Automate the trigger, not the guest list. Pick one moment every job passes through, send one short request to every customer who passes it, and use the same message each time. Google's policy prohibits selectively asking only your happy customers. Beside is building a system that asks after every job.
Answering the phone for small businesses is what we do, and the review question comes up on almost every line. The mechanics are the easy half. The half that gets owners in trouble is the one every vendor in this category sells as a feature.
The stock of reviews a thousand-dollar customer reads is about three months deep, and it empties itself whether or not anybody asks.
What happens when the request goes to the customer whose job went sideways
Here is why most owners never turn this on. You automate the ask, and at six in the evening the text goes out to the one customer whose job ran two hours long and ended with an argument in the driveway. You have just paid software to invite a one-star review.
The message is not creating the risk. It surfaces it early, while you can still do something. A customer who is annoyed and silent is already telling his neighbor and his group chat, and you will never hear about it. A customer who is annoyed and asked has just told you first.
So have an answer ready for that reply. Call him back the same day rather than texting, fix the thing or credit it, then leave the review alone. Do not offer him anything to change it or take it down, which is its own violation. Leaning on him to remove it is where the federal rule starts to apply. A reply saying you called and fixed it does more for the next reader than a missing one-star would.
Why hand-picking who gets asked is the part that backfires
So owners filter. They send the request to the good ones and quietly skip the rest, and most of the software in this category will do the filtering for them, usually described as routing unhappy customers to a private feedback form.
Google's own policy lists that under what merchants may not do, in one line.
“Discourage or prohibit negative reviews, or selectively solicit positive reviews from customers.” — Source: Google Maps User Generated Content Policy
Those are Google's words, not a reading of them. The consequence is not a fine. Google removes the reviews, and account privileges are on the table after that, which costs you the asset you spent two years building.
There is a version of this that is fine, and the distinction is worth getting right. Approving what the message says, and when it goes out, is ordinary quality control. Using that approval step to decide who deserves to be asked is the same filter with your hand on it.
Two more that owners break by accident: offering a discount or a free service in exchange for a review is prohibited outright, and so is asking your staff to hit a number of reviews or to solicit reviews naming a particular technician (Source: Google's review policy).
Two rulebooks, two different trigger points
| Google's review policy | The FTC's reviews rule (16 CFR Part 465) | |
|---|---|---|
| What it governs | Who you ask, and how | What you display, and threats to remove a review |
| Asking only your happy customers | Prohibited by name | Not addressed |
| Offering a discount for a review | Prohibited | Prohibited when the incentive is conditioned on the review being positive (16 CFR 465.4) |
| Running a reviews section on your own site while hiding the bad ones and implying it is all of them | Not addressed | Prohibited |
| Threatening a customer to get a review pulled | Not addressed | Prohibited |
| What happens | Reviews removed, account privileges at risk | FTC action, civil penalties |
Sources: Google's review policy · 16 CFR Part 465 · FTC, 2024.
The fix costs nothing: pick a trigger, write one message, send it to everyone who passes that trigger. The rest of this page is how to do that well.
Those two get run together constantly. Filtering who you ask is a Google policy problem; the federal rule is about what you display and what you do to people who have already posted.
A review stock is perishable
This is the part that makes the ask a standing job rather than a Saturday project. That three-month window is also closing: 32% of consumers now look for reviews written in the last two weeks, up from 20% a year ago, and 18% are only swayed by reviews from the last seven days (Source: BrightLocal, 2026).
Fifteen reviews collected in one push in March are read as history by June, not as evidence. The asking has to be attached to the work itself. Anything that depends on remembering stops the first week the schedule gets heavy.
Owner responses count too (Source: BrightLocal, 2026), and we covered replying to the reviews you get elsewhere.
The five decisions: trigger, timing, channel, link, message
Everything else in this category is a preference. These five are the system, and none of them takes an afternoon to decide.
- The trigger. One moment every job already passes through: job marked complete, final invoice sent, deposit cleared. A trigger you have to remember is not a trigger, which is the same test that decides what's actually worth automating first anywhere else in the business.
- The timing. Same day, next morning at the latest. The memory is sharpest before the invoice starts to feel like a bill.
- The channel. The one they already used to reach you. Decide it once, next to your other post-job messages, rather than bolting on a second stream.
- The link. Your Google review short link, in the message, one tap. "Look us up on Google" is where these requests go to die.
- The message. The same words every time. Sending everyone the same thing is what makes it defensible: there is no guest list to explain.
Setting this up this week with what you already have
Being straight about where we stand: Beside answers your phone, books appointments, and qualifies callers. It does not send review requests today. Asking after every job is part of the system we are building, not part of what you would be using tomorrow.
So here is the version that costs nothing and works now
Pull your Google review short link out of your Business Profile: open it, then Read reviews, then Get more reviews, and copy the g.page link it hands you. Pick the trigger you already have. Save one message as a text template on your phone. If the software you run jobs through can fire that message on the trigger, let it; if it cannot, sending them yourself at the end of the day beats a system you never finish configuring. Then ask everybody who passes the trigger, with no exceptions list.
The message, if you want to steal it
Hi [name], thanks for having us out today. If the job went well, a quick Google review really helps a small business like ours: [your review link]
Either way, if anything wasn't right, just reply to this text and I'll sort it out. [your name], [business]
Two things make it defensible: everybody who passes the trigger gets it, and the second half is offered to everyone rather than used to decide who gets the first half.
That is genuinely most of it, and for a slow month it is enough. What it cannot do is survive the month you are running flat out, which is the month that produces the most jobs worth asking about.
What this looks like in your busiest month
Take the month that hurts, because a quiet one proves nothing.
Nearly half of consumers won't use a business with fewer than 20 reviews (Source: BrightLocal, 2026), so treat 20 as your floor.
Say you close 40 jobs in your heaviest month and ask all 40. If one in seven replies, that is six reviews, and six a month puts you past twenty inside four months. They also keep landing, which is what the recency window rewards. Now hand-pick instead: you decide 12 of the 40 are safe to ask, two come back, and the profile ages faster than it grows. Those are our numbers rather than a study, so run them against your own close rate. Then put your own average ticket against however many of those extra reviews you think win you one job, and you have the number that matters.
Will asking everyone pull your rating down?
It is a fair worry, and the bar moved. 31% of consumers now say they will only use a business rated 4.5 or higher, up from 17% a year earlier, and 68% want four or more (Source: BrightLocal, 2026).
The answer is volume and recency. A broad, continuous ask pulls your average toward what your work actually is and keeps the recent page of your profile full. A filtered ask produces a small, aging, suspiciously perfect set that also happens to be against the rules. And if asking every customer sinks your rating, the reviews are not your problem.
Sources: Google Maps User Generated Content Policy · 16 CFR Part 465 · FTC, 2024 · BrightLocal, Local Consumer Review Survey 2026. Beside product details verified on beside.com on August 5, 2026. Header photograph is a stock image, not a Beside customer.