How do you sell AI visibility to a local business client?

The live check

Asking an assistant, in front of the client, the question their customers actually ask. It takes about a minute and it is the entire pitch.

The gap

The distance between who the assistant names and who your client is. It is either there or it is not, and the client can see which without being told.

Holding the position

The work when a client is already named. Harder to sell than a gap and easier to renew, because the number now has somewhere to fall.

The pitch is one question, asked out loud

Do not open with the category. Open the assistant, type the question their customers type, and read the answer out. For a trade in a metro area that looks like "best HVAC company near Austin", phrased the way a person would phrase it rather than the way a keyword tool would.

Then stop talking. A paragraph naming three competitors and not naming them does the work that a slide about the future of search cannot. They are not being told they have a problem. They are watching one.

Do it live if the room allows it. A screenshot invites the suspicion that you picked the run that suited you, and that suspicion is fair, because the same question asked twice can come back differently.

Be ready for the answer to go the other way

Sometimes the assistant names them. That is not a failed pitch, and treating it like one is how you lose the room. It is a different conversation: they are the recommendation today, and today is all that one check establishes. Ask what they did to get there. If nobody in the room can name it, the position is unexplained, not unmaintained.

It is a defense rather than a rescue, and a defense is easier to renew than a rescue is to repeat. It is also the more honest sale, because a business that is already named is the one with the most to lose from ignoring this.

Why the demo lands when the deck does not

A local business owner has been sold visibility for as long as they have had a website. Every version of the pitch sounded like the one before it, and several of them did not work. Nothing you assert about AI search will land against that history.

What lands is an artifact from their own world: their question, their city, their competitors, in an answer a real customer would have received. You did not make a claim, so there is nothing for them to discount.

The objections that actually come up

"Nobody uses AI to find a plumber." The honest answer is that you do not know the share in their town and neither does anybody else, including the vendor selling you the tool. Say so. Then point back at the screen: this is what their customer sees when they do ask, and the cost of finding out later is that a competitor has been the answer for a year.

"We already rank first on Google." True, and mostly irrelevant to what just happened on the screen, because an answer engine does not show a list of ten and let the customer choose. The good news is worth giving them: the work that earned that ranking overlaps with what gets a business retrieved and named in an answer, so this is a second scoreboard on work they are already paying for rather than a second body of work.

"Can you guarantee we will be named?" No. A guarantee would mean controlling how an assistant weighs an answer, and that control sits with the model vendors. Nobody outside them can see it, so nobody outside them is in a position to promise it. What you can commit to is the measurement, the specific work behind each gap, and a record of what changed. Say the refusal out loud rather than talking around it. It costs nothing, and it is the sentence that makes the rest of what you said believable.

How to frame the price

Price the outcome, not the software. If you itemize the tool on the proposal, you have invited a conversation about the tool, and that is a conversation you cannot win: the client will look up a price, compare it to your fee, and see a markup instead of a service. Nothing stops you telling them what you use if they ask, and you should.

Anchor to a number they already own. Ask what an average job is worth to them and roughly what share of inquiries close. That arithmetic is theirs rather than yours, and it is the version of the value case they will believe. Keep it as a comparison rather than a forecast, and do not put a projection in writing.

Sell the baseline before you sell the retainer. A one-off audit that establishes where they stand today is a much smaller yes than an ongoing fee, it is billable, and it produces the document the retainer conversation runs on. It also has an end, which is most of why it is an easier yes.

What to deliver every month, so the renewal is boring

• The number, with the counts behind it, so the client can see what it was calculated from rather than being asked to trust a percentage.
• The per-engine split, because an average can hide being absent from one assistant entirely, and clients notice that omission later rather than never.
• Who was named instead of them, which makes the next month of work self-evident without you having to argue for it.
• What you changed since the last report, and whether the number moved after it. Some months the honest entry is that it did not.
• What is next and why it is next, so the client is approving a plan rather than receiving a bill.

A client who has read that report for six months does not need to be re-sold. They already know what they are buying, which is the point of sending it.

The one thing that ruins this pitch

Overselling the first month. The demo is dramatic and it is tempting to let the drama carry into the contract. Do not. Tell them at signature that the first weeks establish a baseline, that one reading is not a trend, and that some months will be flat or worse.

The client who was told that in month one reads month four correctly. The client who was not told is the one who cancels in month three, and both clients received exactly the same results.