How to Get Customers to Leave Reviews: The Compliant Playbook
How to get customers to leave reviews: ask timing by business type, copy-paste scripts, the 2024 FTC rules on incentives, and ethical automation.
Getting customers to leave reviews comes down to three moves: ask every customer, time the ask to the moment your product proves its value, and make the review one click away. Most customers are open to leaving one. This playbook covers exact timing by business type, copy-paste scripts, the FTC rules on incentives, and how to automate the ask.
Reviews are an asking problem, not a satisfaction problem
Most thin review pages do not signal unhappy customers. They signal customers nobody asked. In BrightLocal's 2025 Local Consumer Review Survey, 96% of consumers said they are open to writing a business a review, yet only 29% wrote one in the past year. The gap between willing and written is the ask.
Reviews compound in three places at once. They lift conversion on your product pages, they feed the star ratings Google shows next to your listings and Shopping results, and on marketplaces they directly affect rank. If you sell physical products, reviews sit at the center of the product-page work covered in how to sell products online.
That makes review collection a system, not a favor you remember to request twice a year. The rest of this guide builds the system: when to ask, what to say in each channel, where the legal lines sit, and which parts to automate.
Pick the surface before you ask
Every ask should point at one destination, chosen by where your buyers actually decide. A local service business needs Google reviews, because the map pack is where comparison happens. A store selling from its own site needs on-site product reviews, which can also feed Google Shopping star ratings once synced through a review app and Merchant Center.
Marketplace sellers have the least choice: Amazon and Etsy reviews live on the listing and follow platform rules. If you sell across several surfaces, rotate asks by order source instead of splitting one customer's attention across two requests. One customer, one destination, one click.
For Google specifically, generate the direct review link from your Business Profile dashboard and keep it wherever you write customer messages. It turns a three-step request into one tap, and the same link works in email, SMS, and QR codes alike.
When to ask: time the request to delivered value
The most common mistake is asking at the transaction instead of at the value moment. A checkout-page prompt asks people to rate a product they have not even received, which produces silence at best and a vague, useless review at worst.
Ask once the product has had time to prove itself, and that window shifts with what you sell. For physical products like apparel or home goods, 7-14 days after delivery works because the customer has used the item several times. Consumables like beauty and supplements need 14-21 days, long enough for results to show.
Service businesses should ask within 24 hours of finishing the job, while the result is fresh and the goodwill is real. Digital products should wait for the customer's first win, usually 3-7 days in, and never ask at first login.
Send exactly one reminder 7 days after the first ask, then stop. Two contacts is persistence, three is pestering, and pestered customers write the kind of review you do not want.
How to ask: scripts for email, SMS, and the package insert
Email is the workhorse channel because it is cheap, automatable, and expected. If you run a store, the review ask belongs inside the post-purchase flow from our email marketing playbook, not in a standalone blast.
A working ask email is three sentences. Subject: "How is the [product] working out?" Body: "Hi [name], you have had your [product] for two weeks now. Would you take 60 seconds to share your experience? It helps other shoppers decide." One button, linked straight to the review form.
SMS earns faster replies but demands restraint. Only message customers who opted in, keep it under 160 characters, and send it once: "Thanks again for your order, [name]. Would you mind leaving a quick review? [link]" is the entire play.
Service businesses have a fourth channel: the moment a customer compliments the work. Say thank you, mention that reviews decide who finds you next, and send the direct link within the hour while the goodwill is live. A spoken ask followed by a same-day link is why well-run local shops collect reviews at rates most stores envy.
For ecommerce, a package insert card with a QR code catches the customer at the unboxing high point. Keep the card to one job, something like "Enjoying it? Scan to leave a review." Do not print a discount offer next to the review ask; that pairing reads as paying for reviews, and the next section explains why that is a real problem.
Whatever the channel, link straight to the review form. Every extra step (searching for your store, finding the reviews tab, signing in) sheds willing reviewers who owed you nothing in the first place.
The FTC line: what is legal since October 2024
In August 2024 the FTC finalized its rule banning fake reviews and testimonials, effective October 21, 2024. The rule bans writing or buying fake reviews, AI-generated ones included, bans undisclosed insider reviews, and carries civil penalties of up to $51,744 per violation.
Incentives are where honest businesses trip. The rule prohibits offering anything of value conditioned on a review's sentiment, so "leave us a 5-star review for 10% off" is now a federal violation rather than a growth hack. An unconditional incentive avoids that specific ban, but it does not make you safe.
Platform policies go further than the FTC. Google prohibits offering incentives in exchange for posting any review, positive or not, and Amazon has prohibited incentivized reviews since 2016 outside its own Vine program. The practical rule: never offer anything for a public review, and save the discount for a private feedback survey you do not publish.
Review gating, inviting only the customers you already know are happy, sits in the same danger zone. The FTC's Q&A on the rule flags suppression practices, and Google's review policies forbid discouraging negative reviews or selectively soliciting positive ones.
The safe list is longer than the banned list. Asking every customer, automating the ask, responding to every review, using Amazon's Request a Review button, and inviting an updated review after you fix a problem are all fully compliant.
Velocity beats a perfect score
Shoppers do not buy from the highest rating; they buy from the most credible one. A 4.4 average across 300 reviews persuades more than a 5.0 across 6, because a tiny perfect score reads like friends and family. A few critical reviews raise believability instead of sinking it.
The math punishes late starters. A product with 100 reviews averaging 4.2 stars needs 60 consecutive 5-star reviews to reach 4.5, which is months of flawless asking just to repair a slow start. Steady collection from the first order is far cheaper than a rating rescue later.
Velocity also protects you from looking fake. Thirty reviews landing in one week after a year of silence resembles a purchased batch to both shoppers and platform filters, while 10-15 new reviews per month builds the same total without the suspicion.
Handle bad reviews without making them worse
Respond to every negative review publicly within 48 hours. Acknowledge the specific problem, state what you are doing about it, and move the details to email. You are not really writing for the angry customer; you are writing for the hundreds of shoppers who will read the exchange later.
Once the problem is genuinely fixed, ask whether the customer would consider updating the review. Updates are allowed on every platform; what crosses the line is paying for removal or threatening legal action, which the FTC rule treats as review suppression.
Do not quietly delete critical reviews from your own storefront either. The FTC rule also covers displaying only positive reviews while hiding the rest, because a curated review page misrepresents what customers actually said. Let the 3-star review stand; it does more for credibility than one more perfect score.
Recurring one-star themes are operations data, not reputation noise. If late shipping or a clunky refund process keeps appearing in reviews, fix the back office before spending another dollar asking people to rate a broken experience.
Automate the ask without crossing the line
For product reviews, dedicated apps handle timing, sending, and photo collection automatically. Judge.me collects unlimited reviews on its free plan, with a $15 per month Awesome plan that removes branding and adds Google Shopping sync, while Loox starts at $14 per month with a photo-first widget. Either one turns the timing rules above into settings you configure once.
Amazon sellers get one compliant automation built in: the Request a Review button, which sends Amazon's own templated ask once per order. It cannot be customized, and that is exactly why it is safe.
The ask email itself is ordinary marketing work: write it once, trigger it on a timer, revisit it quarterly. It is the kind of job the current crop of AI marketing tools drafts in minutes, and bundled AI operators, our own SoGood.ai among them, draft review-request emails alongside the rest of a store's marketing for you to approve.
Your first 30 days
Here is the rollout order that produces a working review engine in a month:
- Week 1: claim your review surfaces (Google Business Profile, marketplace listings), install a review app on your store, and generate direct review links.
- Week 2: write the ask email, set the timing window for what you sell, and wire in the single 7-day reminder.
- Week 3: add the package insert or an opted-in SMS ask, and start responding to every review within 48 hours.
- Week 4: check your numbers and set a velocity target of 10-15 new reviews per month.
None of this needs an incentive, a gimmick, or a gray-area tactic. Nearly every customer is open to leaving a review. The entire game is asking all of them, at the right moment, with one click in the way.