Quickly
Within a few days: unanswered criticism looks ignored.
Online reviews
A public criticism, a useful reply.
To respond to a negative review: reply quickly and in public, acknowledge the problem without arguing, explain what you will do and move to private only what requires personal data. Find out how to write the reply, handle fake reviews and use criticism to improve.
In brief
How to respond to a negative review.
A public reply speaks to two people: the one who wrote and the one who will read later. For both, speed, respect, facts and a concrete commitment matter, not defending yourself at all costs.
With facts
What happened, what you are doing, by when.
Without personal data
Order numbers and addresses are asked for in private.
One reply, four parts
The anatomy of a good reply.
An effective reply has a recognisable structure: it thanks and acknowledges, gives a fact, states an action and offers a channel. A few lines are enough.
A fictional brand’s example reply to a negative review about the battery, with the four parts of a good reply highlighted and explained.
Question“Two stars: the battery lasts half as long as promised.”
Owner’s reply · example, fictional brand
Thank you for writing this, and we are sorry for the trouble. The latest update caused unusual battery drain on this model. The fix is available in the app from today. If the problem persists, write to us at support@northstar.example: we will follow it up.
Statement
Acknowledges the problem without doubting the customer.
Source
Gives a verifiable fact instead of a generic apology.
Recommendation
States a concrete action the customer can take right away.
Mention
Offers a private channel for personal data, without asking for it in public.
What not to write
Replies that make criticism worse.
The tone of the reply stays visible longer than the review. Some phrasings make a brand look careless or defensive even when it is right.
Four common ways of replying badly to a negative review, each set against a wording that respects the customer and informs readers.
| What to avoid | Better | |
|---|---|---|
| Copy and paste | The same apology under every review. | A reply that mentions the customer’s specific problem. |
| Defensiveness | Explaining why the customer used the product wrongly. | Acknowledge the trouble, then give the useful information. |
| Vague promises | “We will do our best” without saying what. | A precise action and when it will happen. |
| Data in public | Asking for the order number and phone in the reply. | Point to a private channel to continue. |
Fake or abusive reviews
When a review should be reported.
Unwelcome criticism is not a fake review. You report to the platform what breaks its rules or the law, with evidence; the rest is handled by replying.
Under Directive (EU) 2019/2161, anyone who publishes consumer reviews must state whether and how they check that the reviews come from people who actually used the product. Google’s content policies for business profiles also forbid asking only satisfied customers for reviews or discouraging negative ones.
Four common situations with online reviews, the right route for each and the evidence to gather before asking for action.
| Situation | What to do | Evidence to gather |
|---|---|---|
| No customer by that name | Report it to the platform. | No matching orders or visits. |
| Insults or personal data | Report it for breaking the rules. | The text and the rule broken. |
| Suspicious wave within a few days | Document and report together. | Dates, profiles and similar texts. |
| Justified but harsh criticism | Reply and fix, don’t report. | The solution offered, to mention in the reply. |
From criticism to product
Reviews as a list of priorities.
Read one by one, reviews look like noise. Grouped by topic, product and period, they become a list of problems to solve, in order of frequency.
The steps by which a period’s reviews become an ordered list of problems, assigned to an owner and then checked in the following period.
Gather the period
All reviews, positive and negative, with date and product.
Group by topic
Delivery, battery, support: one topic per recurring problem.
Sort by frequency
The most cited problem comes before the loudest.
Assign and check
One owner per topic, then a comparison with the next period.
Frequently asked questions
Frequently asked questions about online reviews.
Short answers to the most frequent questions about online reviews: how quickly to reply, how to handle fake reviews, how to ask customers for them and how much they weigh in AI assistants’ answers.
How soon should you reply to a negative review?
As soon as possible, ideally within a few working days. A quick, precise reply counts for more than a long one.
Can a negative review be deleted?
Only if it breaks the platform’s rules or the law, for example if it is fake, abusive or contains personal data. Justified criticism stays.
Can I offer a discount in exchange for a review?
An incentive tied to a positive rating is unfair. Ask all customers for their opinion, in the same way, with no conditions on the rating.
Do reviews count in AI answers?
They can: assistants sometimes summarise customer experiences. That is why a solved problem should also be explained in public replies.
Related paths
After “Online reviews”, the next step.
The guides explain the concepts with examples on fictional brands. From here you continue with the other guides, the methodology and the features that apply them to your brand.
Guides to reputation, trust and AI
Guides on GEO, AEO, AI visibility, online reputation, reviews and monitoring: what to check, a worked example and the steps to act.
ReadFrom question to action
How TrustInsider works in six steps: brand, questions, quote, evidence, page-by-page intervention and a check on the same questions.
ReadA transparent method, not a black box
How every TrustInsider measure is built: sampling of AI answers, sources and rights, index calculation and declared limits.
ReadFind the right answer
Help with access, credits, collections, integrations and data: guides, FAQs, the site assistant and support requests from your workspace.
Read