The Questions Customers Ask AI Before They Ever Call You

There is a change in how people research a purchase that has not shown up in most businesses' websites yet.

When somebody typed into Google, they typed like a person addressing a machine. Three words. "Plumber Brantford." "Physio near me." The search box trained everybody to strip out context.

An assistant invites the opposite. People type whole situations, because it feels like asking a person. And the questions they ask are the ones they would have been slightly embarrassed to ask you directly.

What they are actually asking

Roughly four kinds, and each one is a content opportunity most local businesses have left completely open.

The diagnosis question. "My furnace is running but the air coming out is cold, what does that mean and is it expensive." They are trying to work out how much trouble they are in before they call anybody.

The price sanity check. "Is $450 reasonable for replacing an outdoor tap in Ontario." They have a quote, or they are about to get one, and they want to know whether they are being taken advantage of.

The permission question. "Do I actually need a physiotherapist for a sore knee or will it sort itself out." They are deciding whether the problem justifies spending money at all.

The vetting question. "How do I know if an electrician is legitimate, what should I check." They have a name and they are checking whether to trust it.

Notice what these have in common. Not one of them is "plumber near me." Every one of them happens before the search you have been optimizing for.

And the business that gets named while the assistant answers a diagnosis question has effectively been recommended before the shopping started.

Why your site probably does not answer any of them

Go and look at your own website with those four questions in mind.

Most service business sites are built around what the business wants to say. Services offered. Years in business. Area covered. A contact form.

They are not built around what a worried person at 10pm actually wants to know. Which is: what is wrong, how bad is it, what will it cost, and can I trust whoever comes.

The gap is not a design problem. It is that nobody wrote the answers down, usually for one of two reasons.

The first is that it feels too basic. You have explained what a cold-air furnace usually means four hundred times and it bores you. That familiarity is exactly why you should write it, because it is not boring to the person experiencing it for the first time.

The second is price nerves. Owners worry that publishing a range costs them the chance to sell. In practice it filters out the people who were never going to buy and reassures the ones who were, and it is the single most quotable thing you can put on a website.

Where the answers should live

Do not put all of this on one enormous FAQ page. Split it.

Give the big diagnosis questions their own pages. "Why is my furnace blowing cold air" is a page. It gets found in ordinary search, it gets read by assistants, and it demonstrates competence better than any about-us paragraph. Two or three of these are worth more than twenty generic blog posts.

Put the pre-booking questions on the relevant service page. Cost range, timescale, what happens on the first visit, whether you charge for a quote. Right there where somebody is deciding, not buried in a separate FAQ.

Keep a genuine questions section for the rest. Ten to fifteen real ones. Use proper question and answer markup so the structure is machine readable, and answer in the words a customer would use rather than trade language.

The rule that makes all of it work: answer in the first sentence, then explain. Do not build up to it. A page that opens with "there are many possible causes of reduced airflow" gives an assistant nothing to lift. A page that opens with "a furnace blowing cold air is usually one of three things, and two of them are cheap to fix" gives it a quote.

How to find your own list in twenty minutes

You do not need keyword research for this. You need to write down what people already ask you.

Ask whoever answers the phone to keep a note for two weeks of every question asked before booking. That list is your content plan and it will be more accurate than any tool.

Then check your own inbox and your text messages for the questions customers ask after hours. Those are the ones your website should have answered.

Then ask an assistant the four question types above about your own trade in your own town, and see what it says and who it names. If it is naming a competitor while answering a question you answer daily, that is the gap, stated plainly. Running a proper check of who gets named turns that hunch into something you can track.

The part that compounds

This work has an unusual property. Every one of those answer pages keeps earning, because the questions do not change. A page explaining what a cold-air furnace means will be just as relevant in three years.

It also does a second job that has nothing to do with search. It answers the question before somebody phones to ask it, which means the calls you do get come from people already halfway to booking.

Common questions

What kinds of questions do people ask AI about local services?

Mostly four kinds: what is wrong with something, whether a price is reasonable, whether the problem is worth paying to fix, and how to tell whether a business is trustworthy. All of them happen earlier than the searches most businesses optimize for.

Should I put prices on my website?

A range, with an explanation of what moves it, is usually better than nothing. It filters out people who were never going to book, reassures the ones who were, and gives search engines and assistants something concrete to quote about you. Very few local competitors do it.

Is a FAQ page still worth having?

Yes, with two conditions. The questions have to be real ones customers ask rather than invented ones, and the biggest questions deserve their own pages rather than a line in a list. Use proper question and answer markup so the structure is readable.

How do I find out what customers ask before booking?

Have whoever answers the phone write down every pre-booking question for two weeks. That list beats any keyword tool because it is your actual market rather than an estimate.

How long should each answer be?

Give the direct answer in the first sentence, then two or three short paragraphs of explanation. The first sentence is the part that gets quoted, so it has to make sense on its own without the rest.