified enquiries or sales.
AI-assisted business discovery is part of a broader research journey that can also involve Google, reviews, referrals and company websites.
TL;DR
- Research confirms meaningful AI adoption, but studies measure different populations, questions and behaviours; their percentages aren't interchangeable.
- ChatGPT may help buyers explore a problem, compare providers or investigate products without being their final verification channel.
- Prioritize AI visibility when your customers have research-heavy decisions, complex choices or clear reasons to compare providers.
- Use actual customer questions, source checks, enquiry conversations and a repeatable sample of AI searches to test relevance.
- If your existing website or follow-up process is weak, fix the commercial bottleneck before purchasing an expensive new visibility package.
Your customers are using ChatGPT.
Well, some of them are.
And that's precisely where the conversation should begin — because there's a rather large difference between people using AI and your next customer using ChatGPT to choose a business like yours.
The distinction matters when an agency is proposing yet another monthly retainer for “AI visibility.”
It isn't that the opportunity is imaginary. Consumer research suggests it's real. But the research also shows that customers continue to use search engines, reviews and other sources to check what AI tells them.
If you run a small business, you don't need another technology trend to admire from a distance. You need to decide whether the trend deserves a place in your marketing priorities.
Here's a practical way to make that decision.
Are People Really Using ChatGPT to Search for Businesses?
Yes, but the answer depends heavily on what researchers mean by “using AI,” “searching,” and “finding a business.”
Consider four pieces of primary research published in 2026:
| Research | What Was Measured | Finding | What It Does Not Prove |
|---|---|---|---|
| Pew Research Center, February 2026 | U.S. adults using chatbots to search for information | 42% reported doing so | That 42% used them to choose a local provider |
| Forrester, February 2026 | U.S. online adults using ChatGPT to look for products they might buy in the previous month | 26%; Google and Amazon remained more widely used for product discovery | That ChatGPT was used for service-provider research |
| BrightLocal, 2026 | U.S. consumers' most recent search for a local business | 23% used AI at some point; 8% started with it | That AI was the only or final source consulted |
| Gartner, June–July 2025 survey | U.S. consumers' attitudes to generative AI versus search engines | Only about one-third considered chatbots as effective as search engines for learning new information | That generative AI has no role in buyer research |
The Pew study surveyed 5,119 U.S. adults. The Forrester consumer snapshot concerns shopping, not all business discovery. BrightLocal's local search research addresses a different use case again. And Gartner's analysis looks at perceived effectiveness rather than how many business referrals AI generates.
Different samples. Different questions. Different outcomes.
You cannot take the largest percentage, paste it into a proposal and claim that nearly half of your prospects are now choosing suppliers through ChatGPT.
Well, you can. Whether you should is another matter.
Why the Distinction Matters for Small Businesses
Imagine a company selling specialist industrial cybersecurity assessments. A relatively small number of qualified buyers may matter far more than thousands of general consumers casually experimenting with AI.
Now imagine a neighbourhood coffee shop. Its buyers may still rely heavily on maps, proximity, opening hours, reviews and recommendations from friends.
Both could appear in AI-assisted searches. But the commercial opportunity, urgency and measurement approach are different.
The relevant unit isn't “AI users.”
It's buyers researching the kind of problem your business solves.
Where Does ChatGPT Fit in the Customer Journey?
ChatGPT can function as a research assistant, comparison tool or product-discovery interface. It doesn't necessarily replace the places where customers verify information and complete purchases.
OpenAI's official shopping guidance says ChatGPT can show relevant products, compare trade-offs and provide merchant links when a question indicates shopping intent. It also warns people to verify price, availability and policies with merchants before buying.
That illustrates the nature of the experience: assistance with narrowing decisions, not a guarantee that every recommendation is complete or correct.
The behaviour is not limited to shopping. A business buyer might use ChatGPT to prepare a supplier shortlist, ask what questions to put in a request for proposal, or clarify technical terminology before contacting anyone.
The conversation often develops in stages.
First: Help me understand the problem.
Then: What options are available?
Then: Which option seems suitable for my situation?
Finally: Can I trust the provider, and what happens if I choose them?
Traditional search, professional referrals and company websites may appear at several of those stages.
BrightLocal found that around three-quarters of consumers used more than one channel in their most recent local-business search. Its July 2026 research also found that many AI users returned to Google or another source to verify the information. See BrightLocal's original findings and methodology.
So a business can benefit from AI-assisted discovery without AI being the last touch before a sale.
And the reverse is also true: an AI mention that never creates interest in the business may have little commercial value.
Which Kind of Customer Are You Trying to Reach?
Before you invest, identify the circumstances in which a buyer would plausibly ask an AI assistant about your category.
Start with three questions:
Does the Purchase Require Explanation or Comparison?
A buyer arranging an emergency locksmith often has a location, availability and urgency problem.
A buyer selecting a penetration-testing consultancy faces more complex decisions: scope, credentials, assessment objectives, commercial terms and provider suitability.
That second situation offers more opportunities for detailed explanatory content during research. It doesn't prove the buyer will use ChatGPT, but it gives a plausible reason they might.
Are Customers Researching Before They Know Your Service Name?
A homeowner may describe recurring damp patches rather than ask for a named remediation service.
A business owner may ask why qualified website leads aren't becoming customers rather than search for “conversion rate optimization.”
Those are problem-led questions. An answer engine can help a buyer identify possible causes and types of provider to investigate.
If your marketing only describes the service you sell, it may miss people who haven't yet learned what to call the problem.
Does Your Business Need to Prove Expertise Before a Buyer Contacts You?
For higher-consideration purchases, provider evaluation matters.
A prospective customer may want evidence of relevant experience, clarity about who is suitable, realistic limitations and explanations of how the work is delivered.
That's an opportunity for useful content—whether discovered through ChatGPT, Google or a colleague's recommendation.
A business with little credible information online has a broader discovery problem, not merely an AEO problem.
A Four-Path Decision Guide for Small Business Owners
Use the following router to determine how much attention AI discovery deserves right now.
| Your Situation | Recommended Path | Why | First Move |
|---|---|---|---|
| Customers research complex problems and compare providers before contacting you | Investigate AI-assisted discovery actively | The research journey offers multiple opportunities for credible expertise | Test high-intent questions and review your buyer guides |
| Customers mostly need a nearby, available provider immediately | Strengthen local SEO and business information first | Accurate location, contact details, reviews and availability are essential regardless of channel | Audit service areas, mobile experience and local profiles |
| You sell products people compare by technical attributes, price or suitability | Improve product data and test AI shopping research | Selection depends on specific, accessible attributes and current details | Review product descriptions, specifications and available merchant data |
| You already attract qualified enquiries but convert too few | Fix conversion and sales follow-up before spending heavily on AI visibility | Additional exposure won't repair a weak buying process | Investigate lost enquiries and response times |
These are prioritized paths, not mutually exclusive categories. A business can have a strong local presence and also benefit from research-led content.
The point is to choose the first commercially defensible action.
Path 1: The Research-Heavy Service Business
If your buyers need advice before they can even specify the service, AI discovery deserves active investigation.
Consider a cybersecurity consultancy, commercial insurance adviser or specialist engineering supplier.
The most useful work may be a clear comparison guide, an explanation of service scope or a decision framework grounded in firsthand expertise.
Start by testing a handful of realistic buying questions across ChatGPT Search and other relevant experiences. Record whether you are mentioned, what sources are used and whether the explanations are accurate.
Then compare those findings with what your service pages already explain.
For the broader technical foundations, see our practical guide to Answer Engine Optimization.
Path 2: The Urgent or Geographically Constrained Business
If people mostly choose you based on proximity, immediate need and basic trust, local discoverability comes first.
A plumbing business should not ignore AI, but it also shouldn't sacrifice accurate business listings, clear emergency contact options and mobile usability to chase a new reporting metric.
BrightLocal found that only 8% of respondents started their most recent local-business search using AI, while 23% used it somewhere in the journey. That's a material distinction between an opening channel and a supporting one. BrightLocal, July 2026.
Keep your services, geography and contact information clear and consistent. Then test AI results for inaccuracies or meaningful omissions.
Path 3: The Product Comparison Business
If customers compare features, specifications and alternatives, strong product information is a sensible foundation for AI discovery.
Forrester reported that 26% of U.S. online adults used ChatGPT in the previous month to search for products they were interested in buying in February 2026, while Google and Amazon remained more widely used for product discovery. Forrester's data snapshot.
That doesn't mean you should rewrite your catalogue for a chatbot. It means buyers may investigate products in additional environments.
Get the fundamentals right: accurate specifications, benefits and limitations, real pricing, availability and clear product distinctions.
And make sure the checkout experience works.
Path 4: The Business With a Bigger Conversion Problem
If customers already find you but don't become customers, diagnose conversion before funding a major AI search programme.
Perhaps prospects submit forms and hear nothing for two days. Perhaps the site doesn't explain what happens after a consultation request. Perhaps the sales team has no consistent follow-up process.
AI visibility might create more interest. But the immediate bottleneck is what happens to that interest.
This is the gap I would ask every business owner to consider.
What you ask for might be more traffic.
What you want is more business.
They're not interchangeable.
How to Find Out Whether Your Own Customers Use ChatGPT
Industry surveys provide context. Your own customers provide better evidence of whether the behaviour matters to your business.
You won't get a complete answer from any single method, so combine three modest investigations.
Listen to Real Sales and Customer Questions
Review the questions arriving through enquiry forms, calls, email and sales meetings.
What are people trying to decide? Which comparisons do they ask about? What misunderstandings delay the purchase?
Ask new customers a neutral question such as “How did you research providers before contacting us?” rather than “Did you find us through ChatGPT?”
The first invites a useful account of the journey. The second risks putting words in their mouths.
Record replies in your CRM or an ordinary spreadsheet. Avoid turning a dozen anecdotes into a market-wide statistic, but use the patterns to improve content.
Test Five Research Scenarios, Not Five Brand-Name Searches
Build a short sample around actual buyer intent:
- Help me understand a problem: “What could cause [the customer's actual issue]?”
- Choose an approach: “Should I use [approach A] or [approach B]?”
- Compare providers: “What should I look for when choosing a [provider category]?”
- Find local fit: “Which kinds of companies handle [need] in [service area]?”
- Verify credibility: “How can I check whether a [provider] has the right experience?”
Keep the prompts and dates. Record whether the answer cites your business, names relevant alternatives, contains factual inaccuracies or identifies questions your website fails to answer.
This is exploratory research, not a valid statistical estimate of your visibility across all users. Responses depend on context, tools and retrieval conditions.
A repeatable sample can identify worthwhile problems. It doesn't establish an immutable AI ranking.
Look for Commercial Signals Without Pretending Attribution Is Perfect
Track identifiable referrals where available, relevant landing-page engagement and qualified enquiries.
Someone may first encounter a company through ChatGPT but return later through Google, an email referral or a direct visit. That can obscure the original influence.
Gartner's 2026 research announcement highlights how generative tools can add steps to research rather than simply replace existing search behaviour.
So report what you can actually observe and keep causal claims modest.
A citation is not a lead. A lead is not a customer. And a dashboard that draws a neat line between the three may be giving you considerably more confidence than evidence.
Three Traps That Make Small Businesses Overspend
Mistake one: Treating a general AI adoption statistic as your buying audience. A national survey may show millions of people use chatbots. It doesn't tell you how many are shopping for your service in Chilliwack, Vancouver or Leeds.
Mistake two: Confusing a recommendation with a sale. Buyers still check reviews, websites, product details and other sources. A mention is at best one potential step in the journey.
Mistake three: Paying for output instead of removing constraints. An agency may offer twenty new AI-optimized articles when your best next move is improving three important service pages, repairing enquiry follow-up or documenting genuine experience.
The antidote is the same in each case: ask what evidence connects the proposed work to a real customer decision.
What If Your Customers Aren't Using ChatGPT Yet?
You don't have to manufacture an urgent investment just to keep up with the conversation.
Keep a sensible watch on how your audience researches. Maintain a technically accessible, useful website. Answer customer questions accurately. Make sure your business information stays current across important channels.
Those activities have value even if ChatGPT never sends you a lead this month.
If the evidence changes—more customers mention AI-assisted research, relevant queries surface your competitors, or AI referrals become commercially meaningful—revisit your priorities.
Not every trend requires a new department.
Sometimes it requires a better question.
Should You Invest in AEO Right Now?
If your customers make research-heavy decisions and the online information about your business is weak or incomplete, investigate AEO now as part of your existing SEO and content work.
If your primary problem is local findability, product-data quality, conversion or slow follow-up, address that constraint first and monitor AI-assisted discovery alongside it.
The right answer comes from your customers' behaviour and the next improvement most likely to help them choose your business—not a vendor's need to sell a new package.
For context on how the disciplines fit together, read SEO vs. AEO vs. GEO: Do Small Businesses Really Need Three Strategies?. For the broader fundamentals, start with our SMB guide to AEO.
Start with five genuine buyer questions, ask a few recent customers how they researched, then investigate one gap worth fixing. That's enough to inform the next decision without pretending you've solved the entire AI search market.
And if someone assures you that every customer is already choosing suppliers through ChatGPT?
Ask them how many of your customers they interviewed.