Ecommerce reputation management: what to fix first
Ecommerce reputation management for online stores: who reads your ratings, disputes and reviews, what to fix first, and the US and UK rules on reviews.
Ecommerce reputation management is the work of controlling what shoppers, marketplaces, payment providers and search engines can read about an online store, and fixing the signals that cost sales or account access first. For most stores the order is payment disputes and refund handling, then marketplace and Google ratings, then review platforms, then scam-alert listings and public complaints.
The order matters because the readers are not equal. A shopper skims stars. A marketplace reads a score computed from your order history. A payment provider reads your dispute rate, and a card network can act on it long before a shopper ever sees a bad review. A store that spends a quarter chasing five-star reviews while its dispute rate climbs is fixing the wrong layer.
This guide maps each reader, shows what each one actually looks at, and gives a decision table and a 90 day sequence. It also covers the limits: the platform rules and consumer-protection laws that decide which review tactics are allowed. Where a rule belongs to one country, the text says so. This is information about published rules, not legal advice.
- Fix in this order: dispute rate and refund handling, marketplace seller scores, Google store ratings, review platforms, then scam-alert and complaint listings.
- Payment providers and marketplaces act on numbers you can measure. Shoppers and search engines act on text and stars you can only influence.
- Asking every customer for a review is allowed. Fake reviews, paid-for sentiment and hidden bad reviews fall foul of the US and UK rules read for this guide, and selective invitations fall foul of Trustpilot's.
- Rules differ by platform and by country. Read the current policy page before acting on any figure in a guide, including this one.
Three figures from the platforms' own documentation set the scale of the work.
Stripe's documentation on how disputes work, and Google Merchant Center's store ratings overview, as read for this guide.
Who reads an online store's reputation, and what each one sees
An online store is read by five audiences at once, and each reads a different signal. Treating them as one "reputation" is the first mistake, because the signals sit in different places and are fixed by different people.
Shoppers read star ratings, the review count, the most recent reviews, the answer a store gave to a complaint, and the first page of results for the store name. They do not read policies. They read the one-star review that says the parcel never arrived.
Related readBrand Ambassador Scandal: The First 24 HoursMarketplaces read computed scores. On eBay, for example, the seller standards page describes metrics such as transaction defects and cases closed without seller resolution, evaluated on a fixed monthly schedule. The score decides placement, fees and selling limits, and no review text is needed.
Payment providers and card networks read disputes, refunds and fraud reports. They never look at your star rating. They look at the ratio of disputed payments to captured payments.
Search engines and their shopping programs read ratings from several sources and combine them. Google says its store rating draws on its own Customer Reviews program, on shopping reviews from supported third-party partners and users of Google Search, and on aggregated performance metrics from Google-led shopping research.
Scam-alert and trust-score sites read technical setup, reviews from across the internet, location and contact details. A shopper who is unsure about an unfamiliar store often checks one before paying.
The order to fix things: a decision table
Fix the signal that can stop you trading before the signal that only slows conversion. The table ranks the signals by consequence and by how directly you control them.
| Signal | Who acts on it | What you control | Priority |
|---|---|---|---|
| Dispute and fraud rate | Payment provider, card network | Shipping clarity, descriptors, refunds, fraud screening | First |
| Marketplace seller score | The marketplace, automatically | Dispatch, tracking, returns, case handling | First, if you sell there |
| Google store rating | Shoppers and ad systems | Review volume and spread, accurate site claims | Second |
| Review platform profile | Shoppers, search engines | Invitations, replies, reports of rule breaches | Second |
| Scam-alert and trust scores | Cautious first-time buyers | Contact details, company information, owner feedback | Third |
| Public complaints and social posts | Shoppers, journalists | Reply speed, resolution, platform reports | Third, faster if viral |
One exception reverses the order. If a false or abusive post is spreading and shoppers are quoting it, handle that first for a week while the upstream fixes run in parallel. The other signals are slower to move than a public reply is.
Payment providers read your dispute rate before they read your reviews
Card networks track disputes and fraud reports per merchant, and a store that crosses their thresholds can face monthly fines and, in the end, loss of card acceptance. This is the highest-consequence signal and the one most stores never look at until a notice arrives.
Related readFounder reputation affects company: separate or alignStripe's documentation explains the mechanics. A dispute usually lets the cardholder's bank pull the payment back, the disputed amount plus a fee is debited from your balance, and the dispute counts toward your dispute rate. Stripe says card networks typically allow cardholders to open a dispute within 120 days of the payment, you usually have 7 to 21 days to respond depending on the network, and the whole cycle can take two to three months. The outcome is decided by the cardholder's bank, not by the payment provider.
Stripe's monitoring-program page, which it describes as a general guide and not a full reference, adds two points a store owner should know.
- The page says monitoring programs do not consider refunds when identifying disputes, so a refund after the dispute is filed does not remove it from the count.
- Early fraud warnings are not disputes, but Visa's VAMP program includes them in its calculation.
Stripe also lists Visa and Mastercard thresholds. As of the page read, the Visa VAMP table lists a non-compliant ratio of 0.5 percent with a count of 5, and an excessive ratio of 1.5 percent with a count of 1,500, for regions other than CEMEA (Central and Eastern Europe, Middle East and Africa), where the excessive column lists 2.2 percent and a count of 150. The Mastercard Excessive Chargeback Merchant level needs both a chargeback count of 100 to 299 and a rate of 1.5 to 2.99 percent. Thresholds are set by the networks and change, so the networks' own current documents govern.
A worked example with stated assumptions
Assume a store captures 4,000 card payments in a month. It receives 12 disputes and 8 early fraud warnings that Visa counts, and no payment appears in both groups. That is 20 events, and 20 divided by 4,000 is 0.5 percent. On the Visa figures above, that matches the listed non-compliant ratio, and 20 is above the listed count of 5, before any review has been written. At 60 events the ratio is 1.5 percent, which matches the ratio in the excessive column, but that column also lists a count of 1,500 outside CEMEA, which a store this size would not reach. Stripe's page does not spell out how count and ratio combine, so the store should read Visa's own documents for its status.
Related readLayoffs and employer brand reputation: what stays publicSource: Stripe's card network monitoring programs page, as read for this guide. Applies only to merchants at the Excessive Chargeback Merchant level; the High level carries higher fines.
What lowers the rate before it is counted
Stripe's prevention list is plain operations, not reputation work, and it is where the effect is largest. It also notes that unrecognized statement descriptors and confusing receipts can show up as disputes labeled general or duplicate.
- State shipping times before checkout, send tracking, and announce delays before the customer has to ask.
- Keep a refund or return policy flexible enough that a customer takes it instead of calling the bank.
- Answer any inquiry from a card issuer at once. Stripe's how disputes work page says an unanswered inquiry can escalate to a likely unwinnable chargeback, and that American Express and Discover use inquiries most often, while Visa and Mastercard no longer do.
Marketplace seller scores are computed, so fix the operation
A marketplace score is arithmetic on your orders, and a reply or a review request does not move it. If you sell on a marketplace, fix dispatch speed, tracking, returns and case handling, because that is what the score is built from.
eBay's US seller standards page is a clear example, read for this guide. It describes three levels: Top Rated, Above Standard and Below Standard. For Below Standard it lists consequences that include lower placement in search, possible lower selling limits, no Promoted Listings, and possible payment holds until tracking shows shipment, with possibly higher fees from the following month and, after an extended period, selling restrictions. It states that performance is evaluated on the 20th of each month, and that the window depends on volume: sellers with more than 400 transactions in the previous three months are assessed on those three months, and sellers with fewer than 400 on the previous twelve.
Google store ratings: how the rating forms and what blocks it
Google's store rating combines several sources, filters reviews it considers untrustworthy, and has volume and quality thresholds before it can appear. A store with too few reviews may show no rating, and a store under the ad threshold may not show it in ads.
From Google Merchant Center's store ratings overview, as read for this guide:
- The rating shows stars out of five, the number of reviews, sometimes a qualifier such as average delivery time, and a link to recent reviews.
- Sources include Google Customer Reviews, shopping reviews for the store's domain from supported third-party partners and users of Google Search, and aggregated performance metrics. Google says it assigns different weights to different sources and does not state the weights.
- A merchant generally needs at least 100 eligible, unique reviews, though the number can vary.
- Ads require an average rating of at least 3.5 out of 5.
- Ratings are calculated per country. Reviews from one country count toward that country's rating.
- Google says it does not modify existing ratings. It calibrates them and filters reviews it considers untrustworthy, such as spam and abuse, or questionable, such as reviews that do not indicate a transaction took place.
The same page gives two time windows for the reviews it counts, a rolling 24 months in one section and five years in the troubleshooting section, without reconciling them. Check the current page before planning around either figure.
Related readLegal Dispute and Brand Reputation: A Company PlaybookGoogle's separate Product Ratings policies for review feeds are stricter on method. They allow incentives such as gift cards or discounts if the incentive does not depend on the sentiment of the review and the review is disclosed as incentivized. They require retailers to share all reviews, including low-star ones, and not to delete old reviews from the feed. They remove reviews with conflicts of interest, such as those from employees, and reviews that are primarily automated or AI-generated. Google may remove reviews, block feeds or suspend accounts for repeated or egregious violations. That is a feed program, separate from the store rating, but it shows how Google reads the same behavior.
Review platforms reward consistency, not volume tricks
A review platform profile is read by shoppers and ranks in search for your store name, so it is worth running well. The rules on how you ask are stricter than most store owners assume.
Trustpilot's guidelines for businesses, as read for this guide, say not to be selective with invitations by picking and choosing which customers you invite, and to invite consistently and fairly. They say not to pre-screen by sending happy customers to the platform and unhappy ones to a private channel. They state that incentivized reviews are not allowed on Trustpilot, which includes discounts, promo codes and prize draws tied to writing or editing a review. They bar reviewing your own business, and bar inviting family, employees or shareholders. For misuse they list actions such as suspending features, placing a consumer warning on a profile, hiding the score and legal action.
Related readNew company online reputation with no reviewsReplies are the one fully legitimate lever. Trustpilot's guidelines ask for polite, professional replies without personal information, threats or aggressive language, and say that a business that disagrees with a review that does not breach the guidelines can reply and tell its side. A good reply in a store context does three things: it states what happened in two lines, says what was done or offered, and moves the case to a named channel. It is written for the next shopper, not the reviewer.
Scam-alert sites and trust scores: read what they say and fix what you control
Scam-alert and trust-score sites compute a score from public signals, and a legitimate store with thin contact details can score badly. ScamAdviser's FAQ, as read for this guide, says an automated algorithm weighs technical setup, customer feedback, business location and contact details, plus other factors, and that its rating is a guide. It also says site owners who disagree can contact it.
You cannot buy a better score, and you should not try to write reviews to move one. What you control is the input. Put a full company name, registered address, working contact channels and clear policies on the site. Keep the domain, hosting and certificates consistent. Respond to genuine feedback on the platforms a trust-score site draws from. Then request a re-check if the score reflects stale or wrong data, and accept that it is a guide.
The rules that limit review tactics: US and UK
Both the US and the UK treat fake reviews and several related tactics as unlawful for businesses, and both regulators have published rules and guidance. These differ from platform rules: platforms can suspend you for a breach of terms, and regulators can pursue a breach of law.
Related readPartner and supplier due diligence on your companyUnited States. The FTC's Rule on the Use of Consumer Reviews and Testimonials (16 CFR Part 465) took effect on October 21, 2024, according to the FTC's questions and answers page. As read, it covers:
- Fake or false reviews and testimonials, including buying them or spreading ones you knew or should have known were fake (section 465.2).
- Paying or rewarding someone for a review with a particular sentiment, positive or negative (section 465.4).
- Insider reviews without a clear disclosure of the material connection (section 465.5).
- A site or entity you control presented as an independent source of reviews (section 465.6).
- Review suppression through groundless legal threats, intimidation or certain false public accusations, and misrepresenting a filtered set of reviews as representative of all of them (section 465.7). Withholding reviews is not suppression when the criteria apply equally to all reviews regardless of sentiment, such as personal information, abusive content, a reasonable belief that a review is fake, or a review wholly unrelated to the products sold.
- Buying or selling fake social media indicators, such as bot-generated followers or views, when you knew or should have known they were fake (section 465.8).
The FTC also says that organizing reviews is not suppressing them, and that showing five-star reviews first is not covered by the rule, but that arranging reviews so that consumers find negative ones hard to see could violate the FTC Act's ban on unfair or deceptive practices. The FTC's materials say the rule allows courts to impose civil penalties for knowing violations. The pages read for this guide give no per-violation figure, so none is stated here.
United Kingdom. The ban on fake reviews sits in the Digital Markets, Competition and Consumers Act 2024, Schedule 20, paragraph 13, and, according to legislation.gov.uk, came into force on 6 April 2025. The CMA's guidance (CMA208) calls it automatically unfair and illegal. As read, it covers:
- Submitting or commissioning fake reviews, and concealed incentivized reviews. Commissioning includes incentivizing by any means.
- Interfering with reviewers, with threats of harm or legal action listed among the practices traders should avoid.
- Publishing reviews in a misleading way, with suppression and cherry-picking as examples, and failing to update star ratings or review counts after fake reviews are removed.
- Publishers, which the guidance says include traders who display reviews on their own sites and intermediaries such as review sites and marketplaces, having a duty to take reasonable and proportionate steps to prevent and remove banned reviews.
- Incentivized reviews: they are not banned outright if the incentive is disclosed and the review still reflects the reviewer's genuine experience. Where a platform allows them, they must be clearly identifiable as incentivized, and where a platform does not allow them, submitting one is likely to be misleading.
The CMA's news release of 27 March 2026 says its sweep covered over 100 review publishers, that it sent advisory letters to 54 firms, and that it launched five new consumer law investigations. The same release says the CMA can fine a business up to 10 percent of its global turnover for infringing consumer law, and that it has not reached any conclusions on whether the law was broken; no fine has been imposed. Check the CMA's current pages for the state of those investigations.
| Tactic | US (FTC rule) | UK (DMCC Act) | Platform rule |
|---|---|---|---|
| Writing or buying fake reviews | Prohibited | Prohibited | Removed, penalties |
| Incentive for any honest review | Allowed if sentiment is not required | Allowed if disclosed and genuine | Varies: Trustpilot bars it |
| Incentive tied to positive sentiment | Prohibited | Review must still reflect a genuine experience | Prohibited |
| Showing only good reviews | Misrepresenting a filtered set is prohibited | Cherry-picking is a banned practice | Google feeds require all reviews |
| Threatening a reviewer | Groundless threats prohibited | Threats of harm or legal action listed as a practice to avoid | Policy breach |
Other laws on harassment and defamation also apply to what you say to a reviewer, in either country.
Related readRebrand reputation search results: what carries overA realistic 90 day sequence
Ninety days is enough to stop the damage that can restrict trading, set up honest review collection, and start moving visible ratings. It is not enough to reverse a poor rating built over years, and a plan that promises that is selling something. The sequence below assumes a store with both its own site and at least one marketplace presence.
- Days 1 to 14: audit every readerExport disputes, early fraud warnings and refunds for the last 90 days. Record each marketplace score and its next evaluation date, the Google store rating and review count, each review platform profile, and what the first page of results for the store name shows. Note which complaints repeat.
- Days 15 to 30: fix the causesCorrect shipping-time statements at checkout, add tracking emails, review statement descriptors and receipts, and publish a plain returns policy. Set rules for answering inquiries the same day. This is the step that lowers disputes and the future one-star reviews.
- Days 31 to 45: open honest review collectionSend one invitation flow to every order, at the same point after delivery, through a channel your Google store rating and review platform accept. No filtering, no incentive tied to sentiment, and a disclosure wherever an incentive of any kind is offered.
- Days 46 to 60: reply and reportAnswer every open review in the format of facts, action, channel. Report only reviews that break the platform's rules, with order evidence. Correct wrong marketplace entries through each platform's own dispute process.
- Days 61 to 75: fix the trust inputsComplete company details, contact channels and policies on the site. Ask scam-alert sites to re-check where their data is stale. Check that site claims about delivery, stock and promises are true.
- Days 76 to 90: measure and decideCompare dispute ratio, marketplace score trend, store rating and review count with the day 1 baseline. Decide what to keep monthly: dispute tracking, review reply cadence and a search check of the store name.
What to do this week
Start with the one number that can stop you trading. Find your dispute and early fraud warning counts for last month, divide by captured payments, and compare the result with the thresholds on your payment provider's monitoring page. Then list the marketplace evaluation dates and the first page of results for your store name.
If a payout hold, a marketplace restriction or a wave of false reviews is already live, a private audit shows which signals are driving it and in what order to act, before you spend on reviews that cannot help yet. Reputation Partners starts that with the private audit.
Monitoring is the part stores skip. A monthly check of ratings, disputes, marketplace scores and the search results for the store name catches a change in the month it starts. The reputation monitoring page describes that routine for brands.
Questions people ask
Do reviews or chargebacks hurt an online store more?
Chargebacks hurt faster, because card networks can fine merchants and, in the end, refuse to process their payments when thresholds are crossed. Reviews hurt conversion and trust gradually. Fix disputes first, since a store that cannot take cards has no use for a good rating.
Can I offer a discount in return for a review?
It depends on the country and the platform. The FTC rule does not prohibit incentives that do not depend on the review's sentiment, and UK guidance allows disclosed incentives for genuine reviews. Trustpilot bars incentivized reviews, and Google requires disclosure in its feeds. Follow the strictest rule that applies.
Can I only ask happy customers for reviews?
No. Trustpilot's guidelines tell businesses not to pick and choose which customers they invite. The FTC rule treats misrepresenting a filtered set of reviews as representative as review suppression, and UK guidance lists cherry-picking as a banned practice. Invite everyone the same way, and deal with unhappy customers in your reply and support queue.
How many reviews does a store need for a Google rating?
Google says a merchant generally needs at least 100 eligible, unique reviews, and the number can vary. Ratings are calculated per country, so reviews from one country do not count toward another. Ads need an average of at least 3.5. Check Merchant Center for your own eligibility status.
Is it legal to threaten a reviewer with legal action?
A groundless legal threat meant to stop or remove a review is prohibited under the US FTC rule. In the UK, the CMA's guidance lists threats of harm or legal action among the practices traders should avoid when reviewers leave negative reviews. A genuine legal claim is different and belongs with a lawyer. Take advice before sending anything.
Why does my store show a low score on a scam-alert site?
These sites compute a score from public signals such as technical setup, contact details, location and reviews from across the internet. ScamAdviser describes its rating as a guide. Complete the company details on your site, then contact the site if the data it holds is stale or wrong.
