While Bots Named Timmy Flood Your Feed, AI Search Is Getting Pickier About Who It Recommends

The Slop Problem And The Selectivity Problem

Three bots named Timmy, Ren, and Jackie are currently flooding Mastodon, Bluesky, and X with what Ars Technica described this September as slop-infused spam, all in service of promoting something called the iLands platform. This is the noisy side of AI and social media right now. Volume, repetition, agents posting around the clock to make a human-AI social system look alive.

Meanwhile, the quiet side of AI is doing the opposite. It's getting harder to get noticed, not easier.

The Found by AI test ran 200 queries across five AI assistants back in February. Only 8.4% of businesses showed up in any of them. Not the top result. Any mention at all, across five different tools.

That gap matters more than the bot spam does. Anyone flooding social platforms with agent-generated noise is optimizing for volume because volume used to work. But the AI assistants your customers actually ask for recommendations aren't rewarding volume. They're rewarding something else, and most businesses have no idea what that something else is.

You can post a thousand times a day and still not exist to ChatGPT. You can also barely post at all and show up consistently, if the signals underneath are right. The slop problem and the selectivity problem are happening on the same internet, at the same time, and they are not related to each other.

Ranking On Google Won't Save You Here

if you show up in the Google local 3-pack, you're covered. AI search will just follow along. It won't.

SOCi's 2026 Local Visibility Index found that ChatGPT recommends only 1.2% of multi-location brands. Google's local 3-pack recommends 35.9% of those same brands. That's not a small gap. That's two completely different filters looking at the same businesses and reaching almost opposite conclusions.

The overlap between the two systems sits at roughly 45%. Flip that around and think about what it means. More than half the time, a business that ranks well in Google's local pack does not show up when someone asks ChatGPT or Perplexity the same question. Your map pin doesn't transfer. Your review count doesn't transfer. Whatever got you into that 3-pack was built for a ranking algorithm, and AI search isn't running a ranking algorithm. It's making a recommendation, and it's being roughly 30 times more selective about who earns one.

If your entire local visibility strategy is "get into the map pack," you've optimized for one system out of several your customers are already using. The other systems are asking different questions entirely.

Gemini, Perplexity, And ChatGPT Don't Agree

Assume Gemini and Perplexity think alike, and you'll build the wrong list of things to fix.

Surfeo ran 13,094 queries in September across 30 industries comparing the two. Gemini recommended 31% of businesses. Perplexity recommended 26%. Five points sounds small until you remember these are two AI assistants looking at the same businesses and landing on different answers nearly a quarter of the time.

ChatGPT adds another wrinkle. Businesses it recommends average 4.3 stars. That's not a coincidence, and it's not a low bar either. ChatGPT appears to weight review quality more heavily than the other two, which lean on different combinations of content depth, structured data, and citation patterns scattered across the web.

So you've got three engines, three different filters, and no single move that satisfies all of them. A business optimized purely for star rating might do fine with ChatGPT and still miss Gemini. A business with strong technical content might clear Gemini's bar and fall short on Perplexity, which seems to weight source diversity differently.

None of this means the situation is random. It means "AI visibility" isn't one target. It's three separate scorecards, each grading on different criteria, and none of them are publishing their rubric. Chasing a single number is a losing game. Chasing the underlying signals — reviews, content depth, structured data — pays off across all three, just not evenly.

What A Local Business Should Actually Do This Week

Start with your reviews, not your website. ChatGPT's recommended businesses average 4.3 stars. If you're sitting at 3.6, no amount of schema markup or blog content closes that gap by itself. Pull up your Google Business Profile and your Yelp listing this week and count what's actually there. Respond to the recent negative ones. Ask your last five happy customers for a review today, not next quarter.

Then go ask the assistants yourself. Open ChatGPT, Gemini, and Perplexity and type the exact question a customer would type — "best plumber in [your city]" or whatever applies. Do this on all three, not just the one you happen to have open. Remember the Surfeo numbers: Gemini and Perplexity disagree on roughly a quarter of businesses. If you only check one, you're only seeing one-third of the picture.

Check whether an AI Overview shows up on your core local searches too. Early 2026 data puts AI Overviews on more than 30% of searches, and local queries trigger them even more often. If an Overview is answering your customer's question and you're not in it, that's lost visibility Google's regular rankings will never show you.

Treat this as its own audit, separate from your SEO checklist. Different scorecard, different homework.