The 7 review sources that dominate every category (and how much to trust each)

By Ryan Siegal · Founder and Principal

Published 2026-04-23

The pattern across every category

Every consumer category falls into a similar review-source ecology:

• Professional critics — editorial publications with paid reviewers and editorial standards. Small sample size but high per-review information density.
• Verified-purchase aggregators — platforms where reviews are tied to actual transactions. Large samples, medium rigor, high inflation pressure.
• Enthusiast crowd platforms — community-driven sites where passionate hobbyists rate. Large samples, medium rigor, slightly less inflation than general consumer platforms.
• General consumer platforms — anyone-can-review sites (Google Reviews, Yelp). Very large samples, low rigor, highest inflation pressure.

Our answer to that ecology is not to rank the tiers against each other. It is to read whichever of them we can reach, normalize every reviewer against their own history — subtract their personal mean, divide by their personal standard deviation — and rank inside a peer cohort, with every qualifying reviewer counting equally. The method is written out at /methodology. The tiers in the tables below are a separate thing: an argument about which sources are worth your attention, not a coefficient anything gets multiplied by.

What Rankquant reads today

None of the professional publications in the tables that follow is ingested anywhere, in any category — so those tables are landscape analysis, not a description of a shipped computation. The live inputs are crowd platforms:

So the rest of this article is an assessment of the review landscape: which sources repay your attention when you go read them yourself, and which do not. None of it feeds a Rankquant number.

Wines — who to trust

Movies & TV — who to trust

Books — who to trust

Amazon consumer products — who to trust

Hotels — who to trust

Restaurants — who to trust

Why publish an opinion no score depends on

Every other review aggregator treats its weighting as proprietary secret sauce. Metacritic doesn't tell you why a New York Times review is weighted more than a Variety review. Rotten Tomatoes doesn't explain which reviewer gets which weight.

The opposite move is to have no secret sauce, because there is no weighting to keep secret. Every qualifying reviewer we read counts once, and the whole computation — admission rules, per-reviewer z-scores, the confidence correction and the percentile mapping — is written out at /methodology, so a reader can rerun it line by line. The tier lists above are an opinion you can disagree with at no cost to the arithmetic, precisely because no score reads them. The output is deterministic: same inputs, same arithmetic, same score.

The same transparency is how we answer accusations of bias on the scores that are live. If a brand claims Rankquant is under-ranking their product, we can point to the reviewer set, the raw ratings, each reviewer's own baseline, the published CI-floor diagnostic and the peer set — and demonstrate that the normalized output is mechanically correct given the inputs.

Related: The full methodology · Rating inflation explained · Bayesian averaging