The Reviews That Do the Selling for You

A customer picks up her car at an independent repair shop on a Thursday afternoon. The brakes were grinding, the shop found it was the rotors too, called her before doing the extra work, and finished a day early. She tells the service writer it’s the best experience she’s had with a mechanic. Then she drives off.

The shop’s Google profile has 41 reviews. The newest is seven months old. A one-star review from April, about a misunderstanding over a quote, still sits there without a reply. The chain shop two miles away has a few hundred.

The difference isn’t the quality of the work. It’s that one business asks every customer, and the other hopes they’ll think of it on their own.

I’ll be honest about where I sit. I haven’t run a repair shop, a salon or a restaurant. I spent almost 18 years running a digital agency, Infographic World, and these days I use AI to help businesses fix the things that eat their week. What I do know is how small, repeatable tasks fall through the cracks when everyone’s busy. Asking for a review is the textbook example.

Why reviews matter more than they used to

People have read reviews before choosing a business for years. Two things have grown around that habit.

Google says plainly that reviews feed its local results. Its page on improving local ranking says more reviews and positive ratings can help a business’s local ranking. It also lists responding to reviews among the things an owner can do.

And more people now ask an AI assistant for a recommendation instead of scrolling a map. When an assistant describes a local business, it can only work with what’s been written about that business somewhere, and reviews are a large share of that writing. A review that says “they fixed my brakes, called before adding the rotors, and finished early” gives an assistant, and a reader, something specific to repeat. “Great service” doesn’t. I’ve written more about how AI assistants choose local businesses elsewhere.

None of this guarantees a ranking or a recommendation. Nobody can promise either. Reviews are one input among many, but they’re an input you can actually work on.

Where the reviews get lost

Follow one happy customer from the counter to the review that never got written.

1. Nobody asks. Staff mean to, but the phone rings or the next customer is waiting.
2. The ask comes at the wrong time. Weeks later in a newsletter, or never. The best moment is soon after the job, while the experience is fresh.
3. The ask is generic. A mass email to everyone with “Review us.” It feels like marketing, so it gets ignored.
4. The link is hard. The customer has to search for the business, find the profile and figure out where the review button is.
5. Replies never happen. The owner means to answer, doesn’t know what to say about the bad one, and it sits.
6. Nobody reads the pattern. Three reviews this year mention long waits on Saturdays. Nobody has connected them.

Each step is small. Together, they leave a good business with a thin, stale profile.

The rules first

Before any system, know what the platforms allow. Google’s Maps content policy is specific, and these are the points that matter for a local business:

• No incentives. No discounts, free items, payments or prize drawings in exchange for a review.
• No gating. You can’t ask only the customers you think are happy, or steer unhappy ones away from posting. Ask everyone.
• Don’t tell customers what to write. The policy says merchants shouldn’t request that specific content be included. You can’t ask them to mention your town, a service or a staff member’s name.
• No pressure on the premises, and no review quotas for staff.
• No reviews from employees or family. Those are conflicts of interest.

The good news is that a plain, personal ask sent soon after a specific job tends to produce detailed reviews on its own. People describe what happened to them. You don’t have to script it.

What an AI-assisted review routine looks like

It asks every customer, at the right moment

When a job is marked complete or an invoice is paid, the system sends a short, personal message by text or email to every customer who has agreed to hear from you. Their name, the job, a thank-you, and a direct link to your review form. Google’s own tips page suggests sending customers a Google link or a QR code, which removes the search step.

Something like: “Hi Dana, thanks for trusting the shop with your brakes this week. If you have a minute, a Google review helps other drivers nearby decide where to go. Here’s the link.” One reminder a few days later if they haven’t responded, and then it stops.

The owner approves the wording once. The system just doesn’t forget to send it.

It drafts replies for the owner to approve

Every new review gets a drafted reply that mentions the specifics the customer mentioned. The owner reads it, edits it and posts it. Positive reviews get a short, real thank-you. Negative reviews get a calm draft that acknowledges the problem, skips any private details about the customer, and offers a way to talk offline. The owner always decides what goes out on those.

It summarizes what customers are telling you

Once a month, a one-page summary: what people praise most, what they complain about, and anything new. That’s free market research most businesses never read.

Run the numbers

The formula is monthly customers, times the share you actually ask, times the share who leave a review when asked. That gives you reviews per month.

Say, for the sake of the math, a business serves 300 customers a month and staff ask about one in five, so 60 people get asked. If a made-up 10 percent of them leave a review, that’s 6 reviews a month, or 72 a year.

Now ask everyone the system can reach. Say 90 percent of customers have a phone number or email on file, so 270 get a personal message. At the same made-up 10 percent response, that’s 27 reviews a month, or 324 a year. The only change is the ask.

To run it for real, pull three numbers:

1. Customers served last month, from your point of sale, invoices or booking system.
2. How many of them were actually asked. Be honest. For most businesses it’s a guess, and the guess is usually low.
3. New reviews last month, from your Google Business Profile.

Picture a Monday, before and after

This is a made-up scene. Monday morning, the owner of a small salon remembers she meant to answer that two-star review from last week. She opens the profile, writes three sentences, deletes them, and gets pulled onto the floor. The review sits another week. Over the weekend, 60 clients came through. Two were asked for a review at the desk. One left one.

Same Monday with the routine running. Every weekend client got a personal text Sunday evening. Eight new reviews came in, and eight drafted replies are waiting for approval, including a careful one for the two-star review about a late appointment. The owner edits two of them, approves all eight in ten minutes, and notices the monthly summary flags Saturday wait times for the third month running.

What has to be in place first

• A claimed and verified Google Business Profile, with your direct review link.
• Customer contact details and permission to message them. Texting and email rules vary, so check the ones that apply to your business.
• A clear moment in your process that means “job done,” like a paid invoice or a checked-out appointment.
• An owner, or one trusted person, who approves the ask wording and every reply.

Guardrails worth keeping

• Ask everyone the same way. No filtering by mood.
• Nothing in exchange. Not even a small discount.
• Never script the review. Thank them for the job. Leave the words to them.
• No private details in replies. Especially in health, legal or financial services, don’t confirm details about a customer in public.
• A person approves every reply. Drafts save time. They don’t replace judgment.

How you’ll know it’s working

• Reviews per month, against your pre-start baseline.
• Share of customers asked.
• Response rate on the ask.
• Share of reviews with a reply, and median days to reply.
• Themes in the monthly summary that led to an actual change.

Where I’d start

Two moves this week. Reply to every review from the last 90 days, starting with the unhappy ones. Then pick one service and set up the personal ask for it alone, triggered when the job is marked done. Run it for a month and compare the review count against last month.

If you’d like to see how your business shows up when people ask an AI for a recommendation, try the free AI Visibility Check.