Evidence to report a fake review: build the pack

The evidence that helps most when you report a fake review is a short, dated record that does three things: shows the review as it appeared, shows what your own records contain (or lack) about the person or transaction it describes, and names the exact platform rule the review breaks. Everything else, including anger, history and speculation about who wrote it, makes a report longer without making it stronger.

An evidence pack is the document that holds those three things. It is not a legal bundle and it is not a case against a person. It is one page, sometimes two, that lets a moderator who has never heard of you verify your claim in a minute. Most reports fail not because the review was genuine, but because the report gave the person assessing it nothing to check.

This guide covers what to gather, how to check your own records safely, how to capture screenshots, how to match the review to a written rule, and what to keep out. Platform policies change, so it describes the pages read for this guide, with the source named.

The short version

A pack has three parts: a dated capture of the review, a statement of what your own records show, and the exact policy clause. One page is usually enough.
Prove absence from your side: search your own bookings, invoices and messages. Never try to identify the reviewer, contact them or ask others to.
Leave out personal data about the reviewer, guesses about motive, threats and legal posturing. They weaken a report and can create risk for you.
Log every report with date, reason chosen and status, so an appeal is a decision rather than a scramble.
Platforms decide. A strong pack improves the odds on a review that breaks a rule, and does nothing for genuine criticism.

Why a report needs an evidence pack at all

A report is a claim, and a moderator can only act on a claim they can test. Google says on its help page for reporting reviews that only reviews that violate its policies can be removed, that you should not report a review just because you disagree with it, and that it does not intervene in disputes between businesses and customers. That sentence defines the job. You are not arguing that a review is unfair. You are showing that it breaks a rule.

The pack makes that showing quick. A moderator opening a report sees a short reason field and, in some flows, nothing more. In others there is a form, an appeal, or an email thread where extra material can be sent. In every case the same material works: a precise statement, a dated capture, and a clause. Keeping it in one document means you reuse it across platforms, in an appeal, and with a lawyer if the matter ever gets that far.

Which evidence carries weight and which does not

Evidence is strong when a third party can see why it supports the rule you cite, and weak when it only expresses your belief. The table sorts common claims by what you can actually show.

What you claim	Proof you can hold	Strength	Common mistake
No such customer or transaction	A dated note of searches across bookings, invoices, tickets and messages that found nothing	Moderate: shows absence, not intent	Searching only one system, or only by display name
Wrong business	The review text names a service, product, branch or address you do not offer	Strong when the mismatch is explicit	Assuming a vague mismatch proves it
Conflict of interest	A relationship you can document from your own side, such as a former employee or a contractor on your payroll files	Strong if documented, weak if inferred	Naming a competitor on a hunch
Coordinated or repeated posting	Dated captures of several reviews in one window with matching wording, plus your normal baseline	Moderate to strong with a clear pattern	Calling a busy week an attack
Off-topic or abusive content	A capture of the text and the clause it breaks	Strong when the text is plainly outside the rule	Reporting rudeness that the rule allows
Personal information posted	A capture showing the information, with the clause on posting personal data	Strong	Pasting the data again into your report

How to prove you have no record without hunting the reviewer

Check your own systems for the person and the event the review describes, write down what you searched and when, and stop there. The aim is a statement of absence from your records, never a profile of the reviewer.

Start from what the review itself gives you. A display name, a date range, a described service, a location. Then search the places a real transaction would leave a trace, which depend on the business:

The booking or point-of-sale system, by name, date window and described service.
Invoices, receipts and refund logs for the same window.
The shared inbox, contact forms and chat history, searching the name and the service described.
Complaint, warranty or returns records, if the review describes a dispute.
Staff, if the review describes a specific interaction: ask your own team what they remember, framed as checking the record, not as an interrogation.

Write the result as a short, factual note. "On the date of this note I searched our booking system, invoices and inbox for the display name shown and for any job matching the service described between the dates mentioned. I found no matching record." That sentence is accurate, dated and checkable in principle. It claims nothing beyond what you did.

Where the line sits

Checking your own records is ordinary business hygiene. Going further is where people get into trouble. Do not search social networks to work out who the reviewer is, do not match a display name against customer lists beyond what the review itself describes, do not ask friends, staff or customers to identify or contact the account, and do not post or message anything to the reviewer that implies you know who they are. Beyond the ethics, there is a practical reason. If you guess wrong, you may accuse a real customer, and the review now has a true story attached to it. The FTC's staff Q&A on its consumer reviews rule says the rule prohibits making a false accusation about a reviewer, which applies when the business knows the accusation is false or makes it with reckless disregard for whether it is true. A guess about who wrote a review is exactly the kind of statement that can end up there, so do not state as fact what you have only suspected.

If your records do show a real customer who is unhappy, the pack is the wrong tool. Resolve it, reply publicly, and let the review stand or fall on its own terms.

How to capture screenshots that hold up

A screenshot is evidence only if it shows what the review said, where it appeared and when you saw it. Capture the full review, the page address, the date and time, and the reviewer's public display name exactly as shown, and keep the original files unedited.

A usable capture has five parts:

The whole review text and star rating, not a crop of the worst sentence.
The address bar or a copied link to the review, so the capture can be matched to a live page.
The date and time of capture. Your operating system clock in the screenshot, or a note you add in the file name, both work.
The review's own date, as displayed.
The public name and any public profile information shown on the page, and nothing you went looking for elsewhere.

For a pattern across several reviews, capture each one, then add a simple list of the dates and times as displayed. A table of five reviews with their dates tells a moderator more than five screenshots without a summary. Add a baseline: how many reviews the profile usually receives in a comparable period, taken from your own dashboard. Without a baseline, a cluster of reviews is just a cluster.

How to match the review to the exact policy clause

Name one clause, quote its wording, and explain in a single sentence why the review fits it. A moderator reading the clause and the review side by side should reach your conclusion without further help.

Take Google as the worked case, because its policy is written plainly. Google's prohibited and restricted content policy for Maps contributions states that fake engagement is not allowed and will be removed, and it lists content based on a conflict of interest, which it says may include current or former employment or a contractual or consultory relationship. It says to post content based on your experience or questions about experiences at a specific location, which is the off-topic clause. It says not to distribute or post personal information without consent. It says it does not allow content that harasses other people or businesses, and it lists unsubstantiated allegations of unethical behavior or criminal wrongdoing under offensive content. It also says it does allow content that describes negative experiences in a respectful manner, so a harsh but honest account sits on the allowed side. Pick the clause that matches what the review does, not the one that sounds strongest.

For other platforms the same method applies, but the clause has to come from that platform's own page. The policy pages of Trustpilot, Yelp and Tripadvisor could not be read for this guide, so no clause wording from them is given here. Before filing on any of them, open the platform's current content guidelines, find the sentence that matches the review, and quote it in the pack. If you cannot find a matching sentence, that is a finding in itself.

A one-page evidence template

Use one sheet per review, or one per cluster with a table of reviews attached, with every field factual and short.

Field	What to write
Platform and profile	Platform name and the address of your own profile page
Review reference	Link to the review, public display name as shown, star rating, review date as shown
Captured on	Date and time of your capture, and file names of the unedited screenshots
What the review claims	One or two neutral sentences, no adjectives
Records checked	Systems searched, search terms, date window, who searched, date of the check
Result	"No matching record found" or the specific mismatch, such as a service you do not offer
Clause matched	Policy name, section heading, the quoted sentence and the page address where you read it
Why it fits	One sentence connecting the review to the clause
Pattern, if any	Dates of related reviews and your normal weekly baseline
Report made	Date, route used, reason selected, and any reference number

The pack as steps

When the template is filled, the order of assembly matters, because an early step can make a later one unnecessary.

Assembling the pack

Capture before anything elseTake unedited, dated captures of the review and, for a pattern, of each related review, before you reply or report.
Read the review as a stranger wouldDecide honestly whether it describes a real experience, however harsh. If it might, stop and plan a public reply instead.
Search your own recordsCheck bookings, invoices, messages and complaints against what the review describes, and write down what you searched.
Find the clauseOpen the platform's policy page and copy the one sentence the review breaks, with the page address.
Fill the templateComplete the single page, cut adjectives, and check that every statement could be verified by someone else.
Report once, then log itUse the platform's own reporting route, select the reason that matches your clause, and record the date and status.

A worked scenario: a service shop and a job that never existed

This is an invented scenario with generic parties, built to show the sequence. A small bicycle shop receives a one-star review from an account with no profile photo. The review says the shop damaged a wheel during a service job on a stated weekday, refused to refund, and that "the owner clearly runs a scam."

The owner follows the pack. First, a dated capture with the link. Second, a stranger's read: the review claims a specific job, which can be checked, plus a scam accusation. Third, records: the shop's job tickets, till receipts and inbox are searched for the display name and for any wheel job on or near the stated weekday. Nothing matches, and the shop writes the exact search terms and date window down. A staff member also confirms the shop was closed that weekday for stock-taking, which the owner can show from the shop's own schedule.

The clause: the owner cites fake engagement as primary and the unsubstantiated allegations line as supporting. What the shop does not do: it does not search the web for the account holder, ask regulars whether they recognize the name, reply to the review calling it a lie, or message the account. It files one report with the clause and the records note, logs it, and posts a short, neutral reply saying it cannot find a record of the job and inviting the author to contact the shop. Whether the report succeeds is the platform's decision; the pack is built to survive any outcome without harming anyone.

What to leave out of the pack, and why

Leave out anything about the reviewer that you did not get from the public review itself, anything that reads as a threat, and anything that argues the case instead of showing it. A pack is evidence, not a letter.

The same logic covers three other omissions.

Threats and legal posturing. Phrases like "we will sue" or "our lawyers will be in touch" do not move a content moderator and can look like pressure. The FTC staff Q&A says the rule prohibits using an unfounded or groundless legal threat to get a review removed, while a legitimate legal threat is not barred. Either way, a moderator is not the audience: a real legal claim goes through a lawyer, separately.
Offers and incentives. Never offer a refund, discount or gift in exchange for the review being changed or taken down. The FTC Q&A says the rule itself does not prohibit such incentives, but that offering them could be an unfair practice in violation of the FTC Act.
Motive stories. "This is a competitor trying to hurt us" is a belief. Include it only if you hold a documented tie that matches a conflict-of-interest clause.

How to log report status and follow up without losing the thread

Keep one running log, one row per report, and update it every time something changes. The log turns a vague memory of "I reported that last month" into a record you can act on.

Date	Review and platform	Reason selected	Status and next action
Day 0	Review link, platform, capture file names	The clause family from the pack	Reported, awaiting decision
Day 3 to 7	Same row	No change	Check the platform's status view and copy the wording shown
On a refusal	Same row	No change	Re-read the review and clause, decide whether an appeal adds anything
After appeal	Same row	Appeal date	Final outcome, then close the row and note what you learned

On Google, the status view in the Reviews Management Tool shows a report as decision pending, reviewed with no policy violation found, or escalated after an appeal. The help page says evaluation typically takes several days, and that a one-time appeal is available, covering up to 10 reviews in one submission. Because the appeal is single and shared, the log is also how you avoid wasting it: list the reviews you might appeal, rank them by how strong the pack is, and use the slots on the best ones.

When the pack will not be enough

A strong pack improves the case for a review that breaks a rule. It cannot change what the review is, and it cannot make a platform agree.

Genuine criticism. If your own records show the person was a customer, the review is not fake. The route is a reply and, if it fits, a direct attempt to resolve the problem.
Extortion. If a demand for money or favors comes with the reviews, the evidence you keep and the routes you use are different, and the demand itself belongs with the police and your lawyer, not in a content report.

Your next step: one pack, one review

Pick the single review that most plausibly breaks a written rule and build its pack before you report or reply. A thin pack is a useful answer too: it tells you a public reply will serve you better.

If a review looks fake and you are not sure how to prove it

A second pair of eyes helps most before you report, because one report and one appeal are all you may get. Reputation Partners reads the review, your records and the policy side by side, and tells you honestly whether a report is worth filing. Request a private audit.

Questions people ask

Do I need a lawyer to report a fake review?

No. A platform report needs facts and a rule, not legal argument. A lawyer matters when a review makes a serious factual allegation, when someone demands money to stop, or when you are considering a legal claim. For an ordinary report, a clear one-page pack filed through the platform's own route is the right tool.

Can I tell the platform who I think wrote the review?

You can, but it rarely helps and can hurt. A moderator cannot verify your guess, and a wrong guess may name a real customer. State what your records show and which rule the review breaks. Offer a suspected relationship only when you can document it, such as a former employee on your own files.

What if my records show nothing but the review is plausible?

Then the review may be genuine. People book under other names, pay through others or visit without a ticket. If you cannot show a mismatch or a pattern, reply publicly in a neutral tone, invite contact, and keep the capture in case more reviews follow.

Can I report the same review again with better evidence?

On Google, the help page describes a one-time appeal for reviews that were reviewed with no violation found. Beyond that route, repeating the same report with the same material rarely changes the outcome. Improve the evidence first, and check the platform's current help page for any other route before filing again.