Email and customers · Lesson 5.5
5 Automated Review Requests Mistakes and How to Fix Each
Automated review requests save time, but they also repeat a mistake faster. A bad message to one customer is a small problem. The same bad message sent to every customer is a bigger one, and a rule that breaks a platform policy breaks it every time it runs.
These are five mistakes that come up when people set up automated review requests, why each one costs you, and a concrete fix. They come from how the process works and from the official rules, not from a survey, so I will not put a percentage on any of them. Lesson 5.5 teaches the correct version of each step: Get five star reviews without chasing.
Most review request problems come from timing, wording, filtering, silence after the review, and never checking the automation.
Mistake 1: Asking before the customer has the result
What it looks like. The request goes out on the order date, or the moment payment clears. The customer has nothing to review yet, because the parcel has not arrived or the job is not done.
Why it costs you. Best case, the message is ignored and your one chance with that customer is spent. Worst case, you get a review about shipping delays or a late start, not about the product. People also feel pestered when asked to praise something they do not yet have.
The fix. Trigger on completion, not purchase. Pick an event that proves the result: delivery confirmed, job signed off, session finished. Lesson 5.5 uses three days after delivery as its example for products, so the customer has had time to use what they bought. Your own best moment may differ, so start there and adjust.
If your system cannot detect delivery, ask at a fixed number of days after dispatch that covers most deliveries, and accept that some will arrive late. Better yet, send a delayed trigger and exclude refunded or cancelled orders.
Mistake 2: Building a filter that only asks happy customers
What it looks like. A short survey comes first: "How was your experience?" Happy customers get the Google link; unhappy ones get a private feedback form. It feels tidy and protects your rating.
Why it costs you. Google's Maps policy lists, among things merchants may not do, discouraging or prohibiting negative reviews or selectively soliciting positive reviews. A filter built this way does exactly that. If found, it risks your reviews being removed or your profile being restricted. Google also has a page on Business Profile restrictions for policy violations. Beyond the rules, a rating built only from asked, happy customers is a thinner signal for you too, because you lose the honest feedback.
The FTC's rule on consumer reviews separately covers review suppression, such as using groundless legal threats to remove negative reviews, and incentives conditioned on sentiment. A filter is not necessarily the same thing, but it is in the same area, so be careful.
The fix. Send the same ask to every customer who completed a purchase. Gather feedback privately in addition if you like, for example by inviting replies to the email, but do not use it to decide who gets the public link. Keep your attention on dealing with problems directly, personally, and promptly.
Mistake 3: Offering something in return
What it looks like. "Leave a review and get 10 percent off your next order." It seems fair and it is a common idea.
Why it costs you. Google's help pages say offering incentives such as free or discounted goods or services in exchange for posting, changing or removing a review is considered fake and misleading content and is strictly prohibited. Reviews posted due to an incentive can be removed.
The FTC's Q&A on its rule adds a nuance: the rule does not itself prohibit giving incentives for reviews, as long as there is no express or implied requirement that the review express a particular sentiment, and failing to disclose incentives could still be a problem under other FTC guidance. But that does not help you on Google, whose own pages prohibit them.
The fix. Do not offer anything for a review. If you want to reward customers, do it for other reasons such as loyalty, and keep it unconnected to reviews. Thank people genuinely in your reply instead.
Mistake 4: Never replying, or replying with a template
What it looks like. Reviews arrive and sit unanswered for weeks. Or every reply is the same: "Thank you for your feedback! We appreciate your business."
Why it costs you. Replies are public. Google's help says the reviewer is notified when you reply and that customers can still change their review after reading your reply. Its tips say that instead of sending the same thank you to everyone, you should focus on reviews where you can share a useful update or answer a question. Silence on a negative review reads as indifference to everyone who reads it later.
The fix. Reply to every review within a day or two when you can, and keep it short. Use the reviewer's name, mention something they said, and avoid pushing offers in the reply, as Google's tips say to be conversational, not promotional. For negative reviews, apologise where fair, explain what you can and cannot do, and take complicated issues offline. Never share a reviewer's private details.
AI can draft replies from the review text. Read every one before it goes out. A draft that invents a detail ("we have refunded your order") when you have not is a bigger problem than no reply.
Mistake 5: Setting it up and never checking it
What it looks like. The automation is running, so you assume it works. Months later you discover the link was broken, emails go to spam, the same customer got three messages, or the trigger stopped after a software change.
Why it costs you. A broken link means every request in that period was wasted. Duplicate messages annoy customers. You lose the ability to spot a problem early, and each week lost is a week of customers you cannot ask again at the right moment.
The fix. Add a short weekly check, ten minutes at most:
- Open the last few sent messages and click the link on your phone.
- Check that no one was asked twice, using an "asked on" column.
- Confirm new reviews have a reply.
- Write down requests sent and reviews received.
Also do a full test after any change to your email tool, your list, or your website. The post on the metrics that matter explains what to record.
A quick self audit of your automated review requests
| Question | Healthy answer |
|---|---|
| When does the ask go out? | After the result, on a fixed rule |
| Who gets it? | Every completed customer |
| Is anything offered? | No |
| Who replies and when? | You, within a day or two |
| When did you last test the link? | This week |
If you answer differently on any row, that is your first fix. Do one at a time and re check after a few weeks.
Smaller slips worth a mention
- Scripting the review. Asking for "five stars" or particular words. Google's policy says merchants should not request that specific content be included or try to influence the rating.
- Pressure in person. Google's policy says not to require or pressure users to leave ratings while on the premises.
- Pointing Yelp traffic at a review. Yelp's business help says you shouldn't ask customers to post reviews on Yelp.
- A long message. Cut it to three lines.
- Messy data. A wrong name or product in the message looks careless. Lesson 5.1 shows how to keep one clean customer list.
Fixing things you already got wrong
If you recognise a mistake from above, you do not need to panic or start over:
- Pause the automation.
- Fix the single problem.
- Test with your own details.
- Switch it back on.
If you ran a filter or offered incentives, stop both now. For anything involving legal exposure, check the current rules for your country or ask a professional, since this post summarises official pages and is not legal advice.
Questions people ask
What is the biggest mistake?
Filtering out unhappy customers, because it breaks Google's policy as well as hiding useful feedback. The others are fixable with a bit of care.
Can I reply to reviews with AI?
Use it to draft, then edit and read every reply before it is posted.
How many reminders can I send?
There is no official figure. One reminder is a sensible limit for most businesses.
How do I stop asking the same customer twice?
Keep an "asked on" column and exclude any customer with a date in it.
Next step
Choose the mistake that applies to you most, fix it this week, and leave the others for next week. Lesson 5.5 shows the right setup step by step, including timing, the ask, the link and the reply: Get five star reviews without chasing. For the whole process in one place, read the complete setup guide.