We took four well-regarded Arlington restaurants — Green Pig Bistro, Ruthie's All-Day, Ambar Clarendon, and Palette 22 — and tested how each one appeared when we asked ChatGPT, Claude, Perplexity, and Gemini to recommend places to eat in Arlington and Clarendon.
The gap between the most visible and least visible was wider than we expected.
The two questions we asked
We focused on the queries a real diner would use:
- "What are the best restaurants in Arlington?"
- "What are the top dining spots in Clarendon?"
Both are genuine discovery queries — the kind someone asks before they've decided where to go. We recorded which of our four restaurants appeared in each response, across each model.
What each model said
"Best restaurants in Arlington":
ChatGPT: "Check out The Liberty Tavern for delicious American fare, or go to Ambar for some fantastic Balkan cuisine."
Claude: "Ambar for outstanding Balkan small plates... Pupatella for some of the best Neapolitan pizza in the DC area."
Perplexity: Lists Ambar Arlington in its top picks, and separately cites the Washingtonian's best-of list which includes Green Pig Bistro and Ruthie's All-Day.
Gemini: "Consider Green Pig Bistro for upscale gastropub fare, Ambar for delicious Balkan small plates and a lively atmosphere, or The Liberty Tavern."
"Top dining spots in Clarendon":
ChatGPT: "Ambar for Balkan cuisine with a modern twist."
Claude: "Ambar is also a standout for unlimited Balkan small plates at a fixed price."
Perplexity: "The most consistently recommended picks are The Liberty Tavern, Ambar, Lyon Hall, CIRCA at Clarendon..."
Gemini: "For a unique all-you-can-eat Balkan small plates experience, be sure to visit Ambar."
The result
| Restaurant | ChatGPT | Claude | Perplexity | Gemini | |---|---|---|---|---| | Ambar Clarendon (56) | ✓ both queries | ✓ both queries | ✓ both queries | ✓ both queries | | Green Pig Bistro (48) | — | — | ✓ Arlington | ✓ Arlington | | Ruthie's All-Day (2) | — | — | — | — | | Palette 22 (45) | — | — | — | — |
Ambar appeared in every single response, for both queries, across all four models. Palette 22 didn't appear in any of them.
Why Ambar wins
Ambar's AI visibility isn't an accident. A few things are working in its favour that the other restaurants don't have to the same degree.
A specific, memorable concept. "Unlimited Balkan small plates" is a distinctive identity that's easy for a model to recommend for a specific occasion or preference. Every model described Ambar the same way — it has a clear, consistent reputation that's been written about enough times that the description is settled. "Upscale gastropub" or "gallery restaurant" are harder to own.
Editorial coverage in authoritative outlets. Ambar appears in Washingtonian dining coverage, local roundups, and is consistently cited across review platforms. The models aren't guessing — they're retrieving from sources that have mentioned Ambar enough times that the recommendation is confident.
High review volume. Ambar's Yelp and Google presence is substantial. Perplexity in particular relies on live retrieval, and a restaurant with hundreds of reviews on multiple platforms gets surfaced reliably.
Why Palette 22 is invisible
Palette 22 has real customers and good reviews, but its AI footprint is thin. It didn't appear in a single discovery query across four models — not because it's a bad restaurant, but because:
- Its concept (a gallery space serving food and cocktails) is harder to categorise, which makes it harder to recommend for a specific query
- It hasn't appeared in the editorial roundups that the models are drawing on for "best of Clarendon" answers
- Its footprint on review aggregators is smaller than Ambar's
Claude did mention Palette 22 once — a late-position reference for "top dining spots in Clarendon" — but that's the extent of its category visibility. For a visitor to Arlington asking where to have dinner tonight, it effectively doesn't exist.
The Ruthie's result
Ruthie's scored a 2 — effectively invisible, even with location-specific breakfast and brunch queries. That's a surprise given how well-loved the restaurant is locally, and it illustrates something important: real-world reputation and AI visibility aren't the same thing.
Ruthie's does appear in some AI responses when the query is very specific — Perplexity cited it as the top vote-getter in ARLnow's readers' choice for Arlington breakfast spots when asked about that exact list. But that's a narrow window. Across a broader range of breakfast, brunch, and dining queries, it doesn't surface. Strong local word-of-mouth hasn't translated into the kind of AI-retrievable footprint that produces consistent recommendations.
What this means for restaurant owners
The restaurants with the strongest AI visibility share two characteristics: a distinctive, repeatable concept that gives models something specific to say, and editorial coverage in the outlets models retrieve from.
Getting onto the Washingtonian's best-of lists, ARLnow's reader polls, and the local food blogs that rank for dining queries in your area matters more than maintaining your own website. The models are not reading your about page — they're reading what everyone else says about you.
Review volume is the floor. Editorial mentions are what takes you from occasionally appearing to appearing consistently across all four models, every time someone asks.
Want to know how your restaurant appears in AI recommendations? Run a free audit and see your V-Score.