How a Proportionate Response Beat a False Online Review

The situation

Eun-ji teaches at a university in the Niagara region and, alongside her teaching, runs a small independent research consultancy that conducts survey work and data analysis for corporate clients. The consultancy is not large — a handful of contract researchers, no storefront, no advertising beyond referrals and a professional profile on a well-known business review platform. Most years it brings in modest, steady income on top of her teaching salary, and Eun-ji had always treated it as a careful, part-time practice built almost entirely on reputation and word of mouth. This year was different: a long-standing corporate client was weeks away from renewing a multi-year research agreement worth roughly $600,000 over its term, by far the largest contract the consultancy had ever held, and one Eun-ji had spent nearly two years building toward.

Her spouse, Min-ji, a physiotherapist, was the one who first noticed the review. It appeared under a real name — Miriam — and described Eun-ji's consultancy in specific, damaging terms: that survey results on a recent project had been altered before delivery to make outcomes look more favourable to the client than the underlying data supported. It was not a complaint about slow turnaround or a difficult personality. It was a factual allegation of fabricated research, the single most damaging thing that could be said about a firm whose entire value proposition is data integrity. Min-ji screenshotted the review before showing Eun-ji, sensing correctly that it might disappear or change once the reviewer realized how serious the wording was.

The timing was what turned an insulting review into a genuine business emergency. The consultancy's business review profile was one of the first results when anyone searched its name, and the renewal client's procurement team had a documented habit of checking public reviews before signing off on contract extensions. A single unresolved allegation of fabricated data, sitting online for even a few weeks, had the potential to derail a renewal that had otherwise been a formality.

What the review said

The reviewer, it turned out, was a former subcontractor the consultancy had stopped using the previous year after a dispute over invoicing. Ontario defamation law draws a sharp line between opinion and fact. A review that says "I found the reports overpriced" or "communication was slow" is an opinion, and opinions about services rendered are almost always protected, however unflattering. A review that asserts a specific, verifiable fact — that data was altered — is different. If that fact is false, and it damages reputation, it can be defamatory, regardless of the format it appears in.

The review crossed that line clearly. But the team's first task was not to draft a lawsuit. It was to slow down and assess the situation the way any measured response requires: was the statement actually false, was it actually about identifiable conduct, and was it actually causing harm, or merely uncomfortable? Ontario also has protections against lawsuits used to silence legitimate public criticism, sometimes called anti-SLAPP protections. A business that responds to any negative review with a threatening legal letter risks looking like it is trying to bully a critic into silence — which can backfire badly, both in court and in public opinion, even when the underlying claim has merit. The response needed to be proportionate to what had actually happened: a specific false factual claim, not general criticism.

What we did

1. Confirmed the claim was factual, not opinion, before treating it as defamation. We reviewed the underlying project file with Eun-ji, the delivered reports, and the correspondence with the client at the time. The data had not been altered. This mattered — a response built on a claim that turned out to be true, or even partly true, would have collapsed immediately and made the situation worse.
2. Documented the harm rather than reacting publicly. Eun-ji was, understandably, tempted to post a furious public reply the day she saw the review. We advised against it. Public back-and-forth arguments in a review's comment section rarely help and often extend the life of the very content the client wants removed. Instead, we kept a record of the review, its timing relative to the contract renewal, and any signs the renewal client had seen it.
3. Sent a measured written notice to the reviewer. Rather than a threatening demand letter, the first communication to Miriam set out the factual inaccuracy plainly, attached nothing inflammatory, and requested a retraction and removal within a stated reasonable period. The tone mattered as much as the content — it needed to read as a firm correcting a false record, not as a business trying to punish a critic.
4. Reached out directly to the renewal client with evidence. In parallel, Eun-ji contacted her corporate client's project lead, shared the original data files and delivery records, and addressed the allegation head-on before the client could form an impression based on the review alone. Getting ahead of the narrative with the one relationship that actually mattered was, in practical terms, more important than anything aimed at the reviewer.
5. Escalated to the review platform on a separate track. When Miriam did not respond within the stated period, we prepared a formal notice to the platform itself, using its policy for factually false reviews rather than a general complaint. Platforms are usually far more responsive to a specific, evidenced claim of falsity than to a general dispute about tone or fairness.
6. Kept litigation in reserve rather than leading with it. Throughout, Eun-ji was advised that a defamation claim in the Superior Court remained available if the review stayed up and the harm continued, but that filing one immediately — before trying to resolve the matter directly — would have been disproportionate to what a first, carefully worded notice could likely achieve, and would have taken many months to resolve regardless.

The outcome

Miriam removed the review within days of the formal notice and, through an exchange of emails, provided a short written acknowledgment that the data allegation was not accurate. The consultancy's corporate client renewed the multi-year agreement at its full value of roughly $600,000, having already received the underlying project files directly from Eun-ji before the renewal decision was made. No claim was ever filed in court, and no public exchange of accusations ever took place on the review platform itself.

The measured approach did more than avoid a fight — it protected the relationship that actually generated the revenue. Had Eun-ji responded publicly and aggressively the day the review appeared, the renewal client's first exposure to the dispute might have been an angry public exchange rather than a calm, evidence-backed conversation initiated by Eun-ji herself. The order of operations, not just the legal merits, determined how the situation resolved. A technically correct legal position, deployed clumsily or too loudly, can still cost a business the very relationship it was trying to protect.

The total legal cost of resolving the matter was a small fraction of the contract at stake, and the timeline ran to a few weeks from first notice to retraction — fast by litigation standards, though only because the matter never became litigation. A defamation claim taken all the way through the Superior Court, by contrast, would typically have taken well over a year to resolve, with legal costs and reputational uncertainty stretching across the entire renewal window. Min-ji later said the hardest part had not been drafting letters or gathering files, but resisting the urge to respond publicly in the first forty-eight hours, before there was a clear plan. Eun-ji kept the consultancy's review profile largely unchanged afterward, aside from a brief, factual note thanking clients for their patience during "a since-resolved data accuracy question" — worded carefully enough to close the loop without reopening the dispute.

What you can learn from this