How to spot a fake Google review: what to look for in the reviewer's profile

A one-star review lands and something feels off. The complaint doesn't match anything that happened. The name doesn't appear anywhere in your booking system. But "feels off" isn't a case, and Google doesn't remove reviews because an owner is certain.

So before you flag anything, do the thing most owners skip: click the reviewer's name and read their profile properly. That's where the actual evidence is, and it takes about four minutes.

This matters practically, not just intellectually. You get one appeal per review. Spending it on a hunch, in the wrong policy category, means the review stays permanently even if you later find better evidence. Establishing what you're looking at first is not delay — it's the whole job.

Why the profile matters more than the review

A well-written fake review and a genuine complaint from an unhappy customer are, on the page, often indistinguishable. Fabricated text doesn't announce itself, and increasingly it's generated by tools that produce fluent, plausible, specific-sounding prose.

What's much harder to fake convincingly is a history. An account that exists to leave a review has a shape — thin, bursty, geographically incoherent, oddly focused. An account belonging to a real person who happens to have had a bad experience at your business has a different shape: sporadic, local, unremarkable, spread across the ordinary businesses of an ordinary life.

You're not looking for a smoking gun. You're building a pattern.

Reading the review history

Open the profile and look at the whole contribution list, not just your review.

Account age against review count. A young account with a large number of reviews is the single most common signal. Real people accumulate reviews slowly, over years, in no particular hurry.

Whether your review is their only one. An account whose sole contribution to Google Maps is the one-star it just left on you is the clearest pattern there is — particularly if several such accounts arrive together.

Posting cadence. Look at the rhythm. Genuine review activity is irregular: a cluster during a holiday, then silence for eight months, then one more. Accounts that post in tight, high-volume waves — especially repeatedly — are behaving like accounts with a workload rather than a life.

Category clustering. If most of their reviews are of businesses in your specific industry, that's worth noting. Real people review a restaurant, a mechanic, a dentist, a hardware shop. An account reviewing eleven businesses in your exact category, in your exact area, is either a trade professional or something else.

Rating distribution. An account that has left nothing but one-stars, or nothing but five-stars, isn't behaving like a customer. It's behaving like an instrument.

Geography and timing

This is where thin operations usually break down.

Look at where their reviews are. A real person's reviews cluster around where they live, work, and holiday. An account leaving reviews in four cities several hundred miles apart within the same week is describing a travel schedule almost nobody has.

Then look at the timing against your own records. If someone describes a detailed service experience on a Tuesday afternoon and you were closed on Tuesday, that's concrete and worth documenting. Same for a service you don't offer, a staff member who doesn't work there, or a product line you discontinued two years ago.

The reviews that follow a burst pattern — several landing within 24 to 72 hours from unconnected-looking accounts — are worth treating as a single coordinated event rather than as separate problems. If a payment demand follows, you're looking at review extortion, which has its own reporting route and a much stronger case for removal.

Signals in the text itself

Weaker evidence than the profile, but still useful in combination.

• Generic complaints with no verifiable specifics. Real dissatisfied customers name things: a date, a dish, a price, a person, a wait. Fabrications tend to stay abstract because the writer doesn't know the details.
• Details that are wrong for your business. Complaints about parking when you're inside a pedestrianised centre. A reference to a service you've never provided.
• Near-identical phrasing across multiple reviews. Compare the suspect reviews against each other. Shared sentence structures, repeated unusual word choices, or the same complaint restated are a strong signal of common authorship.
• Suspiciously effusive detail on positive reviews. This cuts the other way too. If you're assessing a competitor's profile, unusually long, thorough, generous reviews from young accounts are a known pattern — most genuinely happy customers write two sentences.

The Local Guide badge means less than you think

There's a widespread assumption that a Local Guide badge, particularly a high-level one, indicates a trustworthy reviewer. Treat that assumption carefully.

The badge is earned through contribution volume, not honesty. It measures how much someone has posted, not whether any of it was true. And established accounts with strong histories have real value to people running manipulation operations — which is exactly why aged, high-level accounts get bought and sold.

So apply the same analysis regardless of badge. A Level 7 Local Guide whose reviews cover six countries in a fortnight and cluster suspiciously around one industry deserves more scrutiny than an unbadged account with a decade of ordinary local activity, not less.

What is not evidence

Being honest about this will save you a wasted appeal.

• A harsh review is not a fake review. A genuine customer describing a genuinely bad day at your business is protected content, however unfair it feels and however much it costs you.
• No profile photo is not evidence. Enormous numbers of ordinary people have never set one.
• A name you don't recognise is weak on its own. People book under partners' names, use nicknames, arrive as someone's guest, or bought from you three years ago.
• A first-time reviewer is not automatically suspect. Everyone's first review is a first review. It's the combination with other signals that matters.
• Disagreeing with the facts as stated is not grounds for removal. Google does not adjudicate whose account of an evening is accurate.

If your case rests only on items from this list, it will fail, and you'll have spent your one appeal establishing that.

The honest limits of this

You are building a probability case, not proof.

You cannot see the account's IP, its creation device, or its connection to any other account. Those signals exist, but they sit with Google, not with you. What you can do is document a pattern coherent enough that a human reviewer looking at the same profile reaches the same conclusion you did.

That's a real thing, and it's often enough. It is also frequently not enough, and no amount of forensic care changes that — which is why the sensible posture is to investigate carefully, report once, and then put your energy somewhere with a better return.

What to do once you're confident

Write it down before you flag anything. A short document: the review with its timestamp, screenshots of the reviewer's profile and contribution history, the specific pattern you identified, and your own records showing no customer relationship exists across all date ranges.

Then map it to a specific Google policy — fake engagement, conflict of interest, spam, off-topic — and file it in that category rather than defaulting to spam for everything. Miscategorising is the most common reason a legitimate report fails.

And then, whatever the outcome, the arithmetic. A business with fifteen reviews has to win every one of these fights, because each fake review visibly moves its rating and sits near the top of a nearly empty profile. A business with three hundred genuine reviews, a dozen from the past month, can lose all of them and barely notice — the average holds, the fake is buried within days, and any customer who does read it sees it surrounded by hundreds of accounts saying the opposite.

Learning to read a reviewer's profile is a useful skill. Not needing the outcome is a better position.