Copy your Google review link (Business Profile, then Read reviews, then Get more reviews), then set a trigger in whatever system knows when a job is finished, such as your booking, invoicing or shop software. It should send one short request by text or email within hours, plus one reminder a few days later. Ask every customer, not only the happy ones.
"Every customer" is a rule, not a courtesy. Google's policy prohibits selectively soliciting positive reviews and offering any incentive for a review, and in April 2026 it added two more: merchants must not tell staff to collect a set number of reviews, or to ask for reviews that mention specific content such as a staff member's name. The good news is that volume doesn't come from clever filtering anyway. It comes from asking consistently at the right moment, which is exactly what automation is good at.
The five parts every automated review request needs
Whatever tool you use, a working setup has the same five parts. If one is missing, the automation either sends at the wrong time, nags people, or gets ignored.
- The link. Your Business Profile review link, copied from Read reviews, then Get more reviews. It opens the review box directly, which removes the step where customers search for you and give up.
- The trigger. An event that reliably means "the customer has had the whole experience": invoice paid, car collected, order delivered, results released. Not "job created" or "order placed".
- The message. Short, from a named person, with the link near the top. One site per message.
- The reminder. One follow-up a few days later, only if the first wasn't clicked. Never more than one.
- The stop rules. Don't ask the same customer again for six to twelve months, and don't ask anyone who has opted out of messages or whose job isn't finished.
Picking the trigger that marks a finished job
Timing is the biggest single lever you control. Ask too early and the customer hasn't finished forming a view; ask weeks later and they've forgotten you. The right trigger depends on what "finished" means in your business.
| Business | Trigger | Send | Watch out for |
|---|---|---|---|
| Car repair garage | Invoice paid or car marked collected | 2-3 hours after collection | Status changes that fire while the car is still waiting for parts |
| Laboratory (consumer tests) | Results released to the customer | The next day | Never vary the request by what the result was |
| Packaging supplier | Courier marks the order delivered | 3 days later, so boxes are unpacked | Repeat buyers: ask once per six months, not every order |
| Home-care provider | Six weeks into a new care package, then yearly | To whoever arranged care, with their agreement | Not after individual visits; never to a client who can't consent |
| Funeral director | Usually not automated at all | A card with a QR code a few weeks later, if anything | Treat every family the same way: all get the card or none do |
The last two rows show where automation should bend to judgement. A home-care provider sending an automated text after each of 14 visits a week would be absurd; one request six weeks in is appropriate. A funeral director that asks at all should decide a single approach for every family. That is still fair under Google's rules, because the decision is about the category of customer, not about who is likely to be pleased.
A packaging supplier goes from occasional asks to a steady flow
An illustrative packaging supplier sells printed mailer boxes and shipping cartons to small brands, mostly through its own online shop. It ships about 250 orders a month, around 60% of them reorders from existing customers. After six years it has 38 Google reviews, most of them from customers the owner happened to remember to ask.
The owner's first plan was "email everyone after every order". A quick look at the numbers ruled that out: with most orders being reorders, a buyer ordering every month would get twelve review requests a year, and by the third they would mark the emails as spam. The plan that went live instead:
- Trigger: the courier's "delivered" status, pulled into the shop platform.
- Delay: three days, so the boxes have been opened and used.
- Stop rule: skip anyone asked in the last 180 days, anyone who has opted out, and any order with an open support ticket until it is closed (then send as normal).
- Message: an email from the customer's account manager, with the review link in the second line.
- Reminder: one, five days later, only if the first link wasn't clicked.
With the 180-day rule, requests drop from 250 a month to about 90: first-time buyers plus regulars who haven't been asked for six months. To plan, the owner assumed that somewhere between 5 and 10 of every 100 requests would become a review, which works out at 4 to 9 new reviews a month. That is an assumption to replace with real figures after the first month, not a benchmark. Even at the low end, the profile would pass 80 reviews within a year, against six for the previous year.
Notice what the open-ticket rule does and doesn't do. It delays the request for a customer whose problem is still being fixed, because asking mid-complaint is tone-deaf. It doesn't drop them. Once the ticket closes, they get the same request as everyone else, whatever the outcome. Dropping unhappy customers from the list would be exactly the selective solicitation Google's policy forbids.
Three ways to build it, from free to fully managed
Route 1: switch on the setting your software already has
Check this first, because it often takes ten minutes. Many garage management systems, booking tools, field-service apps and practice systems have a review request option buried in their settings: paste in your Google link, choose the delay, write the message. If yours does, use it; it already knows when the job finished, which is the hard part. Be careful with shop review apps such as Judge.me on Shopify. They are excellent for product reviews that appear on your own store, but those are not Google Business Profile reviews, and they need a separate request.
Route 2: build it in Zapier or Make
If your software has no setting, an automation joins the trigger to an email or text. A typical Zapier recipe for the packaging supplier:
- Trigger: order marked fulfilled or delivered in the shop platform.
- Filter: customer has an email address and hasn't opted out.
- Look up the customer in a Google Sheet of "last asked" dates.
- Filter: last asked more than 180 days ago, or never.
- Delay three days (check how your plan counts delay steps).
- Send the email from the account manager's mailbox.
- Update the sheet with today's date.
A quick sum on Zapier's task counting: triggers and filter steps cost nothing, and each successful action counts as one task. The lookup, the email and the sheet update make three tasks per request. At 90 requests a month that is 270 tasks, well inside the 750 on the Professional plan ($19.99 a month billed annually). Without the 180-day rule, 250 requests would need 750 tasks, the whole allowance, which is another reason the stop rule matters. The Free plan won't do: it allows only two-step Zaps and checks for new triggers every 15 minutes. Make can run the same flow on its free 1,000 credits a month for low volumes, and routers there use no credits.
Route 3: pay a review platform
Review platforms bundle the requests, text messaging, reminders, click tracking and a dashboard. NiceJob's Starter plan is $75 a month and automates review requests and follow-ups by text and email; GatherUp's small-business plan is $99 a month for one location, including up to 300 texts and 3,000 emails. They are worth it when text messages matter (garages, clinics, trades) or when you have several locations. For a fuller comparison, including tools with AI reply features, see AI review management tools for small businesses.
Writing a request people actually click
The message should take less than ten seconds to read. AI is useful here for producing variants to test, not for writing an essay. A prompt that works:
Write 3 versions of a review request for a [business type].
Channel: [text message, under 160 characters / email, under 80 words]
From: [first name, role]
Customer detail available: [first name, what they bought or had done]
Link: [REVIEW_LINK]
Rules: the link must appear in the first two lines. One request only,
for Google. No incentives, no "5 stars", no "if you were happy".
Plain, warm, specific to the job. No exclamation marks.An illustrative result for the car repair garage's text message, and what I'd change:
Illustrative AI output, version 2: "Hi [first name], thanks for trusting us with your car today. If you were happy with the service, a quick Google review would mean the world: [REVIEW_LINK] [your name], [Garage]"
"If you were happy with the service" crept in despite the rule, and it has to go. It invites only happy customers, which is the gating pattern. "Would mean the world" is gushing for a garage. The version that went live: "Hi [first name], thanks for bringing the car in today. A quick Google review helps other drivers find us: [REVIEW_LINK] [your name], [Garage]". That's 119 characters with a short link, so it fits in one text.
The packaging supplier's email, before and after an edit:
- Before: a 190-word email opening "We hope this email finds you well!", three paragraphs about the company's sustainability commitments, and the review link in the final line after a request to also follow them on social media.
- After: "Subject: How did the mailer boxes turn out? / Hi [first name], your 2,000 kraft mailers were delivered on Tuesday. If you have a minute, a Google review tells other small brands what we're like to work with: [REVIEW_LINK]. And if anything isn't right with the order, reply to this email and I'll sort it. [your name], account manager".
The "if anything isn't right, reply to me" line is not gating, because it doesn't replace the review request or send unhappy customers elsewhere; everyone gets the link first. It simply gives someone with a problem a faster route to a fix. Keep the order and the wording like this: the link first, the help offer second.
Merge fields such as first name and order summary come straight from your shop or booking data, and for most businesses that's all the personalisation needed. Only add an AI step to write each message individually if your job data is messy free text, and even then keep the link and the structure fixed so the AI can't drift into "leave us five stars".
One reminder, then stop
A single reminder typically catches people who meant to review and got distracted; a second or third mostly generates opt-outs. Send it three to five days after the first, only if the link wasn't clicked (platforms track this; in a DIY build, use your email tool's click tracking), and make it shorter than the first message:
Hi [first name], just a quick nudge in case the last message got buried.
If you have a minute: [REVIEW_LINK]. Thanks either way. [your name]Then stop, and record the date. Google won't tell you which customers left a review, because reviews show only a display name, so the "last asked" date is your only reliable guard against asking the same person too often. Rules on texting and emailing customers differ depending on where you operate, and whether a review request counts as marketing varies, so include a simple way to opt out and check with your adviser if you're unsure.
The rules that get a Business Profile restricted
Google can restrict a profile that breaks its fake-engagement rules. According to its help pages, that can mean new reviews blocked for a period, existing reviews unpublished, and a warning shown to customers that fake reviews were removed. These are the rules automated requests most often bump into:
- No incentives of any kind. No discount, prize draw, free wash or loyalty points for a review, even an honest one. That includes incentives to change or remove a negative review.
- No gating. Don't send a "how did we do?" survey first and pass only the happy scores to Google. The explainer on what review gating is and why it's penalised covers the variations tools sometimes sell as features.
- No pressure on the premises. Google's policy says merchants must not require or pressure people to review while they are there. A QR code on the counter is fine; standing over someone while they type is not.
- No staff quotas. Since April 2026 the policy explicitly bans asking staff to collect a certain number of reviews. A leaderboard of reviews per mechanic breaks this.
- No scripted content. Also new in April 2026: don't ask staff to request reviews that include specific content, including a staff member's name. "Please mention [staff name]!" is out.
Consumer-protection law in many places also covers fake and incentivised reviews, separately from Google's own rules. If you're considering anything clever, assume it's a problem and ask an adviser.
Mistakes that turn review requests into one-star reviews
These are the failures that show up in real setups, usually in the first fortnight.
- The trigger fires too early. An illustrative garage set its trigger to "job status changed", which also fired when a car was moved to "waiting for parts". Three customers got a review request while their cars were still in the workshop; one replied with a one-star review saying exactly that. The fix was to trigger only on "invoice paid".
- A blast to the whole customer list. Sending a request to 1,500 past customers on launch day reaches people who haven't visited in years and don't remember the job. Start from today's customers and let the flow build.
- Asking regulars every time. The packaging supplier's original "every order" plan. Repeat buyers are your most loyal customers; don't turn them into your most irritated.
- A link to the wrong profile. Businesses with an old duplicate listing sometimes copy the link from the wrong one. Click your own link on a phone before going live.
- Sending from a no-reply address. When a customer replies to the request with a complaint, nobody sees it. Send from a monitored mailbox or number.
- A generic sender name. "Customer Services" gets fewer clicks than "[your first name] at the garage". People review people.
Reading the numbers after 60 days
Keep one row per month in a sheet: requests sent, links clicked, new Google reviews, average star rating of the new reviews, and replies posted. An illustrative first two months for the packaging supplier:
| Month | Requests | Clicks | New reviews | Average stars | Replied |
|---|---|---|---|---|---|
| Month 1 | 96 | 21 | 7 | 4.7 | 7 of 7 |
| Month 2 | 88 | 17 | 5 | 4.4 | 5 of 5 |
Reading it: about 7 in every 100 requests became a review in month 1 and about 6 in month 2, inside the planning range. Clicks without reviews (14 and 12) suggest people open the review box and don't finish, which is normal; a shorter message won't fix that, but asking a few days after delivery rather than on the day might. The average star rating of new reviews dipped slightly, and that is what an honest, ask-everyone system looks like. A new profile rating of 4.4 to 4.7 from real customers is more persuasive to buyers than a suspiciously perfect 5.0.
Two checks are worth doing monthly. Read every new review, because the flow will surface problems you didn't know about. And reply to each one, which is where AI saves real time; the routine in using AI to respond to Google reviews takes about two minutes a review once your reply brief is written. If the numbers stall, change one thing at a time (the delay, the sender, the first line) and give each change a full month before judging it.
Review request questions that come up during setup
Can I put a QR code for Google reviews on my counter or invoices?
Yes. Business Profile can generate a QR code for your review link, although Google's help page says the QR code can currently only be created from a computer browser, not the mobile app. Printing it on invoices, receipts or a card is fine. What Google's policy rules out is requiring or pressuring customers to leave a review while they are on your premises.
Should I ask for reviews on other sites as well as Google?
Ask for one site per request, or the customer does none. Pick Google for most local and service businesses because that is where new customers look first. If your sector has a dominant review site, alternate: send some months' requests to that site. Asking for two reviews in one message splits the customer's attention and makes it more likely that neither gets written.
Is it worth paying a review platform instead of building this myself?
Pay if you want text-message requests, click tracking and reminders without maintaining an automation, or if you have several locations. Platforms such as NiceJob start at about $75 a month. Build it yourself if you already use Zapier or Make and send email requests from one system; the running cost can be under $20 a month.
Further reads
- How to Write Google Business Profile Posts With AI Every Week — Keep the profile active with weekly posts between reviews.
- Getting More Reviews After Every Job With AI Follow-Ups — The same idea applied to trades, where every job ends on site.
- How to Optimise Your Google Business Profile With AI — Make the profile people land on worth the click.
- How to Turn Customer Reviews Into Marketing Copy With AI — Use the growing pile of reviews in your marketing, with permission.
- How to Reply to Negative Reviews With AI: Examples and Pitfalls — Be ready for the critical reviews that asking everyone will bring.
- How to Measure Whether AI Is Improving Your Marketing Results — Fit review numbers into a wider view of what your marketing achieves.
- AI Quick Wins: 12 Things a Small Business Can Set Up This Week — Twelve small AI jobs sorted by setup time, each with a copyable prompt, a real business example and the catch to watch for.
- What Are Diners Really Saying? Using AI to Analyse Your Reviews — A method for turning months of restaurant reviews into a few counted, checked themes, with a codebook, a tagging prompt and a worked nine-month example.
- How to Automate Pre-Arrival and Post-Stay Emails With AI — A six-message guest email timeline, the templates to write with AI, the rules for who gets what, and how to let AI draft replies safely.
- How to Reply to Hotel Reviews With AI Without Sounding Canned — Why AI review replies read as templated, and the voice sheet, reply shapes and prompt that make each one sound written by the person who runs the place.
- How Photographers Can Handle Client Emails With AI — The email map of a photography booking, a voice card for the AI, and prompts for the awkward ones: RAW file requests, late galleries and unhappy clients.
- ChatGPT Prompts for Pet Sitters and Dog Walkers That Win Clients — Twelve ready-to-paste prompts for pet sitters and dog walkers, from profile bio to daily updates, each with what to feed in and what to check.
- How Car Valeting and Detailing Firms Use AI to Fill the Diary — An empty-slot audit, rebooking intervals, photo-quote prompts and standby-list messages that help a valeting or detailing business keep its diary full.
- How Florists Can Prepare for Valentine's and Mother's Day With AI — An eight-week countdown for florists: forecast, pre-order menu, stem maths, cut-off messages, card-message rules and delivery routes.
- How Florists Can Use AI for Orders, Reviews and Social Posts — Worked examples for a flower shop: a rushed phone order turned into a docket, a substitution message, a year of reviews analysed, and accurate captions.
- How Mobile Dog Groomers Use AI for Routes, Reminders and Rebooking — Zone days, route optimisation, reminders with realistic arrival windows and rebooking by coat interval, worked through for a one-van mobile groomer.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Google Business Profile Help (Create a link or QR code to request reviews; Tips to get more reviews); Google Maps user-contributed content policy (rating manipulation, including the staff-solicitation rules); Zapier pricing and task-counting documentation; NiceJob and GatherUp pricing pages.