How do architects get clients now that people ask ChatGPT?

Mostly the same way as before: referrals, reputation and visible past work. The difference is that prospects now verify you, and compare you with alternatives, through Google Maps, directories and AI answers. Firms with clear facts, detailed project pages and genuine third-party mentions are easier for those systems to name.

Is SEO being phased out for architecture firms?

No. Google says its AI Overviews and AI Mode are rooted in its core Search ranking and quality systems, and that optimizing for generative AI search is still SEO. ChatGPT search also depends on crawling the web through OAI-SearchBot. What is shifting is emphasis: being cited inside the answer matters more as fewer people click through.

How do I get my architecture firm recommended by ChatGPT?

Make your firm easy to verify: clear entity facts, detailed project pages, consistent profiles and genuine third-party mentions such as association listings, press, awards and reviews. Make sure your site does not block OAI-SearchBot. No one can guarantee a recommendation, but these steps remove the most common reasons a firm is skipped.

Should an architecture firm create a page for every city it serves?

Only for places where you have real work and something specific to say. A handful of substantial pages, each backed by local projects and knowledge, is safer and more useful than dozens of template pages, which Google treats as scaled content abuse.

How much does SEO and AI search work cost for an architecture firm?

It depends on scope: how many pages and projects need work, whether the site needs technical fixes, and how many cities and project types you target. A sensible approach is a fixed-scope assessment first, with a written baseline and prioritized fixes, so you can decide what to do in-house and what to pay for. Be wary of open-ended retainers with no baseline or exit point.

What should I look for in an SEO agency for architects?

A fixed scope, a baseline you can re-check, and white-hat work: fixing entity facts, improving project pages and earning genuine mentions. Be cautious of guaranteed rankings, paid listicles, review schemes or mass-generated city pages.

AI search guide for architects and engineers

SEO and AI search for architecture firms: how to become the studio ChatGPT and Google can confidently recommend

A homeowner in San Jose no longer only asks a neighbor for an architect. They also ask Google Maps, scroll a directory, or type 'who designs ADUs in San Jose?' into ChatGPT. This guide explains, from first principles, what those systems need to see before they name your firm, and what to do about it without gimmicks.

Updated 2026-10-10 · Kai, TruGoods · 10 min read

Short answer

AI assistants recommend architecture firms they can verify: a clear licensed-professional entity, project pages stating type, location, scope and role, real service-and-city content, consistent profiles, and independent mentions such as AIA chapter listings, local press, awards and genuine reviews. Google says its AI features build on normal Search ranking, so SEO is still the foundation; no special AI file or schema unlocks recommendations.

How do architects get clients today?

Ask most small studio principals how their last five projects started and the answer is still mostly referrals: a past client, a contractor, a structural engineer. What has changed is what happens right after the referral. The prospect looks you up, often looks up your alternatives too, and some of that research now happens inside an AI answer rather than on your website.

A prospective residential or small commercial client usually passes through several of these channels before calling:

Referrals from past clients, builders and consultants, followed by a quick check of your website and reviews.
Google Maps and your Google Business Profile, often reached through searches like 'architect near me' or 'residential architect Palo Alto'.
Houzz-type design directories and portfolio platforms where they browse photos by style and location.
City- and project-specific searches: 'ADU architect San Jose', 'hillside home architect Oakland', 'tenant improvement architect San Francisco'.
Conversational questions to ChatGPT, Google AI Overviews or AI Mode, Perplexity and similar tools: 'Who designs ADUs in San Jose?' or 'Which Bay Area architects specialize in Eichler remodels?'

Why AI search changes the math for small firms

When an AI summary answers the question directly, fewer people click through. Pew Research Center found that when a Google AI summary appeared, users clicked a traditional search result in 8% of visits, compared with 15% without one, and clicked a link inside the summary in only 1% of visits (Pew Research Center, 2025). Being named inside the answer now matters more.

The mechanics are not mysterious. Google's guide to optimizing for generative AI features says AI Overviews and AI Mode are rooted in its core Search ranking and quality systems, and that optimizing for generative AI search is still SEO. For ChatGPT search, OpenAI documents that OAI-SearchBot is used to surface sites and that sites blocking it may not appear in those answers (OpenAI crawler docs). So good SEO services for architects still do the same job, with more emphasis on verifiable facts and independent evidence.

There is an honest headwind: Chen et al. (2025) found that AI search engines strongly favor earned, third-party media over brand-owned content, and that niche brands must overcome a bias toward big names (arXiv 2509.08919). A ten-person studio cannot out-publish a national firm, but it can be the best-documented local option for a specific project type in a specific city.

What an AI assistant needs before it recommends your firm

Think of an AI assistant as a cautious researcher. Before it names a firm for 'ADU architect in San Jose', it needs consistent evidence, in pages it can retrieve, that the firm exists, is qualified, does that work, works in that area, and is well regarded by others. If any link is weak, it names someone else or stays generic.

Five kinds of evidence carry most of the weight:

Clear entity facts: legal firm name, principals and their professional licenses where your state board makes them public, address, service area, phone and the project types you actually do. State them in plain text on an About or Studio page, not only in a PDF or image.
Project portfolio pages with real facts: each project gets its own page stating type, city or neighborhood, approximate size, year, scope and your role, not just a gallery of photos.
Service and city pages with substance: a page about ADUs in San Jose should explain what you have learned doing ADU work there, show related projects and answer real client questions. A template with the city name swapped in is a doorway page, and Google treats mass-produced pages as scaled content abuse.
Consistent profiles: your Google Business Profile, directories, LinkedIn and association listings should match your website on name, address, services and principals. Google explicitly points local businesses to Business Profile for visibility in AI responses.
Third-party mentions: your AIA chapter or other professional association listing, local press coverage, award and juried competition lists, publications that featured a project, consultant and builder pages that credit you, and genuine client reviews.

What clients ask AI, and what your site must make easy to extract

Write down the questions your best prospects would type into ChatGPT or Google, then check whether a stranger could answer each from your site and public profiles in under a minute. Google describes 'query fan-out', where the AI runs several related searches behind one question, so one client question often becomes checks of reviews, location, specialties and examples.

What a client asks AI	What the AI has to verify	What your site and profiles must make easy to extract
Who designs ADUs in San Jose?	The firm does ADUs, works in San Jose, and has examples	An ADU service page with San Jose context; project pages for completed ADUs that name the city, size, year and your role
Which Bay Area architects specialize in Eichler remodels?	A real specialty, not a keyword	Several Eichler project pages with neighborhood, scope and before/after photos with descriptive alt text; an article on what you have learned remodeling them
Is this architect licensed and reputable?	Licensed principals, track record, independent opinions	Named principals with license information, AIA or association membership, awards, press mentions and reviews on independent platforms
How much does it cost to hire an architect for a home addition here?	Honest, specific process and fee information	A fees-and-process page explaining how you price (percentage, fixed, hourly), typical phases and what drives cost, without invented averages
Which engineering firm can handle seismic retrofits for commercial buildings in Oakland?	Technical capability, credentials, comparable projects	Capability pages, licensed engineers by discipline, project sheets with building type, location, year and scope

How to structure project pages for Google and LLMs

Most architecture websites are beautiful and nearly silent. A full-screen slideshow with a project name and no text tells a retrieval system almost nothing. The fix is not to clutter the design but to add a short, factual summary people and machines can both read. Google also notes that its AI features can surface images and video, and that following image SEO best practices already counts as optimizing for generative AI search. Each project page should include:

A descriptive title such as 'Detached ADU, Willow Glen, San Jose', alongside any poetic project name.
A fact block in plain text: project type, city or neighborhood, approximate size, completion year, scope (new build, addition, remodel, adaptive reuse) and your role (architect of record, design architect, interiors).
Team credits: builder, structural engineer, landscape architect, photographer. These create verifiable connections to other sites.
Two or three paragraphs on the design problem and how you solved it. This is the first-hand, non-commodity content Google says matters most.
Photos with descriptive file names and alt text, for example 'Kitchen with clerestory windows in a remodeled Eichler, Palo Alto'.
Structured data as a supporting layer, such as Organization or ProfessionalService markup for the firm. Google says no special schema is required for AI features; it supports rich-result eligibility, not a ranking switch.
Links to related service and city pages, and back again.

What to avoid: tactics that waste money or backfire

When you evaluate an SEO agency for architects or a marketing agency for architecture firms, the tactics matter more than the dashboard. Google warns that seeking inauthentic mentions is not as helpful as it seems, and the C-SEO Bench study (2025) found most 'conversational SEO' content tricks ineffective or harmful, with traditional ranking dominating (arXiv 2506.11097). Avoid:

Fake or incentivized reviews. They violate platform policies and can get a profile restricted.
Paying for placement in 'top 10 architects in [city]' listicles. These are inauthentic mentions.
Mass-generated city pages: fifty near-identical pages for every town in the county. Google names this scaled content abuse.
Hidden text or instructions aimed at AI models, and cloaking (showing crawlers different content than visitors).
Rewriting your site into keyword-stuffed Q&A chunks. Google says there is no need to chunk or rewrite content for AI.
Any vendor that guarantees rankings or ChatGPT recommendations.

SEO for engineering companies: what is different

Engineering consultancies, and architecture firms doing mostly institutional or commercial work, are usually B2B and RFP-driven. The searcher is often a developer, facilities director, general contractor or design firm assembling a team. They ask things like 'MEP engineering firms in the Bay Area with lab experience', and they care about capability and risk more than style.

For these buyers, digital marketing for architecture and engineering firms should read like a public, well-organized qualifications package. Formats such as the SF 330 used in US federal architect-engineer procurement already ask for the facts AI systems look for: key personnel, relevant projects, roles and capabilities.

A credentials page listing licensed engineers by discipline and the states they are licensed in, plus firm certifications you actually hold, such as small or disadvantaged business certifications.
Capability pages per discipline or sector (healthcare, labs, K-12, multifamily) with relevant project sheets, not generic paragraphs.
Project sheets that state client type, building type, location, size, year, delivery method and your scope.
Named experts with short bios, publications, talks and committee roles, which create the earned, third-party evidence AI systems favor.

A 90-day plan for architecture and engineering firms

You do not need to do everything at once. This sequence puts the cheapest, highest-leverage fixes first and leaves room to measure. It is the same order we follow in a fixed-scope AI visibility audit.

Days 1-10: Baseline what AI and Google say nowWrite 10-15 real client questions by project type and city. Record, with dates, what Google, AI Overviews, ChatGPT and Perplexity return. Verify your site in Search Console and check that robots.txt does not block Googlebot or OAI-SearchBot unintentionally.
Days 10-20: Fix the entity factsRewrite the About or Studio page with plain facts: legal name, principals, licenses, address, service area, project types. Make your Google Business Profile, directories and association listings match it exactly.
Days 20-45: Upgrade your ten most representative project pagesAdd the fact block, team credits, design narrative, descriptive alt text and internal links. Prioritize the project types and cities you most want to win.
Days 45-60: Build a few service and city pages with substanceWrite one page for each real combination of service and place you serve, grounded in your own projects and local knowledge. Three strong pages beat thirty thin ones.
Days 60-80: Earn third-party evidenceUpdate your AIA chapter or association profile, submit eligible projects to awards, pitch a project story to local or trade press, ask builders and consultants to credit you on their sites, and ask recent clients for honest reviews on independent platforms.
Days 80-90: Measure and decideRe-run the same questions under the same conditions and compare. Check the Generative AI performance report in Search Console. Keep what moved, drop what did not, and set a quarterly re-check rather than chasing daily fluctuations.

What we have seen, and optional extras

We prefer small, honest data over big promises. For a Bay Area design studio we worked with, a 7-query Chinese-language Google check on 2026-10-04 (non-personalized desktop, San Jose location) showed the studio on page one for 3 of the 7 queries, at positions 3, 4 and 7, and 2 of the 7 AI Overviews cited it. That is a snapshot, not a before-and-after claim, and we do not attribute inquiries to it. The details are in our case study.

An llms.txt file is a cheap, optional summary for AI tools; Google says Search does not use it, and its effect elsewhere is unproven. WebMCP agent-readiness work can let browser agents prepare a consultation request on your site; it is early and experimental. Both come after the fundamentals, never instead of them. See also how to get recommended by ChatGPT and our SEO and GEO services.

Frequently asked questions

How do architects get clients now that people ask ChatGPT?: Mostly the same way as before: referrals, reputation and visible past work. The difference is that prospects now verify you, and compare you with alternatives, through Google Maps, directories and AI answers. Firms with clear facts, detailed project pages and genuine third-party mentions are easier for those systems to name.
Is SEO being phased out for architecture firms?: No. Google says its AI Overviews and AI Mode are rooted in its core Search ranking and quality systems, and that optimizing for generative AI search is still SEO. ChatGPT search also depends on crawling the web through OAI-SearchBot. What is shifting is emphasis: being cited inside the answer matters more as fewer people click through.
How do I get my architecture firm recommended by ChatGPT?: Make your firm easy to verify: clear entity facts, detailed project pages, consistent profiles and genuine third-party mentions such as association listings, press, awards and reviews. Make sure your site does not block OAI-SearchBot. No one can guarantee a recommendation, but these steps remove the most common reasons a firm is skipped.
Should an architecture firm create a page for every city it serves?: Only for places where you have real work and something specific to say. A handful of substantial pages, each backed by local projects and knowledge, is safer and more useful than dozens of template pages, which Google treats as scaled content abuse.
How much does SEO and AI search work cost for an architecture firm?: It depends on scope: how many pages and projects need work, whether the site needs technical fixes, and how many cities and project types you target. A sensible approach is a fixed-scope assessment first, with a written baseline and prioritized fixes, so you can decide what to do in-house and what to pay for. Be wary of open-ended retainers with no baseline or exit point.
What should I look for in an SEO agency for architects?: A fixed scope, a baseline you can re-check, and white-hat work: fixing entity facts, improving project pages and earning genuine mentions. Be cautious of guaranteed rankings, paid listicles, review schemes or mass-generated city pages.

Sources

Related guides

If you want to know what Google and AI assistants currently say about your firm and which fixes matter most, start with a fixed-scope assessment.

See the SEO & GEO service Discuss your site