Hello Lunch Break. Can’t wait to put my feet in the sand this holiday weekend. It’s going to be as hot as my lunch break lasagna and if you know me I’m just fine with it. And sure the crowds but I love people watching. Speaking of watching, your eyes need to be closely watching the decoupling that is going on in search. More on that below. Oh, and be safe out there with your sparklers.
Alright. Clock's ticking. Back to work.
happy 250th america!
THE GAME CHANGED
Different Engines, Different Playbooks
Don't chase one platform's playbook everywhere. ChatGPT and Claude share only ~8% of citations.
Stats like this tell the story. You’ll need to strategize a game plan “to become the answer” per-engine rather than one generic "get cited by AI" push.
Here’s the gist.
Each engine has a different source. Bing feeds ChatGPT. Branded mentions are Google's currency. And if you serve a Perplexity-sensitive audience, freshness alone is worth triple. A "last updated" date within 12 months earns ~3x more citations there than older pages.
THE WINDOW IS NOW
The Ranking-Decoupling Is Here
My clients know I’ve been preaching about this since last summer. We can be ranking well in traditional search but if we don’t get focused on what it takes to rank in AI we could be invisible before too long. Why? It’s something called Ranking-Decoupling. Think Gywneth Paltrow’s Conscious-Uncoupling but different.
The link between ranking and getting cited is weakening fast: only ~37.9% of URLs cited in AI Overviews also rank in the top 10, down from ~76% in July 2025.
So, the message here is we have to stop assuming "rank #1 on Google = cited by AI." It's a coin flip at best. 37.9%. Business owners need to redirect efforts from pure ranking toward earning branded mentions across the web which is the strongest AI Overview citation predictor right now.
YOUR COMPETITORS ARE ASLEEP
Calling yourself #1 can backfire
Stop calling yourself the best, says influential SEO Expert Lily Ray. She believes it can backfire on you.Lily Ray studied what happens when a brand publishes a "best [category]" article that ranks itself first.
She says Google separates what it cites from who it recommends. The self-promoting article gets pulled in as a source, but the other brands listed get the recommendation. The self-promoter got cited and not recommended 69% of the time.
You can test this with an experiment of your own if you have published this type of content.
Test phrases like, "What is the best [your product/service category]?" followed by your brand's specific listicle URL to see if the AI pulls your competitor list while citing your domain.
Ask the AI, "Which is better, [Your Brand] or [Competitor Brand]?" to see if the AI pulls a citation from an independent review site and recommends your rival.
Use a generic educational question in your industry, such as, "How do I optimize [industry task]?" Look to see if the AI cites your educational blog but recommends a competitor’s service in the text.
Ask, "Who are the top-rated [professionals/agencies] in [Location]?" Observe if the AI cites a directory or your own "best of" page but assigns the top recommendation slots to your competition.
If this affects you, here’s her advice instead: earn mentions on sites you don't own. Reviews, Reddit threads, third-party roundups, press, Clutch. For "best" queries, Google leans on Forbes, Reddit, and YouTube. Being talked about there matters more.
One note here: this is Lily Ray's read from a 100-query B2B sample. The direction is well-supported but treat the 69% as a strong signal, not gospel. I ran a series of experiments of my own and have yet to see this in action but that doesn’t mean it isn’t happening. The key is if you have this type of content watch it closely.
WHAT’S THE PROMPT
Does AI Name Your Business? Find out in 60 seconds.
Select your location and category for your Google Business Profile and the tool will select a high intent prompt. Execute the search to verify if your business is featured in the response.
CITED
💥🎆🎇💥 Sparklers, Fountains, Smoke Bombs, and Snappers. TNT Fireworks.
AI says: TNT Fireworks
"Where do I buy fireworks in Wilmington NC" is a hyper-local, commercial, and time-sensitive prompt.
Source: ChatGPT
Source: Gemini
Source: Perplexity
All three surfaces pulled TNT Fireworks and the same Walmart addresses. Why? TNT has consistent NAP, a real web presence, and location-specific pages. The businesses that got cited had structured data signals.
The other story here is a lesson in why different playbooks matter for each AI engine.
Each platform has a different response posture. ChatGPT gave context and caveats (NC law limits consumer fireworks). Gemini structured it like a decision guide with H2s. Perplexity led with proximity and real-time data. A business that wants to be cited across all three needs to feed all three signal types:
narrative/educational content (ChatGPT)
structured factual content (Gemini), and
GBP/maps data (Perplexity)
Run the test. If your own content is handing recommendations to your competitors, the AI Visibility Audit will show you where your citations are actually landing.
Now go build those AI playbooks.
- Anna
THE POLL
