How ChatGPT, Gemini and Perplexity Choose Which Local Businesses to Recommend

Featured image for the AI SEO Lift article

Ask ChatGPT for “a reliable roofer in Portland” and you will usually get a short list of three to five companies with a sentence or two about each. Ask Gemini or Perplexity the same question and the list may be similar, or it may be completely different. For business owners, the obvious question is: how do these tools decide who to recommend?

No AI company publishes a full recipe, and the systems change often. But by looking at how these tools work, what they cite and what their makers have said publicly, we can build a practical picture. This guide explains it in plain English and shows what Oregon businesses can do to improve their chances.

Step 1: The AI decides whether to search the web

Large language models are trained on huge amounts of text, but that training data has a cutoff date and is not reliable for specific local details like opening hours or phone numbers. So when a question involves local businesses, recent information or recommendations, most assistants run a live web search first.

  • ChatGPT uses its search feature, which draws on a web index built from several sources and partner data.
  • Gemini and Google’s AI Mode use Google Search and Google’s Maps and Business Profile data.
  • Perplexity runs its own searches and shows the sources it used for every answer.
  • Microsoft Copilot relies on Bing.

This is the first key insight: if search engines cannot find you, AI assistants usually cannot either.

Step 2: The AI reads a handful of sources

After searching, the assistant reads a small set of pages, often the top results plus some it judges especially relevant. For local queries, these typically include:

  • Business listings such as Google Business Profile, Yelp, Bing Places and the Better Business Bureau.
  • “Best of” articles and roundup lists from local publications or review sites.
  • Business websites with strong pages for the service and city in question.
  • Forums and community discussions, such as Reddit threads where locals share recommendations.
  • Review aggregators and industry directories.

Because the AI reads only a few sources, being present on several of the sites it tends to trust multiplies your chances. A business that appears in a local roundup, has a strong Google profile and a clear service page has three chances to be noticed.

Step 3: The AI weighs trust and agreement

When several sources agree about a business, the AI can recommend it with more confidence. If one site calls you the best emergency plumber in Gresham, your reviews mention fast emergency response and your website has a clear emergency plumbing page, the story is consistent.

Signals that appear to matter include:

  • Review volume and sentiment. Many positive, specific reviews are a strong signal. AI tools often summarize what customers praise.
  • Consistency. The same name, address, phone and services everywhere.
  • Specificity. Content that clearly connects your business to the exact service and location being asked about.
  • Freshness. Recent reviews and updated content suggest the business is active.
  • Reputation of the sources. Mentions on well-known, trusted websites carry more weight than mentions on spammy directories.

Step 4: The AI writes the answer

Finally, the model writes a response. It tries to be helpful and safe, so it tends to favor businesses that look established and well reviewed and avoids recommending businesses with mixed or unclear information. It may include a short description pulled from your website or reviews, so how you describe yourself matters.

How each assistant differs

Google Gemini and AI Mode

Google’s AI tools have direct access to Google Business Profile data, reviews and Maps. For local recommendations, a strong, complete profile with good reviews is especially important. Traditional Google rankings also strongly influence which websites get cited.

ChatGPT

ChatGPT’s results often draw on a mix of listings, review sites, local articles and business websites. Because Bing data has historically been part of its search sources, it is worth claiming and optimizing your Bing Places listing and checking your site in Bing Webmaster Tools.

Perplexity

Perplexity is very transparent: it lists every source it used. That makes it a useful testing tool. If Perplexity consistently cites a particular directory or article for your industry, getting listed or mentioned there is a clear opportunity.

What Oregon businesses can do today

  1. Strengthen your Google Business Profile. Accurate categories, services, hours, photos and steady reviews.
  2. Claim Bing Places and Apple Business Connect. Many businesses ignore these, but they feed other assistants.
  3. Build clear service and location pages. One page per service, with the city named naturally, and answers to common questions.
  4. Get mentioned on trusted local sites. Local news, chambers of commerce, “best of” lists, sponsorships and community organizations.
  5. Encourage detailed reviews. Reviews that mention services and neighborhoods give AI tools specific material to summarize.
  6. Add structured data. LocalBusiness schema helps machines read your business details accurately.
  7. Do not block AI crawlers by accident. Check your robots.txt and any security plugins that might block bots like OAI-SearchBot or PerplexityBot.

How to test what AI says about your business

Create a list of ten questions your customers might ask, such as “best dentist in Lake Oswego for kids” or “how much does a new roof cost in Salem.” Ask each assistant every month and record:

  • Whether your business is mentioned.
  • Which competitors are mentioned.
  • Which sources are cited.
  • Whether the information about you is accurate.

Over time, this shows whether your efforts are working and reveals which websites the AI trusts in your niche.

Common reasons businesses are left out

  • Very few reviews, or reviews that stopped years ago.
  • A website with one generic page and no specific service content.
  • Inconsistent business information across listings.
  • No presence on the third-party sites that AI tools commonly cite.
  • Technical problems that stop the site from being crawled or indexed.

Frequently asked questions

Does ChatGPT use Google reviews?

ChatGPT can read web pages that contain or summarize reviews, but how it accesses review data depends on its search sources. Strong reviews across several platforms are the safest strategy.

Can I ask ChatGPT to recommend my business?

Telling a chatbot to remember your business does not change what it tells other users. Recommendations come from what it finds on the web.

Are AI recommendations the same for everyone?

No. Answers vary by wording, location, time and sometimes the user’s history, which is why repeated testing matters.

Make your business AI-recommendable

Our AI search optimization service combines local SEO, reputation building and content designed to be cited by AI assistants. Start with our AI search (GEO) checklist or request a free SEO audit.