How a kitchen remodeler shows up in ChatGPT, Perplexity and Google AI answers
Homeowners now ask an AI assistant who should remodel their kitchen and what it costs. What those systems read, and the page-level changes that get a design-build firm named.
Type this into ChatGPT: "I'm in Bethesda and want to remodel a 200 square foot kitchen, mid-range finishes. Who should I talk to and what will it cost?" You get a cost range, a short list of firms, and a sentence about each. No ads, no map, no gallery. If your firm is not on the list, you were never considered.
For a trade with a three to nine month research phase, this is the new first step for a growing share of homeowners. The good news is that the assistants are not using a secret ranking. They read the same public material Google reads and quote the sentences that state a fact. Most remodeler websites give them nothing to quote.
What the assistants read, in rough order
- Your Google Business Profile: category, services with prices, hours, and the text of your reviews.
- Your website, specifically pages that answer a question in ordinary sentences with a number in them.
- Directories where your name, address and phone match exactly: Houzz, Yelp, the BBB, NKBA and NARI member listings, the local chamber.
- Anywhere else a human wrote about you: a local news feature, a supplier's "our dealers" page, a neighborhood forum.
Consistency across those sources is the first job. If your showroom moved two years ago and Houzz still has the old address, an assistant trying to decide whether you are one firm or two will often leave you out.
Write pages an assistant can quote
Compare these two openings for a service page.
Before: "At [Firm], we believe your kitchen should reflect your lifestyle. Our award-winning design team brings your vision to life with craftsmanship you can trust."
After: "We design and build full kitchen remodels in Bethesda, Chevy Chase and Potomac. Most of our projects run $55,000 to $110,000 and take 10 to 14 weeks after cabinets arrive. Design consultations are free and include a written scope and budget range."
The first paragraph cannot be quoted because it contains no facts. The second has the services, the towns, the price range, the timeline and the offer. An assistant can lift it whole, and so can Google's AI Overview.
Do the same for every question homeowners type. Give each its own page or H2 and answer it in 40 to 80 words with a number:
- How much does a kitchen remodel cost in [city]?
- How long does a kitchen remodel take?
- Can I stay in my house during a kitchen remodel?
- Should I move the sink or keep the layout?
- How long do custom cabinets take to arrive?
- Do I need a permit to remodel my kitchen in [county]?
- What is the difference between refacing and replacing cabinets?
Six to ten of these pages, each with a local answer, is more than most firms in your metro have published. Only 23.2% of kitchen remodelers even show a public email on their profile; the bar for being a clearly documented firm is low.
Make the reviews do the work
Assistants quote review text to describe what a firm is like. "Reviewers mention that the designer caught layout problems early and that cabinets arrived on schedule" is a sentence built from your reviews. Generic five-star reviews produce nothing for the assistant to say.
So the ask at the final walkthrough should request a detail and a name. The reviews guide has the script. The median kitchen remodeler has 9 reviews; a firm with 40 that mention the designer, the timeline and the budget is a firm an assistant can describe.
Fix the listings
Pick one exact format for your name, address, phone and hours and copy it everywhere: Google, Houzz, Yelp, the BBB, Angi if you are there, NKBA or NARI, the chamber, your own footer. "Suite 200" on three listings and "Ste 200" on two is a small thing to a human and a conflict to a machine. Hours matter here as well; 29.5% of kitchen remodelers list none, and a firm with no hours anywhere reads as possibly closed.
Add schema markup to your site for LocalBusiness (or HomeAndConstructionBusiness), with your address, phone, hours, service area and price range, and FAQPage markup on the question pages. It is a half-hour job for whoever maintains the site and it is how assistants confirm what they read.
Let the crawlers in
Check robots.txt. Some site builders block everything except Google by default. You want GPTBot, PerplexityBot, ClaudeBot and Google-Extended allowed, or the assistants cannot read the pages you just wrote. Add an llms.txt file at your domain root with a plain-text summary of who you are, where you work, what you charge and links to the key pages. It costs nothing and some assistants read it first.
Measure it monthly
Once a month, ask ChatGPT, Perplexity and Google (watch for the AI Overview) twenty local questions: "best kitchen remodelers in [city]", "kitchen remodel cost [city]", "design build kitchen [suburb]", and so on. Screenshot the answers and note where you appear. Movement on specific questions usually shows inside a month of the pages going live; the general "best in city" lists take longer and depend on review count.
Frequently asked
Is this different from SEO? The sources overlap almost completely. The difference is what gets rewarded: SEO rewards a page that ranks, AI search rewards a sentence that can be quoted. Write for the sentence and both improve.
Does Houzz still matter for this? Yes. Assistants read Houzz project pages and reviews for remodelers specifically, because Houzz is the category's directory. Keep the profile accurate and add projects with real budgets.
Can we pay to be in the answer? Not today. There is no ad placement in ChatGPT or Perplexity local answers, and Google's AI Overview pulls from organic sources. The only lever is being the clearest documented firm in your metro.
If you want the question pages written and the listings reconciled for you, AI search optimization is one of the services in the free 14-day trial.
Steven has spent the last several years running growth campaigns for local service businesses, kitchen remodeling companies among them, and builds the benchmark datasets on this site from public Google listings. About the editor