How does AI search route RFQs to fabrication shops?

TL;DR

A buyer asks an AI assistant: ChatGPT, Perplexity, Claude, Google AI Overviews, for a shop that does their work. The assistant returns a short list of named shops with citation links. Buyers click through to whichever shop's site reads most like a credible vendor for the specific job. The shops that show up are the ones whose websites the AI engine could parse and verify; the rest are not in the result set at all.

The flow has three steps. (1) Query, the buyer types something specific: "small-batch aluminum CNC machining shop in Ohio for a 200-unit run, 0.005-inch tolerance." (2) Retrieval, the AI engine pulls from its citation index, which it built by crawling sites with structured data, FAQ schema, and clear capability text. (3) Ranking, the engine returns 3 to 8 shops with brief descriptions and direct links.

Shops show up in step 2 only if the engine can parse them. That means: capability lists in plain HTML text (not PDFs, not image badges), Service and LocalBusiness JSON-LD schema, FAQ page with FAQPage schema, content updated within 90 days, and at least a handful of inbound citations from credible directories. Most small shop websites fail two or three of those checks.

Once a shop shows up, ranking comes down to fit and freshness. AI engines preferentially cite pages whose text matches the query closely, whose schema confirms the service offering, and whose modified date is recent. A shop that publishes a quarterly capability update will outrank a shop with a static 2018 site every time, even if the static site has more domain authority.

Key facts

Perplexity reported 5B+ URLs in its citation index by late 2024 and 15M monthly active users.
Google AI Overviews launched to all US users in May 2024; Search Engine Land tracks AI Overview appearance rates by category.
ChatGPT search launched October 2024 and became default for logged-in users in early 2025.
Pages with FAQPage and Service JSON-LD schema get cited at substantially higher rates per Search Engine Journal's 2026 AI citation study.
Pages updated within 90 days are cited at roughly 3x the rate of stale pages for time-sensitive vertical queries.

Common follow-ups

Is this just SEO under a new name?

Overlapping but not identical. SEO ranks pages on a results page; AEO ranks pages inside an answer where there may be no click. AEO weights structured data, lead-with-the-answer content, and freshness more heavily than traditional SEO does. Both want fast, useful, well-structured content.

Can a 2018 brochure site catch up?

Yes, with focused work, adding a Capabilities page in plain text, a Certifications page, a 6-question FAQ block with FAQPage schema, and a quarterly update cadence will move the needle within 60 to 90 days. A full rebuild compresses that to 2 to 4 weeks.

Do AI engines favor big shops or small shops?

Neither. They favor shops whose websites match the buyer's specific query precisely. A 4-employee shop with a clean capabilities page will outrank a 200-employee shop with a flash-driven 2014 site every time.

When this doesn’t apply

Buyers procuring through approved-vendor lists, RFQ portals (e.g., Xometry-style platforms), or referrals do not pass through AI search. Shops whose pipeline is entirely those channels see no impact from AEO.

Sources

By Isaiah Grant, Founder, Rebuilt StudioUpdated Apr 29, 2026