June 5, 2026
AEO for trades: what showing up in ChatGPT actually requires (and why most contractors won't)
Thom Van Dycke · Van Dycke Strategic Business Architecture
How do you show up when your customer asks ChatGPT for a contractor?
You show up when a language model can tell, straight from your website, exactly who you serve, what problem you solve, and why you're the obvious choice in your service area. Answer engine optimization rewards one thing above all: clarity. It's the same problem you've always had with your message, except the judge is now a machine that recommends someone else the moment it can't read you.
The conversation everyone in B2B is having, and nobody in the trades has started
Spend a week inside the marketing world right now and you'll hear one acronym over and over. The sharpest B2B marketing podcasts and newsletters are mid-conversation about it: how to structure a website so AI engines recommend it, how teams are building their own AI tooling, how answer engines are turning into infrastructure rather than a side channel. The smart B2B operators have stopped asking whether AI changes how customers find vendors. They're asking how fast they can get cited in the answers.
Now walk across the street to the trades. The most-repeated advice this month is about holding price at the close. Good advice, and also a conversation from a different decade. The trades segment simply hasn't started the AEO conversation. Nobody has translated it into their language yet. The behavior is already here. The answer isn't.
Homeowners are already doing this
The temptation is to file AEO under things to worry about in two years. Don't. Go read how homeowners actually talk online and you'll find it already happening: people asking for reassurance about a repair quote, then admitting, a little sheepishly, that they already asked AI to sanity-check it first. The buyer is standing in their own kitchen, holding a quote they don't understand, and instead of calling three more contractors they're asking a language model whether they're being taken.
That homeowner is going to get an answer. The only question is whether your business is in it. For most contractors right now the honest answer is no, because the model has nothing to go on. And here's the uncomfortable part. AI doesn't recommend the best contractor. It recommends the most legible one. The business whose website makes its position unmistakable gets named. The business with a handsome homepage that says "quality you can trust" gets skipped, because that sentence describes nobody and answers nothing.
AEO is not SEO with a new name
Let me kill the most comfortable misreading first. AEO is not SEO with the serial numbers filed off. SEO was about ranking a page so a human would click it, and the human did the rest of the work: scanning, comparing, judging, deciding. The page only had to earn the click.
An answer engine removes the click. ChatGPT, Perplexity, Google's AI Overviews all read the page for the buyer and hand back a recommendation. Now the question is whether a machine can read your site and confidently name you to a stranger who asked for help. If your homepage needs a human's generosity to interpret, if someone has to infer what you're great at, the model won't extend that generosity. It picks the site that already said it plainly.
That's why AEO rewards clarity over keywords. You can't keyword-stuff your way into a recommendation. You earn it by being the clearest answer to a specific question.
The four questions every trades site has to answer
Here's the framework. It's the same guide-not-hero spine good marketing copy has used for decades. Who is the customer, what's their problem, what path do you offer, and what does staying stuck cost them. The audience reading those answers changed. The questions didn't. Answer these four plainly on your site and you've done most of AEO. Skip them and no amount of technical work saves you.
1. Who, exactly, is this for? Not "homeowners and businesses." A model can't recommend you to everyone. Name the customer narrowly enough that the machine can match a query to you: homeowners in a specific metro with older homes, property managers running multi-unit buildings, restaurants that need emergency service. Specificity is what makes you matchable. "We serve everyone" reads to a language model as "match me to no one in particular."
2. What problem do you solve, in their words? Write the problem the way the customer would type it into ChatGPT, not the way you'd describe it to another contractor. They don't search "hydronic system diagnostics." They ask why their upstairs is always cold. The site that mirrors the buyer's actual question is the site the engine pulls from, because the engine is matching language to language.
3. What's the path you offer? A short, clear sequence of what working with you looks like. Three steps, not a brochure. Models cite structured, sequential answers, like "first we assess, then we quote a fixed price, then we do the work and you approve each stage," because they're easy to summarize and hand back. A vague "contact us for a consultation" gives the engine nothing to repeat.
4. What does staying stuck cost them? The stakes. What happens if they keep delaying, keep hiring the cheapest bidder, keep ignoring the problem. This is the part almost every contractor site omits, and it's the part that makes an answer persuasive. A recommendation with stakes attached, like "wait on this and a small leak becomes a framing-replacement job," is the one a buyer acts on.
Notice what those four questions never mention: schema markup, page speed, backlinks. That technical layer matters, and you should fix it. But it's the last ten percent. If the four questions above aren't answered clearly in your actual copy, the technical work is polish on a message the machine still can't read.
Why most contractors won't do this
Because it requires a sacrifice most won't make. Answering question one honestly means naming a customer you're for, which means admitting there are customers you're not for. Most contractors would rather stay legible to nobody than risk turning somebody away. So they keep the homepage broad, the language safe, the position blurry, and the model keeps recommending the competitor who was willing to be specific.
This is the same architecture problem that shows up everywhere in a founder-led business. The instinct to stay open to all comers feels like keeping your options open. In practice it makes you invisible. AEO just makes the cost of that choice impossible to ignore, because now there's a machine making the recommendation in real time, and it will never pick the business that refused to say what it is.
Try this: ask the engine about yourself
Open ChatGPT today and type the exact question your best customer would ask: best [your trade] in [your service area] for [the specific job you want]. Read what comes back. Are you in it? Is anyone you respect in it? What did the businesses that got named have in common on their websites?
Then read your own homepage the way the model would, with no generosity and no inference, just the words on the page. Can you answer the four questions from the copy alone? If you can't, neither can the engine, and neither can the homeowner standing in their kitchen asking it for help.
That gap between what you know about your business and what your website actually says is the architecture problem. AEO didn't create it. It just hired a machine to enforce it.
Frequently asked
Do I need to hire an SEO agency to do AEO?
Not first. Most of AEO is a clarity-of-message problem your agency can't fix for you, because they don't know who you're actually for. Get the four questions answered plainly in your own copy, then bring in technical help for schema and structure. Doing it in the other order is paying to optimize a message that isn't built yet.
Does AEO replace Google rankings?
No. It sits alongside them. Plenty of buyers still search Google, and AI Overviews now sit on top of those results too. The same clarity that earns an AI recommendation also strengthens traditional SEO, so you're not choosing between them. You're fixing the message that feeds both.
How long until this actually matters for my business?
It already does. Buyers are using AI to vet contractors right now, and you can watch them admit it in homeowner forums. You won't see it in a dashboard, because the model recommends a competitor silently. There's no "lost to ChatGPT" line in your numbers. The absence is the symptom.
What's the single biggest mistake contractors make with AEO?
Treating it as a technical checklist when it's a clarity problem. They add schema markup to a homepage that still can't say who it's for, then wonder why nothing changes. The machine needs a clear answer to recommend, and you can't mark up an answer you never wrote.
Ready to look at the architecture honestly?
Book the conversation. We'll tell you what we see, and whether the work we do fits where you are.
Ready to apply this?
One working call. We'll tell you what we see — and whether the work we do fits.