One More Google Review Star Is Worth $30K a Year

There is a list you keep in your head of customers you would feel fine asking for a Google review. The retired couple on Oakwood who brought the crew lemonade. The guy whose deck you redid for free. Everybody else gets skipped, because why go poking at it. That list is the most expensive habit in your business, and since December it has a second problem: the FTC sent warning letters to 10 companies over its Consumer Review Rule, which has been on the books since October 2024 and is now being enforced. Ten minutes from here you will have a dollar figure for the reviews you are not collecting and a same-day text that goes to every customer instead of the six you picked.

The setup: One automatic text after every completed job, sent to all of them, with no screening question in front of it. What the gap is worth: On $613,000 a year in work, the low end of the research on star ratings puts one star at $30,600. Start here: Open your profile and write down two numbers, your star rating and your review count. Most owners know the first one and have never looked at the second.

Your Google rating is fine. Nobody can see it.

Here’s an illustrative example, built from typical numbers in the trade rather than one specific business: a two-crew exterior cleaning company, 78 jobs a month, $655 average ticket, about $613,000 a year. The owner has 17 Google reviews and a 4.6 average, and he is quietly proud of that 4.6.

He built it by asking eleven people out of the last nine hundred.

BrightLocal surveyed 1,002 US consumers for its Local Consumer Review Survey 2026 and found that 97% read reviews before choosing a local business and, more to the point here, 47% won’t consider a business with fewer than 20 reviews. Seventeen reviews puts him under the line for roughly half the people who find him. The 4.6 never gets evaluated, because the count disqualifies him before anyone reads a word.

What one star is actually worth, with the fine print attached

Michael Luca at Harvard Business School matched Yelp ratings against Washington State tax records for Seattle restaurants from 2003 to 2009. His paper, Reviews, Reputation, and Revenue: The Case of Yelp.com, found that a one-star increase drove a 5 to 9 percent revenue increase, and that the entire effect belonged to independent restaurants. Chains got nothing.

Restaurants, Yelp, Seattle, twenty years ago. Applying it to exterior cleaning is an inference and I’m labeling it as one. It’s still the best-identified number anybody has here, because Yelp rounds to the half star: a 4.24 displays as four, a 4.25 displays as four and a half. Same business, same quality, different number on the screen, measurably different revenue.

Run the bottom of his range against $613,000 and you get $30,600 a year. The top gets you $55,000. I will keep using the bottom, and you should too, for the same reason you should make a vendor show you the study behind their slide.

Here is the part of that paper nobody quotes, and it should change your Monday. Luca tested whether review volume changes how much the rating matters. It does: a rating change has 50% more impact once a business has at least 50 reviews, compared to one with fewer than 10.

Read that against 17 reviews. The stars our man worked for are running at a discount, because the number next to them tells a shopper not to trust the average yet. Volume is what switches the rating on.

Hand-picking is now two separate problems

Google has banned it outright for years. Its prohibited content policy lists “discouraging or prohibiting negative reviews, or selectively soliciting positive reviews from customers” under fake engagement. The penalty isn’t a fine. It’s review removal or suspension of the profile that half your phone calls come through.

The FTC part is newer and worth getting right. Review gating is the habit you already have, with a name on it: screening customers first and only asking the ones you expect to give you five stars. The Consumer Review Rule doesn’t ban it outright, whatever you’ve heard. It bans fake reviews, insider reviews without disclosure, and paying for reviews that carry a particular sentiment, and its suppression section covers reviews on a website you control, not which customers you chose to ask.

What the agency does say sits in its own staff Q&A on the rule, answering this exact question. Can a business ask for reviews only from customers it thinks are happy? “The rule does not contain a specific prohibition against such conduct. But this practice could violate the FTC Act.”

That is live ammunition without a rule citation attached. The December 22 letters show the posture: ten companies, five days to respond in writing, civil penalties of up to $53,088 per violation. The conduct flagged was fake reviews and incentives paid for five-star sentiment, a step past where you are, on the same road.

In plain terms: screening customers before you ask is against Google’s rules today, the FTC says it may be illegal, and it’s also the reason you only have 17 reviews. Three arguments, one fix.

Send a text, not an email

The fix is boring. After every completed job, same day, one message with a direct link to leave a review. Every customer, no exceptions list. No survey in front of it, no “how did we do, one to five” fork.

Text beats email badly here, though the numbers floating around deserve a squint. Review platforms repeat that SMS converts at three to five times email for review requests, with near-98% open rates against 20-30% for email. No study under it, no sample, no methodology, and it comes from companies that sell texting. Even the honest voices in that industry concede the 98% figure is estimated from delivery data and lock-screen analytics rather than measured.

Treat it as direction. The direction isn’t controversial: the message arrives on the device the customer is already holding when your crew pulls out of the driveway, and it takes one tap. An email lands in a tab they open Thursday.

Two things make or break it. Send it within a couple of hours of finishing, while the clean driveway is still the most interesting thing that happened today. And point the link straight at the review box, not at your website, not at a “leave us feedback” page that asks a question first. Every screen between the text and the star rating costs you reviews, and one of those screens is the illegal one.

Yes, you will collect some bad ones

This is the actual objection, so let’s answer it instead of talking around it. At 78 jobs a month, you’ll ask 936 people this year, including the ones you argued with. If even one in eight writes something, that’s roughly 117 reviews, and you cross 50 in about four months, right where Luca’s rating sensitivity kicks up.

A one-star inside 117 reviews moves your average by a rounding error, and BrightLocal found 74% of consumers weight reviews from the last three months most heavily. Volume and recency are what make a bad review survivable. Hoarding 17 hand-picked ones is what makes a single angry customer expensive.

Reply to the bad one within a week, specifically and without arguing, and it does more for a browsing customer than five more five-stars. That workflow, plus using AI to draft replies without sounding like a robot, is the other half of this system.

What to do this week

1. Write down three numbers, four minutes. Open Google Business Profile on your phone: star rating, review count, date of your most recent review. If that last date is older than a month, automation is the whole conversation. Then multiply monthly jobs by average ticket by 12 by 0.05 and put that dollar figure beside them.
2. Count the last 90 days of finished jobs nobody asked. In Jobber, Housecall Pro, ServiceTitan, or whatever holds your job list, filter completed jobs since late June and compare that count to the reviews you got in the same window. The gap is the list you have been hand-picking from, and seeing it as a number is what makes step three happen.
3. Turn on the automatic request and set it to fire on every job. Jobber’s Reviews tool sends a text and email when a job closes, adds up to two reminders, and holds messages overnight. Housecall Pro sends requests on job completion in its marketing tools. Whatever you run, check that the trigger is job completed or invoice paid and that no filter, tag, or approval step sits in front of it. While you are in there, confirm what texting consent your software captured at booking, the same question that governs any customer text you send.

Then leave it alone for ninety days and look again. The version of this that works is the one where nobody in your shop ever decides who deserves to be asked.