A woman has her dishwasher fixed on a Tuesday morning. The technician is polite, the price is what the quote said, and she is genuinely pleased. By the time he has loaded the van, a text has arrived: “Thanks for choosing us! Tap here to leave a 5-star review.” She does not tap. Not because she is unhappy — she would have said something nice if anyone had asked in a way that felt like asking — but because that message is indistinguishable from the four others sitting above it from her dentist, her groomer, and a pizza place.
This is the state of review requests in 2026. The tooling got good, everybody adopted it at once, and the channel got saturated on roughly the same timeline. The businesses still accumulating reviews are not asking more often. They are asking in a place nobody else is.
Why most review requests fail
Three separate problems get blamed on one thing.
- ✓Channel saturation. The automated post-service text is now universal in local service. A customer receives several a month, has learned what they look like in the notification preview, and dismisses them without opening. The message is not being rejected; it is not being read.
- ✓The ask is about you. Most templates say some version of “reviews help our small business grow.” That is true and it is also a request for a favor. The requests that convert explain what the review does for the next person deciding — the one reading twelve results at 11pm trying to work out which plumber will not overcharge them.
- ✓It arrives with nothing behind it. A review request that is the first and only contact after a job is a transaction asking for a gift. The businesses with the best conversion rates earned something first, then asked separately.
Asking more people is not the fix when the problem is that nobody is reading the ask. Volume against a dead channel is still zero.
The rules, before anything else
Two things will get a business profile penalized, and both of them show up constantly in review-generation advice written by people who have not read Google's policy.
No incentives. None.
A discount, a free oil change, a gift card, an entry into a drawing — all prohibited, and the prohibition covers incentives offered in exchange for “honest” reviews too. Offering ten percent off for feedback “good or bad” is still an incentivized review under the policy. Enforcement ranges from review removal to profile suspension, and the FTC's rule on fake and incentivized reviews puts civil penalties on top of whatever Google does.
No gating.
Review gating means filtering who gets asked based on how you expect them to answer — routing unhappy customers to a private feedback form while sending happy ones to Google, or letting staff nominate which customers get the request. It violates the policy even when it is done informally, and “only ask the ones who seemed pleased” is the informal version almost every business does without realizing it is the thing being prohibited.
The practical consequence is that the ask has to go to everyone in a defined group — every completed job, every first visit, every closed matter — not to a hand-picked subset. Which makes the quality of the ask matter more, not less, because you no longer get to cherry-pick your respondents.
Where a handwritten card fits — and where it does not
The obvious move is to put a QR code on the thank-you card you were already sending. It is also the wrong one, for a reason worth being precise about.
The thank-you card carries no ask
A thank-you note with a request attached is not a thank-you note. It is a mailer, and customers classify it as one within about a second of opening it — the same trap covered at length in thank-you note examples for business. Worse, it retroactively reframes every card you send after it: once someone has opened one of your envelopes and found an ask inside, the next one gets opened with that expectation. You spend the credibility of the entire channel to buy one review.
The review request is its own piece of mail
Send it separately, a week or two after the thank-you, and let it be honest about what it is. A card that says plainly it is asking for something reads far better than one that pretends to be gratitude and turns out to have a QR code stapled to the bottom of it. The gratitude already happened. This is a different conversation.
Make the action take five seconds
The single largest drop-off in any review funnel is between intending to leave one and finding the page. A short QR code on the card, pointing at the direct review link rather than the profile, removes that entirely — the customer is holding the card and holding their phone, which is the one moment where paper and a phone screen are not competing. It also makes the card measurable: a scan is a real signal, and without one you are mailing on faith. Printed on a review card, a QR code is a doorway; printed on a thank-you, it is a toll booth. Same code, different letter.
What to actually write
Compare two. The first: “Thank you for your business! Reviews help small businesses like ours grow. Would you mind leaving us a 5-star review on Google? Scan below!” The second: “Sarah — the dishwasher job last month was a straightforward one, which is not always how it goes. Most people finding us are standing in a kitchen at 9pm trying to decide who to call, and what you'd write would be the thing they read. If you have two minutes, the code on the back goes straight to the page. — Dave”.
The second one costs the same to produce. Three things make the difference.
- ✓It names the specific job. A card that could have gone to any customer this month gets filed as marketing on sight, which is true of every note in this channel and especially true of one that wants something.
- ✓It explains who the review is for. Not the business — the next person standing in a kitchen at 9pm. That reframing moves response rates more than any other single change, and it is also the honest description of what a review does.
- ✓It never says five stars. Asking for a specific rating is the tell that you want a score rather than an opinion, and depending on how it is phrased it edges toward the incentive and gating lines. Ask for the review. Let them decide what it says.
One more thing to leave out: an apology for asking. Sorry to bother you, but planted at the front of a request does the doubting for the reader. The ask is reasonable. Make it like it is.
When to send it
Timing is the other half. The window where someone feels warmly about a business closes faster than most owners assume, but a card takes a few days in the mail — so the trigger has to fire earlier than it would for a text.
| Business | Trigger | Why this moment |
|---|---|---|
| Home services | Job marked complete | The repair is fresh and visibly working; a week later it is just a thing that happened |
| Auto repair | First repair order for a new customer | A stranger just trusted you with their car and has no way to judge whether they were treated fairly |
| Dental and medical | New patient first visit completed | The anxiety of a first appointment resolving well is the strongest feeling a practice ever produces |
| Restaurants | First visit, via POS or reservation | The meal is the product; the memory of it is what a review is actually made of |
| Salons and spas | First appointment with a new stylist | Forty-five minutes of conversation is more relationship than most businesses ever get |
| Professional services | Matter or engagement closed | Low frequency means one review has to carry for years, so it is worth the stamp |
Every row is a defined event rather than a judgment call. That is deliberate — it is what keeps the program on the right side of the gating rule, and it is also what makes it automatable.
Automating it from the system you already run
The reason this does not happen consistently is never that owners do not care. It is that writing a card competes with a full schedule and loses every time. The trigger is already in your software.
- ✓A job or work order marked complete in field-service software fires the thank-you, and a second trigger fires the review card ten to fourteen days later.
- ✓A first-time customer flag — first repair order, first appointment, first visit — narrows the program to the cohort where a review is most likely and the relationship is least established.
- ✓A QR scan writes back to the customer record, so the next person to open it can see the card landed and nobody sends a duplicate ask.
- ✓A suppression rule prevents a customer who has already been asked from being asked again within a year, which is the difference between a program and a nuisance.
Where your platform connects through Zapier the build is direct, and follows the same pattern as automating handwritten notes with Zapier. Where it does not, a weekly export of completed jobs works nearly as well — none of these moments need to land same-day, and a card that arrives on a Thursday with nothing else in the mailbox may do better than one that arrives in the Monday pile.
How to start
Pick one trigger and one cohort — new customers, completed jobs — and run it for ninety days. Keep the thank-you card and the review card separate from the start rather than merging them to save a stamp, because merging them is the failure mode this whole approach exists to avoid. Track scans per card mailed and reviews posted in the window; both numbers will be small and both will be larger than what the text message is currently producing.
Keep sending the automated text too. This is not a replacement for it — it is the second ask, in the channel nobody else is using, aimed at the customers who ignored the first one. What a card costs is a rounding error next to what one review is worth to a local business over the years it stays up.
Related reading by trade: handwritten notes for auto repair shops, home service businesses, salons and spas, dental and medical practices, and restaurants. If you would rather have the referral than the review, asking for referrals with handwritten notes covers that ask instead — it is a different conversation with a different set of mistakes.
Can I offer a discount or a gift card for a Google review?
No. Google prohibits incentives of any kind in exchange for a review, including incentives offered for honest or neutral feedback. Penalties range from removal of the affected reviews to suspension of the business profile, and the FTC's rule on fake and incentivized reviews adds civil penalties independent of anything Google does.
Is it against the rules to only ask customers who seemed happy?
Yes. That is review gating, and it violates Google's policy whether it is done with software that routes unhappy customers to a private form or informally by letting staff pick who gets asked. Ask everyone in a defined group — every completed job, every first visit — rather than a hand-picked subset.
Should the review request go on the thank-you card?
No. A thank-you with an ask attached reads as a mailer, and it changes how a customer opens every envelope you send afterward. Send the thank-you first with nothing in it, then send the review request a week or two later as its own piece of mail that is honest about being a request.
Does a QR code on a review request card actually help?
Substantially. The largest drop-off in any review funnel is between intending to leave one and finding the page, and a code pointing at the direct review link removes that step while the customer is already holding both the card and their phone. It also makes the card measurable, since a scan is a real response signal.
Why mail a review request when a text is free?
Because the text is no longer being read. Automated post-service review texts are now universal in local service, which means a customer receives several a month and dismisses them from the notification preview. A card is the second ask in a channel almost nobody is using, aimed at the customers who ignored the first one.
How long after the job should the review card arrive?
Aim for roughly ten to fourteen days after the work is finished, which usually means firing the trigger about a week out to allow for mail time. That puts it far enough past the thank-you card that the two are clearly separate, and close enough to the job that the customer still remembers the specifics.
Scribble fires a thank-you when a job closes and a separate review request a week later, drafts each one from what that specific job actually was, prints a QR code that links straight to your review page, and writes the whole thing in real ink. Book a demo and send one to your own address first.
Book a Demo- Marketing & RetentionAutomated Birthday Cards for Customers: When They Work, When They BackfireThe birthday is the one date every business already has and almost every business wastes. Why most birthday cards read as marketing, the four ways the program goes wrong, and how to run it off your CRM without it looking automated.
- Marketing & RetentionPhysical Mail Has a 98% Open Rate. Here's What That Means for Your Business.Email open rates hover around 20%. Physical mail gets opened 98% of the time. Here's the data on physical mail vs. digital outreach — and why handwritten notes outperform both.
- IntegrationsHow to Send Handwritten Notes Automatically from ActiveCampaignActiveCampaign already knows the moment a deal closes, a tag fires, or a lifecycle stage changes. A walkthrough for turning those automations into real handwritten cards — which triggers earn a stamp, how deal and contact fields become personal copy, and how to log every card back on the record.