Does ChatGPT favour big brands over small local businesses?

ChatGPT doesn't favour big brands across the board. In our Malaysian test, big brands didn't crowd small businesses out: ChatGPT named independents about as often as chains and big brands, and which side led depended on the trade.

In our test, three in four answers that named a business included at least one independent.

How often does ChatGPT name big brands instead of small local businesses?

ChatGPT named small local businesses and big brands about equally often in our test of Malaysian buying questions: 191 mentions to independents, 188 to chains and big brands. Trades differed a lot:

Mentions we coded "unclear" make up the rest of each row. "Chains and big brands" means a brand in three or more countries, a company listed on Bursa Malaysia, part of a listed group or a government-linked company, or three or more outlets in Malaysia. "Independents" are privately owned with one or two locations, including a few large single sites.

Why does ChatGPT recommend big brands in some trades and small businesses in others?

ChatGPT's mix of big and small names varied by trade in our test. Our reading is that it names whoever the pages it finds talk about, and KL renovation is mostly small contractors. In solar, one listed installer alone got 11 mentions.

The question can matter too. Asked for the best accounting firm in Malaysia for small businesses, ChatGPT gave about half its mentions to networks and multi-office firms. Asked for a KL firm to file tax for a Sdn Bhd, it named independents four times in five. That's one pair, not a rule.

Ask ChatGPT itself and it answers "Not inherently", then says it can still look biased because big brands show up more online.

How did small local businesses get into ChatGPT's answers?

Small local businesses in our test were usually named in answers that also cited their own website: a mention and a citation together. ChatGPT cited a page from an independent's own site in the same answer for 59% of their mentions, against 33% for chains and big brands.

Outside hotels the gap held, at 71% against 44% (CSV). Some were free Google Sites pages, mostly renovation contractors'.

Independent hotels were the exception: only 1 of their 34 mentions came with their own site cited, as hotel answers cite booking and review sites most.

Big brands were often named with no page of their own cited. Our reading: ChatGPT already knows them, or picks them up from other pages. It's a pattern, not proof a page causes a mention.

Are big brands steadier than small businesses in ChatGPT's answers?

Big brands were clearly steadier in ChatGPT's answers only for hotels in our test, where each question was asked three times. A big hotel brand came back at least twice 46% of the time, against 19% for independent hotels.

Outside hotels it went the other way, on small numbers: independents 19%, big brands 7%. Every list changes a lot between runs anyway.

What does other research say about AI and big brands?

Other research finds AI pulled towards famous brands, though less so for ChatGPT in one large study.

• Brand bias in the models. An EMNLP 2024 paper probed AI models about global and local brands. It found "a consistent pattern of bias", with the models "disproportionately associating global brands with positive attributes". It also found "country-of-origin effects which may boost local brand preference"¹.
• ChatGPT is less tied to brand fame, but big names still lead. Ahrefs studied 75,000 brands and found "ChatGPT shows the weakest correlations of all three AI assistants for almost every traditional brand authority signal". It also found "the same brands generally appear across all AI assistants"².
• Venues with a website were named more. A 2026 audit of 4,776 cafes, restaurants and bars in Bali found "85.6% of venues were never recommended by any system". Venues with "an own website" were more likely to be named³.

What can't our test tell me about ChatGPT and big brands?

• Each side's fair share. We didn't count how many big and independent businesses each trade has. If small firms far outnumber chains, half the mentions is still more than big brands' share.
• The app, other trades and dates. We used OpenAI's API, five trades, short keyword questions and one evening. The app, or food and retail, may differ.
• Borderline sizes. Two AI research agents sized each business separately and agreed on 181 of 197. Many independents were coded so because no branches turned up. Moving every "unclear" one way shifts the split by about six points.

How can a small local business get recommended by ChatGPT alongside the big brands?

In most trades we tested, ChatGPT's list wasn't reserved for the big names, and independents usually appeared next to a page of their own.

1. Request EasyFound's free check. See which businesses ChatGPT names when customers ask about your trade, then go through it with us on a free call.
2. Test the specific questions your customers ask, in English, Malay or Chinese, such as "Which accounting firm in KL can do tax filing for my Sdn Bhd?", not only "best accounting firm". Here's how to choose the questions.
3. Put your services, area and prices on a page of your own, even a free Google Sites page, written so ChatGPT can quote it.
4. If a big competitor keeps winning, check which page it was cited from and match that source.

EasyFound, an AI SEO service for Malaysian businesses, checks six AI tools in English, Malay and Chinese for your name beside the big brands, then works on the pages behind it. Start with the free check below.