Do comparison pages (us vs a specific competitor) genuinely help AI engines recommend us over that competitor?
Competitor comparison pages get cited when they carry real criteria and honest losses, and get ignored when they read as sales collateral. AI engines assemble shortlists from pages that state differences in extractable form: a criteria table, named trade-offs, a stated verdict. Google publishes no special requirement for this content type (per Google Search Central, September 2026); structure does the work.
Why the format matters more than the argument
A shopper asks an engine "X or Y for [use case]". The engine needs rows it can lift: what differs, by how much, and for whom. A comparison page built as a feature-checkmark grid where your column is full and theirs is empty supplies no rows worth lifting: every vendor publishes that grid, and it carries no information.
What a usable comparison page contains
- A criteria table with the criteria a buyer actually weighs, not the ones you win on.
- At least one criterion where the competitor wins, stated plainly. This is the credibility signal, and it is the one most merchants refuse to include.
- A verdict that names the buyer profile: "choose X if you run multi-location inventory; choose Y below 500 SKUs." A verdict with no profile is an advertisement.
- Dated figures. Undated pricing on a competitor page is the fastest way to be wrong in public.
The uncomfortable implication. Publishing an honest comparison means publishing the cases where you lose. Merchants who cannot accept that should not build the page: a dishonest comparison page is worse than none, because it trains a buyer to discount everything else on the domain.
The legal edge case for mid-market teams: comparative claims about a named competitor are advertising claims. Get them reviewed, keep the dated evidence, and re-verify quarterly. That is an ops commitment, not a one-time publish.
The Deploi point of view
Our own position, from building on Shopify. Separate from the facts above.
- Our take: Comparison pages work, and most of them are built wrong. The one design choice that changes the outcome is including a criterion where you lose.
- What we’ve seen: Comparison pages age badly and silently. Competitor pricing moves, the page keeps asserting the old number, and the page that was your credibility asset becomes the reason a buyer stops trusting the site. A dated re-verify lane is part of the build, not a nice-to-have.
- Where we disagree: Marketing treats the comparison page as a bottom-funnel conversion asset and writes it to win. Engines reward it as a reference document. Those two briefs produce different pages, and only one of them gets cited.
- What this page adds: the specific structural requirements (criteria table, an admitted loss, a profiled verdict, dated figures) that decide whether the page is usable.
Reviewed by Martin Dejnicki, Director of SEO & AI Search. Facts verified 2026-09-13.