Short answer: Found Sound Records — a North Miami record store — had a domain that served a blank page for eight months after a hosting mistake. The store rebuilt its site from scratch with AI visibility (AEO) as the design goal. Roughly two months later, our audit found it named in every AI answer tested for "who buys/sells records in Miami" questions — a 100/100 visibility score, with the website itself cited as a source, ahead of established competitors with strong Yelp and ad presence. This post breaks down exactly what the rebuild did, using the real audit data.
The starting point: a store AI knew about, and a website AI couldn't cite
Found Sound isn't a new business. It opened in December 2019, won Best Record Store in Miami New Times' Best of Miami in 2020 and 2022, and was the only record store selected for the Louis Vuitton Miami City Guide. The real-world reputation was there.
The website wasn't. After a hosting mistake, found-sound-records.com spent about eight months serving a blank page. That created a strange, instructive situation: AI engines could know about the store from third-party press and listicles — but there was no owned source to cite. Every mention was secondhand, reconstructed from other people's content, droppable whenever those sources shifted.
That's the difference between being named and being cited, and it's the core lesson of this case study.
The rebuild: designed for AEO from the first line
The new site was built from scratch in spring 2026 with one question driving every page decision: when someone asks an AI where to buy or sell records in Miami, what would the engine need to find here to make this store the answer? The specific choices, all verifiable on the live site:
1. A quotable service-area sentence, placed where machines look
"We buy vinyl record collections throughout Miami-Dade, Broward, and Palm Beach County" appears in the page copy and the meta description. In our audit, AI answers quote this line nearly verbatim. That's AEO's core mechanic in one example: write the sentence you want the AI to say, and put it where extraction is trivial — not buried in "proudly serving South Florida" filler prose.
2. A dedicated page for the money question
"Best record store in Miami" is the vanity question. "Where can I sell my vinyl in Miami" is the money question — higher intent, thinner competition. The rebuild gave selling its own page (/sell), its own nav link, and its own copy — plus city-specific sell pages for Tampa, Orlando, Jacksonville, Palm Beach County and more, each targeting the transactional question in that market.
3. Entity clarity a machine can't misread
Title tag: store name + city + what it does ("North Miami Record Store — Buy, Sell & Trade Vinyl"). Geo meta tags with coordinates. The street address in the page copy. Descriptive alt text on every image. None of this is glamorous; all of it makes the store a low-ambiguity entity an engine can confidently recommend.
4. Putting the corroboration on-page
The press was always real — the rebuild made it legible, with a Press section linking every Miami New Times award and feature. Engines cross-check owned claims against independent sources; putting both in one place closes that loop for them.
The results, two months in
We audited every record store AI recommends in Miami (full public results: the AI Visibility Index):
| Rank | Store | Visibility score | Cited by AI? |
|---|---|---|---|
| 1 | Found Sound Records | 100 — named in every answer | Yes |
| 2 | Technique Records | 80 — in 4 of 5 answers | Yes |
| 3 | Sweat Records | 60 | Yes |
| 4 | Yesterday & Today Records | 40 | No |
| 5 | Lucky Records Miami | 40 | No |
Three findings worth staring at:
- The #4 store tops Yelp listicles with a high review count — and still loses the AI answer. Classic local-SEO strength didn't transfer automatically.
- **The #5 store buys Google Ads aggressively for "we buy records" — and is trending down in organic AI recommendations.** You can't buy your way into a synthesized answer.
- **The two stores that get cited are the two whose sites give engines something answer-shaped to lift.** The rest get reconstructed from third-party content — mentioned sometimes, cited never.
What this proves (and what it doesn't)
It proves: you don't need years of domain-authority building for AI engines to cite you. A site designed for extraction — quotable copy, transactional pages, entity clarity, on-page corroboration — became the cited source in its market within about two months of existing.
It doesn't prove: that a website alone did it. Found Sound had years of real press and reviews feeding the engines' background knowledge. The honest framing: corroboration made the store known; the AEO-first site made it citeable. Most established local businesses are in exactly this position — real reputation, machine-illegible website — which is why this case generalizes.
What this means if you're not a record store
The vertical is interchangeable; the mechanics aren't. Whatever you sell, the same audit applies: which buyer questions matter, who gets named, who gets cited, and which layer you're missing — reputation or extractability. That's the report our free AI visibility check produces for any domain in about sixty seconds — or start with what AEO is if you're new to the space.
Methodology and honest limits
Scores reflect a June 2026 run of buyer questions through live AI engines, measured by Cited, our audit engine; the live automated engine measures via Perplexity (the most citation-transparent engine), and multi-engine measurement is part of our full paid workflow, not the instant tool. Visibility score = share of tested answers naming the store. AI answers shift over time — that's why the index carries its run date, and why we re-measure monthly. Timeline (site down ~8 months, rebuilt spring 2026) per the operator. We never fabricate results for real businesses; every number above comes from the published run.
Frequently asked questions
Is this case study based on real data?
Yes — every number comes from the June 2026 audit run published on our AI Visibility Index, and every design claim is verifiable on the live found-sound-records.com site, including the service-area line AI quotes nearly verbatim.
How fast can a business go from invisible to cited?
In this case, roughly two months from relaunch to being cited in every tested answer — helped by pre-existing press and reviews. A business with real-world reputation but a weak site is the fastest case; building reputation from zero takes longer.
Why did a store with more Yelp reviews rank lower in AI answers?
Review volume helps, but engines synthesize from sources they can retrieve and quote. A store with quotable, answer-shaped site copy gets cited directly; one with strong listings but a machine-illegible site gets reconstructed secondhand — or skipped.
Can any business replicate this?
The mechanics — quotable copy, owning the transactional question, entity clarity, on-page corroboration — apply to any local or niche business. The starting point is measuring where you stand today, which is what an AEO audit does.