Is your shop named when drivers ask an assistant? How to check, job by job

What does being named mean for a repair shop?

An answer to "who can fix my brakes in [town]" does one of four things with your shop: leaves it out, lists it among others, puts it forward as the one to call, or names it first. Our methodology does not weigh them against each other. It records whether the shop was named and whether it came first, and its visibility score is the share of the twelve questions in which the shop was named, between 0 and 100, with first place and the other shops reported beside it.

Two things matter to a shop that a single score hides. One is the job. A shop can be named for oil changes and missing for transmission work, and an average of the two describes neither. The other is what the answer says. An answer that names the shop but gives last year's Saturday hours, or says it works on a make it turns away, produces a call the shop cannot help or a drive to a locked door.

So a shop's reading is kept per question, with the answer's text beside each verdict. The mention rate, the share of answers that name you at all, is the coarsest number. The useful part sits under it: which jobs you are named for, which you are missing from, who is named instead, and whether the facts are right.

Why a shop's questions are built from its jobs

A panel of "auto repair near me" questions measures one question many times. The questions worth measuring come in the shape of the car's problem: a squeal when braking, a light that stays on, a failed inspection, an air conditioner blowing warm. Each is a sample of the work you want, in the words drivers use for it, with the place named the way locals name it.

Start from your repair orders. List the jobs that make up most of the work and the ones you want more of, then write each the way a driver would type it. Add the makes you specialize in, the vehicle types you take (hybrids, EVs, diesels, vans), and the time words drivers add in a hurry: today, open Saturday, while I wait.

Leave out questions about fixing the car yourself or what a noise means. They deserve an answer on your site, but the answer is not a shop, so they do not belong in a panel that counts shops.

A starting panel for a repair shop

Our panels mix three kinds of question, as the methodology describes: problem questions, asked by a driver who knows no names; comparison questions, where a driver weighs two shops; and questions that name you. For a shop the problem questions carry the most weight, because that is where a driver who has never heard of you either hears your name or does not.

Eight problem questions to four others is a fair split for a first panel; our own panels for paying clients run to twelve questions. Weight the jobs by the work you want, not only the work you have. If transmissions are the growth you are after, they get a question even when they are a sliver of today's repair orders.

What a shop adds to the method

The method is written down already and is the same for every trade. How to measure AI visibility covers the runs and how to tell a real change from noise, and setting up prompt monitoring covers the file the panel lives in and the fields each answer gets. A shop adds four habits to it.

• Put the town in every question the way your callers say it, and never edit a question once it has results against it.
• Ask with web search on, and note which pages each answer cites. A source citation to a directory, a review platform or a competitor shows where the answer's facts about you came from.
• Check every fact the answer states against your own: hours, makes, services, warranty, the address. A wrong fact goes on a list of its own, with the page it came from where the answer cites one.
• Note what the shop was named for, not only whether it was named. A mention for oil changes and a mention for diagnostics are different results.

If you would rather not run it yourself, monitoring is a narrower panel run for you: twelve questions put to Claude alone once a month, with every answer kept, at $290 a month, which you can cancel in any month.

The one large reading we have published, and its limits

In our audit of 9,483 small business websites we ran a narrow version of this across the whole sample. For 9,240 of the businesses, in 40 trades, one buyer question from their own trade and metro went to Claude with web retrieval on, once. 14.9% of the businesses were named. Where Claude answered with a clearly differentiated list rather than a hedge, which it did in 5,928 of the answers, the share was 17.7%.

That is not a repair-shop rate. The audit publishes no figure for the trade, and one question per business on one day, in one assistant, is a direction rather than a measurement of any shop. What it does show is that most businesses were absent from the answer to a plain question about their own trade in their own metro, so a shop missing from most of its first panel is in the common position.

Of every name Claude gave in those answers, 99.4% were businesses and 0.6% were directories or aggregators. When we inspected the sites of the businesses that were named, they were the ones with answer-shaped content and a consistent identity. For a shop we read that as the jobs, prices, warranty and hours stated in text, and one name, address and phone number used everywhere, which is the list the repair shop overview sets out.

Reading a shop's first results

Read the baseline job by job against your own site. If drivers asking about hybrids never hear your name and your site never says you work on hybrids, the answer had nothing of yours to use, and the fix is a sentence, then a page; specialty pages covers the page. Where an answer cites the shops it names, open their pages and note what they state that yours does not: a diagnostic fee, a warranty term, Saturday hours.

Then read who you are up against, question by question. A driver asking about a timing belt on a [make] may be told about the dealer's service department, an independent specialist and a chain; read what the answer says about each. Your mentions as a share of every shop named is your share of voice; the other names tell you whose pages to read next.

Last, list every wrong fact. A wrong closing time or a make you do not take costs the driver a wasted trip and costs you a first impression. Where the answer cites its source, the wrong fact traces to a page: your own, a directory's or an old listing's. The Google listing article covers holiday hours and the other fields shops forget to update.