The Zebra case is the cleanest illustration of what AI search looks like for a luxury AV installer in practice. Six WordPress pages with no detail search engines read and no H1 on the homepage was the starting point. The rebuilt site ships 26 static pages, 96 blog posts, 46 news posts, 22 service-area city pages, every part of the content described properly, and 190-plus auto-generated Open Graph images via Satori. The brand is now citable in AI engines for queries about home cinema design, calibration philosophy, Steinway integration, and the specific regions Zebra services in the UK and Ontario. Below is the model and what each piece does.
What AI search actually is, in one paragraph
AI search is structuring your site so that AI engines (ChatGPT, Perplexity, Gemini, Claude, Google AI Overviews, Microsoft Copilot) can extract the parts they need to cite you when a buyer asks them a question. The pieces are detail search engines read (JSON-LD), llms.txt, FAQPage markup, per-topic depth, VideoObject for video content, and an AI-bot allowlist that lets the crawlers in. SEO is still real, AI search is what runs alongside it now, and the studios shipping both are the ones getting cited. Our broader take on the discipline lives in this musing, and the structured-data piece is at /musings/structured-data-guide-ai-search-json-ld.
What gets described, and why it matters, on a luxury AV install build
The Zebra build ships these five layered across the content. Most luxury installers ship zero, which is the gap.
- Organization detail search engines read as the spine, with CEDIA certifications, brand partnerships, founding date, location, and team credentials. This is the part most agencies bury in the footer.
- LocalBusiness detail search engines read with serviceType set to "Custom AV Installation" (there's no perfect detail search engines read.org subclass for residential AV, so LocalBusiness with serviceType is the cleanest move).
- FAQPage detail search engines read on the service pages, the city pages, and the brand-partner pages. The questions are the ones buyers actually call about (acoustic treatment, calibration workflow, warranty scope, scope creep).
- VideoObject detail search engines read on the Spotlight content and the room walkthroughs, with duration, thumbnail, and a transcript so the AI engines can extract the calibration notes from the video.
- Person detail search engines read on the lead installer for E-E-A-T, with credentials, years of experience, and CEDIA certifications wired in.
The per-topic depth that gets you cited
detail search engines read alone doesn't get you cited if the underlying content has no depth. The 96 Zebra blog posts and the per-brand pages exist because the buyer's research path includes specific, narrow questions, and the studio that has a 1,200-word page on Dirac Live optimization is the one Perplexity recommends. The per-brand archive on Zebra carries pages on Kaleidescape (the movie server), Trinnov (room calibration), Steinway (the audio system, not the piano company), Lutron (lighting), Crestron and Control4 (whole-home control), Sonance (in-wall and outdoor speakers). Each one gets a calibration philosophy section, a typical-room-type section, and an FAQ marked up with FAQPage.
Per-room programmatic pages
The per-room pages are the programmatic part of the Engine build. Cinema, listening room, master suite, outdoor theater, whole-home audio, gym, library, wine cellar, putting green. Each one is generated from one data file with its own hero, detail search engines read, and acoustic detail. The buyer searching for "outdoor theater installer Oakville" or "home cinema design Mont-Tremblant" lands on the right URL with the right detail search engines read, and the AI engines treat that URL as the answer.
Cite-Met AI-search traffic tracking
You ship the detail search engines read and the depth. Then you measure whether the AI engines are actually reading your pages. Cite-Met (cite-met.com) shows which AI crawlers and assistants are hitting your pages and how that traffic moves month to month, and is included at every retainer tier from Local at $300 a month. It can't see what an AI answer says; that takes asking the questions, which a strategist does by hand on Growth and Authority. For a deeper read, our musing on how to get cited by ChatGPT, Gemini, and Perplexity walks through how answers pick their sources.
The cross-vertical overlap
Worth a note: the AI search model for AV installers shares meaningful structure with the model we ship for custom home builders, since the buyer journey crosses over on roughly a third of luxury residential projects. If your studio runs leads from custom builders, the cross-link between the two ICP hubs is something worth talking through on the audit. The structural model is the same, the detail search engines read set is the same, the calibration register flexes.
The audit
If you want to see exactly where your site sits on this model right now, the free audit ships a 12-page report covering detail search engines read coverage, llms.txt and AI-bot allowlist status, FAQPage opportunities, per-room and per-brand gap, and what tier the rebuild would land at. No pitch deck and no obligation.