Review Request

Also known as review invitation, review request email

Almost nobody reviews unprompted. The reviews a store has are, with rare exceptions, the reviews it asked for, which makes the request the entire collection funnel rather than a piece of housekeeping at the end of it. Operators tend to spend their attention on the copy, because copy is the part that feels like a decision. The copy is the smallest of the three levers. Timing, channel, and sequence are the ones that move the number.

Timing is the largest and the most commonly wrong. The rule is to ask after use, not after purchase. That means anchoring the request to delivery rather than to the order, and then adding however long the product takes to be judged. A phone case can be assessed the day it arrives. A mattress needs a fortnight of sleeping on it. A serum needs a month before the customer has anything to say that is not a guess. Sending at order confirmation, which is where a default flow puts it, reliably collects reviews of the checkout and the shipping speed, because that is genuinely all the customer knows yet.

Channel is a question about consent as much as about open rates. Email is the default because it is cheap, deliverable, and easy to ignore politely. SMS is opened far more often and tolerated far less, and the consent needed to send it is narrower than most stores realise. A card or a QR code in the package reaches people who never open marketing email at all, and costs nothing per order beyond the print. An on-site prompt after login catches repeat buyers. The channels are not interchangeable and stacking all of them on the same customer is how a review programme turns into a complaint.

Sequence is where volume is quietly lost. One ask under-collects, because a message that arrives on a bad day is simply gone. Five is harassment. Two, or three at most, spaced out by a week or so, is the usual honest ceiling, with hard stops: stop when they review, stop when they unsubscribe, stop when they reply about anything at all. Every step of the request should be one tap from the message into a form that already knows the order, so the customer never logs in, never searches for the product, and never has to remember which variant they bought.

Consider a Shopify store selling a vitamin C serum. Around 600 orders a month, and the review request went out at fulfilment. The reviews that came back praised the packaging and the delivery speed, which is what a customer can honestly assess on day one, and left the product page with almost nothing about skin. The store moved the ask to thirty days after delivery, with a single follow-up seven days later to anyone who had not responded, and changed the form's first field from a rating to one open question: what did you notice, and after how long. The count roughly doubled, but the useful change was in the text. Reviews started arriving with timeframes and specifics in them, which is the material a hesitant shopper is actually looking for.

Three constraints sit on the request and none of them are negotiable. The ask cannot be filtered by sentiment: everyone who bought gets the same link to the same form, and a flow that predicts an unhappy customer and routes them somewhere private is review gating, which the FTC has stated is deceptive and which most platforms prohibit outright. Any incentive is paid for the act of reviewing rather than for a rating, and is disclosed. And a review request is marketing rather than a transactional notice about the order, so an email one carries a working unsubscribe and your postal address, and the mobile number a customer gave you for delivery updates is not consent to text them marketing. Rules differ by market and change: in the United States, marketing texts need explicit prior consent and must honour an opt-out keyword; in the EU and the UK, the request rides on the same lawful basis as the rest of your post-purchase marketing. Confirm which basis you actually hold rather than assuming the order created one.

The part that gets least thought is that the request also decides what the reviews say. A form whose first field is a star rating collects adjectives, because a rating primes a summary judgement. A form that opens with one specific question collects sentences with nouns in them, which is what both a wavering shopper and a retrieval system have any use for. Asking a furniture buyer which room it went in, or a running shoe buyer what distance they cover, produces a corpus that answers real questions instead of a wall of the word great. Request design is corpus design, and it is the cheapest lever in the whole review stack because it costs one edit to a form.