How to Get More Google Reviews Without Breaking the Rules

Most businesses don't have a review problem. They have a request problem. The work is good, the customers are satisfied, and nobody ever asks them — or somebody asks in a way that gets the reviews filtered out three weeks later.

This page is about the request system. Where the link comes from, when the ask goes out, what medium carries it, who on your staff does it, what they're allowed to say, and how many you need per month to keep your profile from looking abandoned. If you want the ranking side of the profile — categories, service lists, photos, what Google switched off — that's on the Google Business Profile page. This one is the plumbing.

I'll say up front where my numbers come from, because you should ask that of anyone writing about this. The response rates and the filtering losses below are what I observe across my own client accounts, where I either send the requests myself or built the system that sends them. They are not survey figures. Where I cite a survey or a policy, it's linked, and every policy claim on this page points at Google's own documentation rather than at somebody's blog post about it.

Start here, because almost every review program I inherit has one of these baked into it, and the owner has no idea.

Google's position is not that asking is suspicious. Google publishes its own tips for getting more reviews and tells owners to remind customers. The rules are about how you ask. They live in the prohibited and restricted content policy, and profiles that break them can have new reviews blocked or existing reviews unpublished for a period — Google spells out those restrictions for policy violations in a separate document that is worth five minutes of your time.

The last two arrived on April 17, 2026, and they retired two tactics that half the industry was still selling. Review contests for technicians are out. "Mention Kyle so he gets credit" is out. If your dispatch software has a review leaderboard in it, turn the leaderboard off.

There's a second body of law underneath all this. The FTC's rule on consumer reviews and testimonials took effect in October 2024 and carries civil penalties. Google can hide your reviews. The FTC can fine you. Buying reviews, writing them yourself, and suppressing negative ones are federal problems now, not just platform problems.

One thing worth saying out loud: the compliant version of this is also the version that survives. Reviews collected on a shop tablet come from one device and one IP address, and Google's filters treat them accordingly. Reviews collected through a gated form arrive in a suspiciously clean five-star pattern. Plain asks sent to a customer's own phone are the ones that stick.

You need a short URL that drops a customer directly onto the review form with the stars already showing. Not your homepage. Not "search for us on Google." One tap, five stars, a text box.

Google generates it for you, and the path changed when the old dashboard went away. Here is the current one, from Google's own instructions:

1. Sign into the Google account that manages the profile. Not your personal account, and not the account your last marketing company used — if you can't get in, that ownership problem is the actual first job.
2. Search your business name in Google, or open your Business Profile directly. The management panel appears above the public listing.
3. Select Read reviews, then Get more reviews. Depending on what Google is showing you that week, the tile may read Ask for reviews. Same destination.
4. Select Copy to take the link. To take the QR code, right-click it and save the image.
5. Paste the link into a note, a document, your invoicing software and your phone's text shortcuts. Everybody who asks for reviews needs it in their pocket.

Two details that catch people out. The QR code can only be generated in a desktop browser — you cannot produce it from the phone app. And if you manage more than one location, each location has its own link. Sending the wrong one is how reviews for the Oro Valley shop end up on the Marana listing, and Google will not move them.

Don't go hunting for a short name. The old g.page/YourBusiness custom URL is finished. Google's help page now says flatly that you can no longer create or edit short names. Existing ones still resolve, but they're no longer displayed and you can't claim a new one. If a vendor is charging you to "set up your branded review URL," that's a service that no longer exists.

If you want a link you control — one you can point somewhere else later, or track clicks on — put your own redirect in front of Google's. Something like yourdomain.com/review that forwards to the Google URL. Then the card you printed last year still works after Google reshuffles something. I set this up on client sites as a matter of course; it's a five-minute job on the website side and it has saved two of my clients a reprint.

Timing moves the response rate more than the wording does. There is a window after a job where the customer feels relief, and it closes fast. Miss it and you're asking a stranger for a favor.

The window is not the same in every trade, and this is where generic advice fails. Three shapes I see:

Three rules that hold across all of them:

• Ask after they've left, not while they're standing there. This is both a policy matter and a practical one. Google's policy prohibits pressuring customers to review while on the premises. And a review written in your lobby under your eye is short, generic and easy for a filter to spot.
• Ask once. Then one reminder, three or four days later, if nothing came in. Then stop. A third ask costs you goodwill and buys nothing.
• Don't ask when you already know the answer. Not because you're screening — you're not allowed to screen — but because a customer with an unresolved complaint needs a phone call, not a review link. Fix the problem. Ask later, or don't.

That last point is the line people get wrong. Sending the link to everyone and sequencing the unhappy ones through a resolution call first is fine. Building a form that asks "how did we do?" and only shows the Google link to people who click four or five stars is review gating, and it's a policy violation.

This is the part no national guide will tell you, and it's the reason I don't hand a client a flat monthly target and walk away. Southern Arizona has a seasonal shape, and review flow follows it.

• June through September is your volume. AC failures, monsoon roof and irrigation damage, wind cleanup. These are relief jobs with the sharpest emotional peak of the year, and they produce your best response rates. This is when the pile gets built.
• October through April brings the winter residents. In Green Valley, Oro Valley and Catalina, a large share of your customers are seasonal and older, and that group leaves fewer reviews from a text link. Response rates drop. Print matters more here, and a phone follow-up from a person works where a text does not.
• May and late September are the dead spots. Not because customers are unhappy — because job volume dips between seasons. If you only ask when it's busy, your profile shows a two-month gap twice a year, and that gap is what a prospect in October sees.

The practical move is counter-intuitive: ask hardest in the slow months, not the busy ones. If July gives you eight reviews and October gives you two, your average still clears the target, but the newest review in mid-November is six weeks old and the average is doing you no good. Averages don't show on your profile. Dates do. That one adjustment is the change I make most often on Tucson accounts.

Both work. They work in different situations, and most businesses pick one out of habit rather than fit.

The practical answer for most local service businesses here: text as the primary, email as the fallback for anyone who didn't give you a mobile number, QR code on the invoice so the request exists in a second place. The real exception is the winter-resident accounts above, where the age of the customer base flips the ranking and a printed card outperforms a text.

Text discipline matters. Keep it under two lines. Identify yourself in the first four words. Send it between roughly 9am and 7pm, and don't send at 6am because that's when your dispatcher runs the report.

QR codes get printed and then placed where nobody is holding a phone. That's the whole failure mode. A code works when the customer's phone is already in their hand and something has just been completed.

Placements that earn their printing cost:

• On the invoice or receipt, next to a printed line of text that asks. The code alone does nothing.
• On a leave-behind card the technician hands over with the paperwork. It goes on the fridge, which means the ask survives a day or two past the visit.
• At the payment point, where the customer is already looking at their phone to tap or pay. Not a sign at eye level across the room.
• On the back of a business card, which costs nothing extra on the next print run.
• Inside the email signature of whoever does your billing, as a link rather than an image.

Placements that don't: a laminated sheet taped to a wall behind the counter, a sticker on the front door, a banner in the waiting room, the back of a vehicle in traffic. And a tablet on your counter with the review form open is not a placement — it's the thing Google specifically told you not to do.

One more: if you print the code, print it big enough and test it from four feet away with a cheap phone in bad light. I've seen a run of 5,000 invoices go out with a code that only scanned at six inches.

The April 2026 policy update rewrote this section for everyone. Your staff can still ask. What changed is the machinery around the asking.

What you can't do anymore: set a review target per technician, run a contest, tie a bonus to review count, or coach anyone to request that the customer name a person or mention a specific service. Google names both the quota and the staff-name request as prohibited. If a customer decides to name their tech on their own, that's fine and it's a great review. Asking for it is the violation.

What you can do: teach a short, plain handoff. Three that work:

Notice what's missing from all three. No star count. No "five-star." No suggested wording. No mention of a name. No offer. The moment you tell a customer what to write, you've given Google a reason to filter it and given yourself a compliance problem.

Train it the way you train anything else in the field: say it out loud in the truck until it stops sounding like a script. The technician who mumbles a rehearsed line gets fewer reviews than the one who says it like a person.

This is the part nobody does, and it's the part that decides whether the program works. "Get more reviews" isn't a target. A number per month is.

Work it backwards from what a customer sees. Two calculations.

Calculation one: keeping your recent history full

When somebody opens your reviews, Google loads roughly ten at a time. Those ten are what your reputation is, as far as that customer is concerned.

Google sorts that block by Most relevant by default, not by newest. You don't control that order. What you control is the dates on the reviews available to be shown, and what happens the moment a skeptical customer switches the sort to Newest — which is exactly what a careful buyer does.

So: at a steady rate of r reviews per month, your trailing ninety days holds about 3r reviews. To fill that first block of roughly ten with reviews from the last three months, you need r of about three and a half. Round it up.

Three to four a month is the number. That's the whole answer, and here's why it matters that it's three to four and not one: BrightLocal's 2026 consumer survey found that 74% of consumers look for reviews written in the last three months, and 32% want something from the last two weeks. At one a month, arrival is lumpy enough that you'll routinely have three-week gaps. At four a month, you average one every seven or eight days and the two-week window stays covered even in a slow stretch.

The same survey found 47% of consumers won't consider a business with fewer than twenty reviews at all. If you're under twenty, your first job is volume; get to thirty and then settle into the monthly rhythm.

And the part that surprises owners: the number does not drop because your lifetime total is high. Three hundred reviews with nothing in the last eight months reads worse than forty reviews with one from last Tuesday. Recency is a flow, not a balance.

Calculation two: how many asks that takes

Reviews per month equals asks per month times your response rate. Response rate on a plain text ask, sent same-day with a working link, runs somewhere between fifteen and forty percent depending on trade and how good the moment was. Here's what that means for ask volume:

Read the top row carefully, because that's the one that hurts. A business doing ten jobs a month cannot hit four reviews a month unless nearly half of everyone asked follows through. If that's you, the ask has to be excellent — texted the same day, from a person, at the right moment — and you should expect to be at two or three most months. That's fine. Two or three a month, held for a year, is thirty reviews and a profile that never looks dead.

A business doing forty jobs a month has the opposite problem: plenty of raw material, and it usually asks about six of those forty customers. Fixing the process, not the satisfaction, is what moves that account.

The diagnostic is simple. Count your completed jobs last month. Count your new reviews last month. Divide. If you're under ten percent, you have a request problem, and no amount of doing better work will fix it.

Calculation three, for the honest version

Subtract for filtering. Some percentage of what you collect will never post — wrong account, no Google login, caught by a filter, written and abandoned. Across my client accounts it's usually a small slice, but it's not zero, and it's worst on the accounts with an older customer base. Build in a margin: if you need four, aim to generate five.

You don't need software for this. You need a spreadsheet with five columns and a monthly habit.

• Jobs completed — from your invoicing, whatever number you already have.
• Asks sent — the count of texts and emails that actually went out.
• New reviews — from the profile, counted on the first of the month.
• Response rate — new reviews divided by asks sent.
• Days since the newest review — the single number that tells you whether your profile looks alive today.

That fifth column is the one I check first on every account. If it's ever above twenty-one, the program has stopped and nobody noticed. It's also the number that makes the case for the whole exercise, because it moves within a month of fixing the process.

Review platforms automate the sending and they're worth the money above a certain volume. Below about forty jobs a month, a saved text template and a habit beat a subscription. What software will not do is make a bad ask work, and it will happily help you build a gated funnel that gets your reviews pulled.

• The link is wrong or dead. Somebody saved a URL from an old dashboard, or from a short name that no longer displays. Test your own link on a phone you aren't signed in on, today.
• The ask lives with one person. The office manager sends them, the office manager goes on vacation, the program stops for two weeks. It has to be part of closing a job, not one person's side task.
• It runs as a campaign. Forty reviews in a month, then nothing for a year. That pattern reads as a review push to customers and to Google's filters both.
• Nobody replies. Not a request problem, but it undercuts everything upstream. Reply to all of them, briefly, in your own words.
• Somebody buys some. It shows up in the pattern, the FTC rule makes it a federal matter, and the cleanup costs more than the shortcut saved.

I ask the way I've described above: once, after the work is done, with no script and nothing offered in exchange. These four are unedited.

More of them, with the ranking numbers behind them, are on the case studies page.

Related reading: what still works on a Google Business Profile in 2026 · local SEO services · case studies · more articles