The Design and Build Firm's Guide to AI Search Visibility

Ask ChatGPT, Perplexity, or Claude for an interior designer or a custom builder, and a short list of firms comes back by name. Run the same search on Google and you get ranked links to sift through yourself, pulled from much of the same web the assistants read. AI search visibility measures whether your firm is on the AI list. This guide is about getting your firm named when a buyer asks AI for a designer or builder. It does not cover AI tools that render rooms or create design concepts. The referral, not the renderings.

We measured the referral directly instead of guessing. Four AI engines cite designers and builders in patterns no generic AEO guide has tracked. Below, we share what we found, why it happens, and what your firm can do about it.

What our July 2026 study found

Our July 2026 study queried Perplexity, Claude, ChatGPT, and Google live on July 10, 2026. Our team ran 37 queries in all: 29 find-a-pro buyer queries about the firms Illume Digital Studio serves (interior designers, custom home builders, and kitchen and bath remodelers) plus 8 marketing and AI-visibility queries. The 29 buyer queries produced 1,400 cited sources, the basis for every stat below. Two honest caveats: we ran each engine's developer version with web search on, which comes close to the apps people use but is not identical, and our Google column is standard organic search (the top eight to ten results), not the AI Overview box. Aggregate patterns are the claim, not any single query's list.

Three findings carry the study:

• Visibility is engine-fragile. The four engines cited 443 distinct firms; 321 (72.5%) appeared in only one engine. Just 9.9% appeared in three or more, and 2.5% in all four.
• Gatekeepers carry the answers. Roughly half of the 1,400 sources (47.8%) were third-party pages: directories 23.6%, media listicles 11.3%, associations and government 8.4%, social 4.6%. Houzz was the most-cited domain in the entire study (125 citations).
• Each engine pulls from different sources. Perplexity leaned community and social (14.8% of citations; Reddit was its top non-directory domain, 27 citations). ChatGPT cited directories least (11.2%) and associations and .gov sources most (27.1%). Google (32.6%) and Claude (25.6%) were the most directory-heavy.

Two smaller July 2026 observations: on bare "near me" queries with no city given, Claude and ChatGPT declined to recommend anyone and asked for a location, while Perplexity inferred a location and answered. Claude answered 6 of the 29 buyer queries from model memory, zero web citations.

Buyers now ask AI assistants before they ask Google

Buyers changed the first move. Instead of scanning ten blue links on Google Search, a homeowner types a full question into a chat window and receives two or three firm names back. The tell has already reached discovery calls: a client mentions "I asked ChatGPT" before the first meeting. Somewhere, an engine assembled the shortlist, and every engine assembles it differently.

An AI answer makes the choice for the buyer. A Google results page shows ranked links and lets the buyer judge; an AI answer does the judging first and shows the survivors. A three-name answer gives a buyer no reason to keep digging, so absence from the answer is invisible by design. The firm never learns the question was asked.

Buyers asking AI first is only half of it. The bigger issue is that each engine gives a different answer. ChatGPT, Perplexity, Claude, and Google each read a different slice of the web, so a buyer's choice of app quietly changes which firms exist. The data is blunt here: 72.5% of the 443 cited firms surfaced on a single engine only. The practical translation: most firms win the shortlist decision in one place and lose silently everywhere else.

What AEO actually is (and is not)

Answer Engine Optimization (AEO) means structuring a firm's online presence so AI engines can find, verify, and quote the firm accurately. Some people call the same discipline GEO; the naming is unsettled and not worth debating. AEO is not a replacement for SEO, because the same fundamentals feed both. The difference is the destination: SEO earns a ranking on a results page, AEO earns a citation inside the answer itself.

In our work at Illume Digital Studio, AEO is concrete work rather than theory:

• Finding the exact questions buyers type into AI, then answering them on your own site.
• Writing pages that put the answer first, so an engine can lift it cleanly.
• Structuring content for how people actually ask out loud, in full questions.
• Catching the easy wins where your firm is one edit away from getting cited.

AEO is not a trick, and no honest practitioner promises a guaranteed citation. Roughly half of the cited sources (47.8%) sit on third-party pages outside a firm's direct control. AEO improves the inputs AI engines read; nobody can dictate the output.

How AI engines decide who to recommend

AI engines weigh corroboration more heavily than anything a business says about itself. Four mechanics show up again and again:

• Consistency beats any single signal. When sources conflict on a firm's basics, engines exclude the firm rather than guess.
• Third-party corroboration is the currency. Press, associations, directories, and reviews do the vouching, which matches the 47.8% third-party share in the same dataset.
• Direct-answer content gets quoted. A clear paragraph stating what a firm does, where, and for whom is liftable; brochure adjectives are not.
• Proof of experience and trust carries weight. Google calls this EEAT: experience, expertise, authority and trust. Credentials, awards, and project history make every page count for more.

EEAT is where a firm's offline credibility becomes a search signal. Illume weaves credentials, awards, project history, and industry involvement into a firm's digital presence, so every piece of content carries more weight in search. The same signals feed AI engines and Google alike, which is why the study's third-party finding and the EEAT mechanic point at the same job.

Why your Google ranking doesn't carry over

Google ranking and AI citation are separate games with separate scoreboards. The study shows why: 72.5% of the 443 cited firms appeared on one engine only, and each engine reads a different source mix. Houzz illustrates the spread: Perplexity cited Houzz on 90% of buyer queries, Google on 66%, Claude on 62%, and ChatGPT on just 21%. Doing well on one engine does not help you on the others.

Multi-engine presence gets built from assets no single ranking controls:

• A website AI engines can read and quote.
• A complete Google Business Profile.
• Houzz and trade directory listings.
• Association memberships and active review platforms.
• A real community and social footprint.

We have created dedicated guides for the profile side of the list: the Google Business Profile for interior designers and the Google Business Profile for home builders. If the weak point is the website itself, start with the website mistakes design and build firms make.

What a design and build firm should do about it

Design and build firms already own the raw material AI engines verify. Roughly half (47.8%) of the 1,400 cited sources were third-party proof about firms rather than the firms' own pages, and that proof is exactly what a good firm accumulates:

• A portfolio of finished projects.
• Client reviews on the platforms engines read.
• Credentials and association memberships.
• Project photos and project history.

The work is making those trust assets machine-verifiable and consistent everywhere an engine reads. Same firm name, same service area, same specialties, on every source an engine might consult. Illume Digital Studio treats the checking as standing work: our clients on every path get a standing AI search visibility check across the major AI answer engines and Google AI results.

The honest close: nobody can guarantee AI mentions, and any agency promising guaranteed citations is overselling. What moves the odds is measuring the current answers and improving the inputs the engines read. The only way to learn your own standing is to ask the engines the questions your buyers ask.

If you want this done for your firm, apply to work with us. It takes about three minutes. When the gaps are clear, our Google and AI search work is where the ongoing work lives.