Review policy and rating crises

The mechanics of asking, replying and resetting live on their own page. This page covers the two things that actually get shop owners into real trouble: what's banned outright, and what to do when your rating drops for reasons that have nothing to do with your shop at all.

What one centre bans outright

Its own opening statement is blunt: trying to trick its own review process, steal a shopper's data, copy another shop's work, or rig your rating or how you're found gets your shop removed entirely and you personally banned from operating there again.

More specifically: trying to manipulate comments, or inflate your own board position with paid, incentivised, filtered, or fake feedback, including working with an outside service to do it on your behalf, gets the centre actively stepping in, up to and including removing you permanently.

Rigging any part of the shopping experience (the busiest-shops board, the desk's own answers, the comment book, or referrals) is named directly as a violation of trust and is simply not permitted.

You must never require a shopper to rate or comment (or watch something, download something else, tap an advert, or agree to being tracked) in order to unlock something, or to receive money or any other reward, including gift cards or codes.

Read that carefully: the actual ban is on conditioning access or a reward on the rating itself. That's the real, bright line.

Also worth noting: the word "filtered" appears right alongside paid and fake feedback in the rigging clause. That has a direct, practical consequence. See the section on quietly screening people below.

What the other centre bans outright

Trying to manipulate where any shop sits at all is banned, including (but not limited to) inflating ratings, comments, or visit counts through illegitimate means.

The same rulebook also bans offering any reward in exchange for a positive rating, paying for or automating fake visits, and fake comments posing as real shoppers: including comments carrying links, discount codes, or promotion for some other shop entirely.

Where the real line actually sits

Allowed: asking every single shopper to rate you, at a moment you choose. Asking inside your own writing. Replying to comments. Fixing whatever people are complaining about, then asking again afterward.

Banned: paying for ratings. Offering anything of value in exchange for one. Conditioning a feature or a reward on rating you. Using an outside service that generates comments for you. Quietly screening who even gets asked, based on how you expect them to answer.

The screening question. This means quietly asking a customer privately how they feel first, then routing only the happy ones on to the actual public prompt, and sending unhappy ones to a private support form instead.

It's widely practised, and widely written about, including material that itself flags it as controversial.

Neither centre names it outright by that description. But one centre's own rigging clause lists filtered feedback right alongside paid and fake, and this practice is filtering, by definition. This guide does not recommend doing it. If your own organisation does it anyway, that should be a decision someone made deliberately, with that exact clause in front of them. Not a habit that simply arrived bundled inside some outside toolkit.

A genuinely defensible alternative that achieves most of the same goal: simply pick a better moment to ask everyone. The advice already covered elsewhere is to ask right when someone's likely to feel genuinely satisfied: right after finishing something, completing a task, or reaching a milestone. Asking after a genuine success is choosing a moment. Screening is filtering a person. The two look similar and are not the same thing at all.

When your rating collapses and your shop didn't actually change

Some drops have nothing to do with anything you shipped: an outage during your busiest possible moment, a controversial decision, a pricing change, some real-world event that's made people angry at you specifically, or coordinated, organised comment activity aimed at you.

A live sports shop is unusually exposed to this. A failure during a major event produces a concentrated burst of very low comments, all from people upset about the exact same short window of time.

What to actually do, roughly in order:

Work out whether it's genuine first. Look at the volume, the timing, and whether the comments actually describe a real problem. A burst of low comments describing a genuine outage is honest feedback, not an attack.

Fix the actual thing, if there is one. Everything that follows is worthless while the real cause is still live.

Reply to comments about the current version. The shopper gets told when you reply, and can then revise what they originally wrote. This is the only documented mechanism that actually turns an existing bad comment into a better one. Prioritise recent, low, current-version comments specifically.

Report anything that's genuinely abusive or fraudulent to the centre itself, rather than replying to it publicly yourself. Reporting is the sanctioned route; arguing back in public is not.

Don't reach for wiping your number clean first. It resets the visible number but not the written comments themselves, and the centre's own advice is to use it sparingly, because too few ratings can itself put people off. After a genuine relaunch that actually fixed the complaint, it's a reasonable move. As a way to simply escape one bad stretch, you lose the volume that made your number credible and keep every word explaining why it was low.

Remember ratings sit separately per country on the locked side. A crisis in one place doesn't follow you into another, and a country-specific problem shows up as a country-specific number, which is itself a useful diagnostic tool.

Write down what actually happened. The date, the cause, how deep it went, and how long recovery took. The next time this happens, the single most useful thing you can have is knowing exactly how long the first one took to recover from.

What tends to go wrong

Paying for ratings. Being removed permanently is the stated consequence here, not merely a warning.

Offering any reward for a rating. Explicitly banned on both sides.

Screening customers just because a tool made it easy. "Filtered" feedback is named directly in one centre's own rulebook. Decide it deliberately, or don't do it at all.

Replying angrily to a coordinated pile-on. Public, permanent, and covered by the same code of conduct that asks you to treat every single person with respect in a reply.

Wiping your number clean just to escape one bad stretch. Keeps every comment, loses all the volume.

Not replying to the comments that can still actually change. Recent, low, current-version comments are the only ones a reply can genuinely move.