GEO for Architecture Firms and Engineering Consultancies: Getting Cited by AI

🇫🇷 Version française

"Which architect handles wood-frame extensions near me," "which structural engineer for a four-story apartment building": more and more clients are asking AI these questions before they ever open Google. Here's what ChatGPT, Perplexity, and Gemini check before citing a firm.

Architecture firms and engineering consultancies have traditionally lived off referrals and a strong portfolio. The problem is that this portfolio — often a photo gallery or a downloadable PDF — is exactly the kind of content a generative AI can't use. It can't "see" a building in a photo: it needs text to understand what the firm does, where, and for whom.

Why this industry is particularly unprepared for GEO

The visual culture of the industry — full-page image sites, PDF portfolios, photo-driven social media — is a direct barrier to AI visibility. Beyond the homepage, many firm websites have almost no usable text: no project descriptions, no dedicated page per service type, no clear statement of the geographic area covered. An AI searching for "who does energy-efficient renovation for condo buildings in Lyon" simply has nothing to read.

What AI checks before recommending a firm

• The specific specialty — renovation, new construction, extension, commercial work, heritage buildings, concrete or timber structures. A generic line like "architecture and consulting" doesn't help AI match a specific request.
• The service area — a firm that never states its actual working radius is hard to recommend for a local search, even if it operates there in practice.
• Past project types — described in text, not just shown in photos: size, typology, specific constraints, outcome.
• Professional licensing — architectural board registration, structural or engineering certifications, professional liability insurance mentioned clearly.
• External trust signals — client reviews, listings in professional directories, awards or coverage in trade press.

The trap of a 100% visual portfolio

A portfolio made entirely of high-resolution images with no captions or descriptions gives AI nothing to work with — even if the project name is in the image URL. Alt text helps a little, but it's no substitute for an actual paragraph explaining the project: context, technical constraints, materials, size, timeline. It's that text, not the photo, that lets AI connect a user's request to a past project.

Engineering consultancies: technical depth pays off, if it's written down

For an engineering consultancy (structural, MEP, thermal, civil), technical precision is actually a potential GEO asset — AI favors specific, demonstrative content. But that asset only materializes if capabilities are spelled out in writing: types of calculations performed, standards mastered, sectors served (residential, commercial, industrial) — rather than summarized as a row of software logos.

Does GEO replace networking and word-of-mouth?

No, and it isn't meant to. Referral networks — real estate agents, developers, notaries — and word-of-mouth will stay central to this business. But a growing number of clients, especially for residential renovations or extensions, start their search online — and increasingly through AI rather than a traditional search engine. Not showing up there means missing that early screening stage, before the first contact is even made.