Something changed in how people look for suppliers, and most businesses have not noticed yet.
A year ago someone needing a web designer in Calgary typed that into Google and worked through the results. Today a growing share of them open ChatGPT and ask “who should I hire to build a website for my dental practice in Calgary,” read the three names it gives back, and start there.
Those three names came from somewhere. Understanding where is worth more than most of what passes for SEO advice right now.
How do AI tools decide which businesses to name?
They look for businesses that other sources have already described clearly, and they strongly prefer information they can state as fact.
The useful way to picture it is the difference between a phone book and a newspaper. Search engines were a phone book. Being listed was most of the battle, and the listing was mostly your name and category.
AI answers work more like a journalist writing a piece on deadline. They are not reading a directory, they are gathering what has been said about you and turning it into a sentence. What gets quoted is whatever is clearest, most specific, and easiest to attribute.
So the question stops being “am I listed” and becomes “is there a clear sentence about me somewhere that a machine can lift.”
Is this different from normal SEO?
It overlaps, but it rewards different things.
Traditional search rewards authority and links. An AI answer rewards being quotable. Those are not the same. A page can rank third for a term and never be quoted, because nothing on it can be extracted as a clean statement.
Three differences matter in practice.
Structure beats decoration. AI tools read the underlying structure of a page, not the design on top of it. A beautifully designed page with vague headings gives them nothing to hold.
Specificity beats polish. “We deliver bespoke solutions tailored to your needs” cannot be quoted, because it says nothing. “We build WordPress sites for dental practices, typically in six weeks, from our studio in Dehradun” can be quoted directly.
Being described elsewhere matters more than ever. What other sites say about you carries real weight, because a model trusts corroboration.
What can I actually do about it this month?
Five things, roughly in order of effort against return.
Write your headings as questions people actually ask. Then answer each one in the first sentence underneath. This is the single highest return change you can make. AI tools lift answers, and a heading that asks the exact question with the answer directly below is the easiest thing in the world to lift.
So “How much does a website cost in Calgary?” works. “Our Pricing” does not.
Add structured data to your site. This is a small block of code that states, in a format machines read directly, what your business is, where it operates, what it offers and how to reach it. It removes guesswork. Most sites do not have it, and it takes a developer an hour.
Say the specific, boring facts out loud. Where you are based. What you actually do. Who you work with. How long things take. Businesses tend to write around these because they feel obvious. They are exactly what gets quoted.
Get described accurately in a few other places. Your Google Business Profile, your LinkedIn, a directory or two that matters in your industry. Consistency across them matters more than volume.
Answer the questions your customers ask before they buy. Every one of those is a question someone is now typing into an AI tool. If your site holds the clearest answer, you become the source.
Does this mean my website matters less now?
The opposite, and this is the part most people have backwards.
If AI answers are increasingly how people find suppliers, then the thing those answers are built from becomes more valuable, not less. Your website stops being only a destination and starts being a source.
There is a real change in traffic, and it is worth being honest about it. Some searches that used to send you a visitor now get answered without a click. Usually the simple ones. Opening hours, a phone number, a definition.
What has not changed is that nobody hires a web designer, or a dentist, or an accountant, straight out of an answer box. They get a name, then they go and look. The visits you lose are the ones that were never going to buy. The visits you keep are the ones that were already half convinced, arriving with your name given to them by something they trust.
That is a better visitor than the one you lost.
How do I know if it is working?
Imperfectly, and anyone claiming precise measurement here is overselling.
There is no ranking report for AI answers yet. What you can do is check directly. Ask ChatGPT, Gemini and Perplexity the questions your customers would ask. Ask them once a month. Note whether you are named, and what they say about you when you are.
That last part is worth attention on its own. Sometimes a business is named but described wrongly, and a wrong description is a fixable problem once you know it exists.
Also watch your referral traffic in Analytics. Visits from chatgpt.com and perplexity.ai now show up there, and while the numbers are small today, the trend line tells you something.
Is it too early to bother with this?
It is early, which is exactly the argument for doing it.
Almost nobody in any local market has done this deliberately yet. The work is not expensive, it mostly overlaps with writing clearly, and the businesses that do it now become the ones that get quoted while everyone else is still arguing about whether it matters.
There is a version of this from twenty years ago. Search engines existed, most businesses ignored them, and the few that took them seriously spent the next decade being found first.
The same window is open now, and it will not stay open long.
We build this into every site we make, from the structured data to the way the headings are written. Tell us what you are working on and we will tell you where you currently stand.
