1.2% of Businesses Get Recommended by ChatGPT. Here's What They Do Differently

Slop Bots and the Trust Gap

Timmy, Ren, and Jackie showed up on social media in mid-September and started posting nonstop. They are AI agents, part of a startup pitching mixed human-agent platforms, and their entire output is slop. Generic captions, recycled takes, content built to fill a feed rather than answer a question. Nobody asked for this, and yet here it is, clogging timelines everywhere.

This matters more than it looks like it does. AI models are trained on the internet, and the internet is filling up with exactly this kind of noise. When ChatGPT or Perplexity has to decide which local plumber or bakery or law firm to recommend, it is sorting through an increasingly polluted signal pool. Slop bots do not build trust. They just add more junk to filter through.

That filtering problem is why so few businesses actually get recommended. As of April 2026, only about 1.2% of local businesses get surfaced by ChatGPT for local service queries. Not 1.2% of a niche. 1.2% of everyone competing for that spot.

AI models are getting pickier, not because they are smarter, but because they have to be. There is too much low-quality content and not enough verified signal. The businesses that get mentioned have figured out what counts as signal. Most have not.

Why Ranking First Doesn't Matter Anymore

For years the whole game was page one. Rank first on Google, get the clicks, get the customers. That playbook is dead for AI search, and most business owners have not caught up to that fact yet.

Google ranking has almost no direct impact on ChatGPT results. That is worth sitting with. The site you spent three years optimizing for backlinks and keyword density does not automatically carry weight when someone asks ChatGPT to recommend a plumber nearby. Different system, different inputs, different rules.

Part of the reason is that Google itself has changed underneath everyone's feet. As of April 2026, AI Overviews sit at the top of roughly 30% of Google searches. That summary box is answering the question before anyone scrolls down to the blue links. So even the ranking game you used to win is competing against an AI-generated answer sitting above it.

Two separate systems, two separate sets of rules, and neither one cares much about the other. Google ranking helps you inside Google's own AI layer somewhat, but it does not transfer to ChatGPT or Perplexity, which pull from entirely different data sources.

If you are still chasing rank one and calling it a visibility strategy, you are optimizing for a shrinking slice of how people actually find businesses now.

Who's Actually Feeding the AI

So who is ChatGPT actually pulling from when it makes a recommendation. Not Google. Not your website copy. Foursquare.

In April 2026, Foursquare supplied roughly 70% of the local business data ChatGPT uses to make recommendations. Most business owners have never logged into Foursquare in their life. Some still think of it as the app people used to check in at bars a decade ago. But that backend data layer never went away, and now it is doing the heavy lifting for one of the biggest AI models on the planet.

Perplexity runs on a different source entirely. A July 2026 analysis of over 28 million AI responses found Yelp holding roughly 62% of Perplexity's local citation share. So if you have been ignoring your Yelp profile because you decided reviews there did not matter as much as Google reviews, you have been ignoring the exact platform feeding a huge share of Perplexity's answers.

This is the part that trips people up. Each AI model has its own preferred data source, and those sources are not the ones businesses have spent the last decade optimizing. You cannot treat directory listings as an afterthought anymore. Claim them, fill them out completely, keep the hours and categories accurate. That is the data ChatGPT and Perplexity are actually reading.

Reviews, Not Ratings, Win the Mention

Most business owners obsess over getting to five stars. Wrong obsession.

Search Engine Journal published an analysis on September 10 covering 120,000 AI mentions across five models, and the finding cuts against what everyone assumed. Review volume predicts whether ChatGPT or Perplexity mentions you more than star rating does. A business with 400 reviews averaging 4.2 stars is beating a business with 20 reviews averaging 4.9 stars. That is not how the old system worked. That is not how most owners still think it works.

It makes sense once you think about what these models are actually doing. A handful of five-star reviews looks thin no matter how glowing. Hundreds of reviews, even with a few three-star complaints mixed in, reads as an established, verifiable business with a real transaction history. Volume is a trust signal. Perfection with no volume behind it is not.

So the task this week is not chasing your rating up half a star. Claim your directory listings on Foursquare and Yelp and make sure the hours, categories, and business name match everywhere. Add schema to your site so review content is structured, not just displayed. Then ask more customers to leave a review, any honest review, rather than quietly filtering out anyone who might leave four stars instead of five.