Type your company name into ChatGPT and it will probably describe you accurately. Ask it who to hire for your service in your city, and the same engine will likely leave you off the list. A study by Cheers, published November 1, 2026, tracked 455 businesses across 361 cities and found that ChatGPT, Gemini and Perplexity described a company in 96.1% of by-name answers but recommended it in only 39.7% of buyer answers.
That gap is the whole game. Being known by AI is not the same as being recommended by AI, and the difference matters for every business owner in Chandigarh, Mohali and Panchkula who is counting on AI-driven leads.
Two Different Questions, Two Different Answers
A by-name question tests whether an engine can find you and describe you. A buyer question, the kind a customer types without naming anyone, tests whether the engine picks you out of everyone it could name. The Cheers study asked both, using developer APIs with web search, not the consumer apps.
ChatGPT on its own described businesses in 94.9% of by-name answers but named them in only 39.3% of buyer answers. Among 325 home-service companies, the gap was wider: 97.5% by name, 35.8% when the question was who to hire. Roofers had it worse, appearing in just 18.3% of buyer answers.
107 Businesses Were Never Named
Of the 455 businesses, 107 appeared in none of their 12 buyer answers. These were not strangers to the engines. At least one engine described all but one when asked by name, and all three described 90 of them. The engines knew who they were. They just never put them on the shortlist.
The median business appeared in 4 of 12 buyer answers. That means half the businesses showed up in a third or fewer of the answers. If you are a local business owner who checked ChatGPT once, saw your name, and assumed you were covered, you may be in this group without knowing it.
Every Shortlist Comes With a Crowd
One business’s 12 buyer answers named about 40 other businesses on average. The businesses that were never named faced even bigger crowds, with 47.5 other names on average, against 37.7 for those named at least once.
The engines rarely agreed on those rivals. Of 18,182 business and competitor pairs, 81.1% of competitor names came from one engine only, and just 4.3% came from all three. When you write down who is named instead of you, the names that appear on two or three engines are your real competition for that question. No single engine represents the whole market.
Why This Matters More This Fall
In October 2026, ChatGPT switched to a building-block layout that includes phone numbers in local answers. Sistrix found the layout appeared in 97% to 100% of sampled answers from October 23. Before the switch, 12.6% of US ChatGPT local answers carried a phone number. After October 9, that jumped to 92.5%.
Google and Meta also have AI that can place calls. Google’s price-check help page says “Google will call multiple service providers near you,” and Gemini Call for Me is in testing for US Pixel 11 owners. Meta’s Muse can place calls to US businesses in a beta that expanded in September.
Customers are already asking AI for recommendations. BrightLocal’s 2026 survey found 45% of US adults had used AI for local business recommendations in the past 12 months, up from 6% in 2025. And 97% of those users double-check AI picks against reviews at least sometimes.
How to Run the Shortlist Test This Week
You do not need a paid tool for this. Ask four plain buyer questions without your company name, swapping in your trade and town:
- “Who’s the best [trade] company in [town]?”
- “My [specific problem] needs fixing, who should I call in [town]?”
- “Which [trade] companies in [town] are fair on price?”
- “Who’s best for [specific service] in [town]?”
Put each one to ChatGPT, Gemini and Perplexity. That is 12 answers. Count how many name you, and write down every business named instead, engine by engine. Re-run it monthly, because the lists move week to week.
For a business with several branches, run the test per town. Roll it up as one number per market in your monthly review: named in X of 12 buyer answers.
What to Do If You Are Not Being Recommended
If the test shows you are missing from the shortlist, the work falls into three areas:
- Your website: When an answer named a business, it cited the business’s own website 63% of the time. Make sure your service pages clearly answer the buying question for your city. A generic homepage is not enough.
- Your reviews: Reviews feed AI recommendations. Ask satisfied customers for reviews on Google, and make sure the review text mentions the service and location.
- Your listings: Directories and lists are where AI engines build shortlists. Make sure your business appears on the directories that matter for your industry and area.
This is the core of AEO and GEO work: making your business the answer AI engines give when buyers ask. Our SEO services cover the website and content side, and we can tell you more about our approach on a call.
Want to know where you stand? Contact VishBill for a shortlist test in your market.
Frequently Asked Questions
Why does ChatGPT know my business but not recommend it?
They are different questions. ChatGPT described businesses in 94.9% of answers when the question named the business, but listed it in only 39.3% of buyer answers. A buyer answer listed about six businesses, and one business’s 12 answers named about 40 rivals, so an engine that knows you has plenty of others to pick.
How do I check whether AI recommends my business?
Ask what a customer would ask, without your company name, such as “who’s the best furnace repair company in [town]?” Ask it four ways in ChatGPT, Gemini and Perplexity, count how many of the 12 answers name you, and write down who is named instead. Re-run it monthly.
Do AI answers say negative things about local businesses?
Rarely. In October 2026, only 29 of 70,624 business mentions in US AI answers carried a negative tone label. Engines do not put a business on a list to warn people off it. They leave it off. Complaints come up when a question names the business and asks what customers say.
What percentage of people use AI for local business recommendations?
BrightLocal’s 2026 survey found 45% of US adults had used AI for local business recommendations in the past 12 months, up from 6% in 2025. Of those users, 97% double-check AI picks against reviews at least sometimes.
Does this study cover Google Maps or AI Mode?
Google Maps rankings are not part of the Cheers study. Google AI Mode appears only in the tone data, which comes from a separate panel. The recommendation study covers ChatGPT, Gemini and Perplexity through their developer APIs.