Most businesses discover a fake review the same way: a notification, a founder forwarding a screenshot, a sudden question in a board meeting. The instinct is to demand it down immediately. The more useful first question is which of two very different processes actually applies.
What Google will remove on its own
Google's policy covers seven categories: spam and fake content, reviews unrelated to a genuine customer experience, illegal content, conflicts of interest such as a review from an employee or a competitor, personal information, sexually explicit material, and — added in early 2026 — AI-generated review content, even where the underlying sentiment happens to be genuine.
Flagging through the standard reporting tool resolves a real share of these, particularly obvious spam and reviews from accounts with no transaction history. It does not resolve genuine-sounding negative reviews Google has no policy basis to touch, however unfair the account.
Where POHA applies, and where it doesn't
Singapore's Protection from Harassment Act was amended in 2020 to add several court orders dealing with online falsehoods. Section 15C — the disabling order — is made against an internet intermediary, which includes Google, requiring it to disable access to the offending content. This is the mechanism that applies when a review states something false as fact, not when it expresses an unfavourable opinion about a genuine experience.
An Expedited Protection Order can compel action within days for clear cases. A standard POHA application moves on a normal court timeline, typically weeks, and requires evidence the statement is both false and causes alarm or distress — not simply reputational inconvenience.
A POHA application and a defamation claim are not the same instrument, and using the wrong one wastes the time advantage either offers.
Fact versus opinion is where most disputes actually sit
"The food was cold" is an opinion about a real experience and stays up regardless of accuracy. "This company stole my deposit" is a factual assertion, and if false, it's the kind of statement POHA and defamation law are built to address. Reviews that blend both — a real complaint plus a false factual claim — usually need the false portion addressed specifically rather than the whole review contested.
The realistic sequence
1. Flag through Google first. Free, fast where it applies, and resolves clear policy violations without escalation.
2. Establish whether the reviewer exists as a real customer. A review from someone with no record in your systems is a different problem from a genuine customer's unfair account.
3. Separate fact from opinion in the text itself. This determines whether POHA or defamation is even the right tool.
4. Use a Letter of Demand for fast-moving cases. Against a named, contactable individual, a properly drafted letter frequently produces a takedown faster than litigation, because most people don't want a formal record.
5. File the court application only where the above fails. It works, but it's the slowest of the available routes.
What doesn't help
Responding publicly with a detailed rebuttal usually keeps the review visible longer and adds content search engines index alongside it. A calm, brief public reply that invites the reviewer to contact you directly, paired with the removal process running privately, is the better sequence.