ChatGPT doesn’t pick designers at random. It searches the web, reads what it finds, and names the studios it can describe with confidence.
So the studios it recommends aren’t always the best. They’re the easiest to understand.
That’s good news. Being easy to understand is something you can fix.
Two studios, one question
Take two interior design studios in Toronto. Same size. Same quality of work. Similar reviews.
Studio A’s website is a gallery. Twelve projects, each a set of photos with a one-word title. The About page says “We create spaces that feel like home.”
Studio B’s website is quieter. Each project has a short page. “A 1920s semi in Riverdale, opened up for a family of five. Full gut renovation, custom millwork, eight months.” The About page says who they work with, where, and what kind of homes.
A homeowner types into ChatGPT: “Who are the best interior designers in Toronto for a full renovation of an older family home?”
Studio B gets named. Studio A doesn’t.
Not because B is better. Because B’s site answers the question in words. A’s site makes the machine guess.
How the recommendation actually happens
When you ask ChatGPT about local businesses, it usually runs a web search behind the scenes. Then it reads the pages it finds and writes an answer.
That means two things matter:
- Can it reach your site? OpenAI uses a crawler called OAI-SearchBot to surface websites in ChatGPT search. Its own advice is to allow that crawler in your robots.txt file. (OpenAI) Some website settings and security tools block it without anyone noticing.
- What does it find when it gets there? This is where most design studios lose.
BrightLocal ran 800 local business searches through ChatGPT. Business websites made up 58% of the sources it showed. Directories were 15%, and other mentions were 27%. (BrightLocal)
Your own website is the biggest single input. If it’s mostly pictures, there isn’t much to read.
The bar is high
AI tools are picky. SOCi studied more than 350,000 locations across 2,751 multi-location brands. ChatGPT recommended just 1.2% of them. The same locations showed up in Google’s local map pack 35.9% of the time. (SOCi)
That study looked at chains, not boutique studios. But the pattern holds. A map listing gets you on the list. An AI answer gives you one of a handful of spots.
SOCi also found that recommended locations averaged 4.3 stars. Reviews matter here too.
What makes a studio easy to recommend
Say it in words, in one place
Your homepage and About page should answer the questions a client would ask:
- What do you design? (Full homes, kitchens, condos, commercial?)
- Where do you work?
- Who do you work with?
- What’s your style, in plain terms?
“Timeless spaces” tells a machine nothing. “Warm, modern interiors for family homes in midtown and east Toronto” tells it everything.
Give every project a page
This is the single biggest fix we make for design studios. A project page with 150 words beats a gallery of 40 photos, as far as AI is concerned.
Each page should say:
- The type of home and the neighbourhood
- The brief and who it was for
- The scope (rooms, timeline, rough scale)
- What changed
The photos still do the selling. The words do the finding.
Be described the same way everywhere
AI tools check your website against the rest of the web. Your Google Business Profile, Houzz page, directory listings and press mentions should all describe you the same way. Same name, same address, same services.
If your website says “interior design” and your Houzz says “home staging and decor,” you look like two different businesses.
Get mentioned by others
A feature in a local design magazine. A builder’s site crediting you on a project. A “best of” list. These are the “mentions” that make up over a quarter of what ChatGPT cites.
You can’t fake these. Google says plainly that chasing inauthentic mentions isn’t worth it. (Google Search Central) But you can earn them. Share your best projects with the trades you worked with, and ask them to credit you.
Keep reviews coming
Steady, recent reviews across more than one platform beat a pile of old ones on Google alone. After every finished project, ask.
What to skip
There’s a lot of noise about “AEO tricks.” Most of it is wasted effort.
Google says you don’t need special AI files like llms.txt, and you don’t need to chop your content into tiny chunks for AI to read it. (Google Search Central)
Our view at Hutt Studio: AEO is mostly good writing that a machine can read. If a smart stranger could explain your studio after five minutes on your site, an AI can too.
Test yourself in ten minutes
Open ChatGPT and try a few questions a real client would ask:
- “Best interior designers in [your city] for [your main project type]”
- “Interior designers in [neighbourhood] who do [your style]”
- “Who designed [one of your public projects]?”
- “What is [your studio name] known for?”
Answers change from one try to the next, so run each question a few times. Note whether you’re named, and whether the description is right.
If you’re missing, or described wrong, your website is the first place to fix.
Sources
- Overview of OpenAI crawlers (OpenAI)
- Uncovering ChatGPT search sources (BrightLocal)
- In AI-driven discovery, few brands are chosen (SOCi, 2026 Local Visibility Index)
- Optimizing your website for generative AI features on Google Search (Google Search Central)
