Magrios / Knowledge / Market Growth / What to do when a competitor publishes a compari

What to do when a competitor publishes a comparison against you

Guide · Market Growth · 4 min read · last verified 2026-07-27

Reviewed before publication Editorial board Independent commercial review
In shortBefore responding to a rival's comparison page, check whether AI answers actually cite it. If it has no footprint, stand down; if it does, publish your own honest comparison — concessions and all — and never start a rebuttal war.

A competitor comparison page is a page a rival publishes on their own site comparing their product to yours — their name versus your name — written on their terms, structured around criteria they chose, and reliably concluding in their favour. Discovering one aimed at you feels like an attack, and the instinct is to respond immediately and in kind. The right first move is neither. It is to find out whether the page actually influences anything — because a comparison nobody cites is a problem that solves itself, and a comparison AI assistants repeat is a different problem from the one your bruised instincts think you have.

Evidence before reaction

The page's existence is not the harm; its influence is. Before anyone drafts a response, run three checks.

First, do AI assistants cite it? Ask the comparison questions a buyer would ask — the two product names together, best tool for the shared use case, whether your product is worth it — and record whether the rival's page appears among the cited sources. Second, does the page surface for the comparison searches buyers actually make, or only when you go looking for it? Third, and most subtle: do the page's characterisations of you leak into answers without attribution — the framing absorbed even where the link is absent? That last pattern is the most consequential, because it shapes buyers who will never see the page itself.

This is a scan, not a vibe. Magrios runs exactly this kind of check as part of competitor research — which pages assistants cite for each buyer question, with every claim openable to its source — so the decision that follows rests on evidence rather than on how the page made your team feel the morning it was found.

If it is not cited, mostly stand down

A comparison page with no footprint in answers or search is a rival's content bet that has not paid off. Responding to it elevates it: linking to it gives it authority, naming it gives it reach, and an internal fire drill gives it exactly the attention its authors hoped for. Add it to the watch list, re-check on your normal scan cadence, and spend the energy on your own comparison content built around your buyers' questions rather than their framing.

Standing down is organisationally hard — sales will be upset, someone senior will forward it with a one-line email — which is why having the citation evidence matters. The answer to should we respond is not a matter of pride; it is a measurement.

If it is cited, publish your own honest comparison

When the page does shape answers, the response is not a rebuttal of their page. It is a better page answering the same buyer question: your own comparison, structured around the criteria buyers in your category actually weigh, stating plainly where you win, where the rival genuinely wins, and which kind of buyer each product fits.

The concessions are not a courtesy — they are the mechanism. A comparison claiming total superiority reads as marketing and is easy for both buyers and answer engines to discount; a balanced page that concedes real weaknesses is the kind of source that gets drawn on when an assistant needs to answer the comparison question. Your rival's page is usually one-sided by construction. Balance is the one advantage they have structurally chosen not to compete for.

Fix flat factual errors, narrowly

Biased framing and factual error are different problems. If the rival's page states things about your product that are checkably false — a capability you have that it says you lack, a pricing model you no longer use — make the truth easy to verify: your own documentation should state the current fact plainly, dated, where any buyer or crawler can find it. A brief, unemotional correction request to the competitor is sometimes worth sending for outright errors; stale claims are more common than invented ones, and some get fixed. What you are doing is narrowing the page's advantage to pure opinion, which ages badly. What you are not doing is arguing.

Never a rebuttal war

The point-by-point rebuttal page — their claims quoted, your objections appended — fails on every axis that matters. It adopts their framing wholesale, so even a winning argument happens on their chosen ground. It amplifies the original, which must be named and linked to be rebutted. It produces content organised around a dispute rather than around a buyer's question, and buyers researching a comparison want a decision aid, not a transcript of two vendors arguing. And it never ends: they revise, you revise, and the freshest wound is always theirs to choose. Every hour spent rebutting is an hour not spent answering the questions your buyers are actually asking.

This is not the unknown-vendor problem

It is worth keeping this playbook distinct from a neighbouring one. When an unknown vendor outranks you in AI answers, you face a visibility gap: someone you were not watching is winning share of answers across many questions, and the response starts with figuring out who they are and why answers favour them. A competitor's comparison page is the opposite shape — a known rival, a deliberate artifact, concentrated on one question surface. The comparison-page situation is narrower and more tractable: you know the author, the target, and the battleground. Diagnose with the citation checks, respond with your own honest comparison if warranted, and record what their page claims in the rep-facing battlecard so sales walks into deals knowing what buyers may have read. Then go back to the wider map — one rival's page is one square on it.

Frequently asked questions

A competitor published a comparison page against us — should we respond at all?

Only if it has influence. Check whether AI assistants cite it on comparison questions, whether it surfaces in the searches buyers make, and whether its framing leaks into answers. No footprint means no response — monitoring and your own content are better uses of the energy.

Do competitor comparison pages actually affect AI answers?

They can, in two ways: as cited sources on comparison questions, and as absorbed framing that shapes how answers characterise you even without a link. That is why the first step is measuring citations rather than reacting to the page's existence.

Why not publish a point-by-point rebuttal?

A rebuttal adopts the rival's framing, amplifies the page it must name, and produces a vendor dispute instead of a decision aid. Buyers researching a comparison want help deciding. A standalone honest comparison answers their question; a rebuttal answers your grievance.

Why would conceding weaknesses in our own comparison help us?

Because balance is what makes a comparison usable as a source. One-sided pages are easy for buyers and answer engines to discount. Conceding where the rival genuinely wins earns credibility for your claimed strengths — an advantage their page structurally gave up.

Further reading — chosen for this article
Entities in this research
Magrioscompetitor comparisonresponse playbookAI answershonest comparison
Related knowledge

How to find competitors you didn't know you had · shared entities

Where are you losing buyers you never see · shared entities

How AI affects your category narrative · shared entities

What AI answers reveal about competitor positioning · shared entities

How to prep a sales call with AI answer research · shared entities

Recently updated

What is Citation surface? A practical definition · 2026-07-27

Why market intelligence needs a memory · 2026-07-27

Why AI referral traffic is undercounted · 2026-07-27

The five growth questions for fintech companies · 2026-07-27

Where does your brand stand?
Check your AI visibility free — real evidence, not a score.
Check my visibility or run the full analysis →