Methodology

How PROOF ratings are built

PROOF is an independent rating standard for the beverage-alcohol industry — distributors, retailers, and brands. This page explains exactly how a score is produced, what it does and does not mean, and the rules that keep the standard trustworthy. Every score on PROOF links here.

Last updated June 2026 · Prototype methodology

1. Scores are statements of opinion

Every PROOF score, grade, and the commentary attached to it is a statement of opinion derived from a transparent, published methodology applied to aggregated public data and verified, brand-submitted reviews. Scores are not statements of factabout any company's conduct.

Where commentary appears alongside a rating, it is framed as our opinion or tied directly to the underlying rating or reviews (for example, "rated lower on reporting cadence") — never as a bare factual accusation.

Plainly: a PROOF score is our considered editorial assessment, the way a critic rates a film or an analyst rates a stock. Reasonable people can disagree, and we publish the inputs so you can judge for yourself.

2. What goes into a rating

Each rated party is scored on a set of weighted dimensions that differ by tier (distributor, retailer, brand). Every dimension is tagged by the kind of evidence behind it:

The full dimension list, the weight of each dimension, and its evidence tag are published on every entity's profile so the score can always be traced back to its inputs.

3. How the composite score is computed

Each dimension is scored 0–100. The published composite is the weighted average of a tier's dimensions, using that tier's standard weights (which sum to 1.0):

composite = round( Σ (dimension_value × weight) ÷ Σ (weight) )

You can re-weight the dimensions yourself in the product to see how the composite would change under your own priorities. That recomputation is for your exploration only — the published, standard score always uses PROOF's standard weights.

Achievements and badges (Top X, Rising Star, Fastest Mover, #1 in region, A-Rated, Independently Verified, and similar) are computed from a company's scores, its percentile within its tier, and momentum. They are recomputed at read time, not paid for or manually assigned.

4. Grades and color bands

The 0–100 composite maps to a letter grade and a color band so a rating is readable at a glance. The same published thresholds are applied to every company in a tier — there are no per-company exceptions.

A · Exceptional · 85–100B · Strong · 75–84C · Solid · 65–74D · Watch · 55–64E · Caution · below 55

Exact numeric cutoffs are fixed in the methodology and applied uniformly. They are not adjusted for any individual company.

5. Reviews and reviewer verification

Reviews are third-party content. They reflect the views of the brands and operators who submit them, and we display them substantially as submitted — we do not rewrite a review into a factual claim of our own.

6. Notice and takedown

Any company can flag a review it believes is false or violates our policy. A flagged review moves into review and may be removed while we assess it. We aim to respond promptly.

To flag a review or raise a dispute, email hello@proofratings.com with the company name, a link to the rating or review, and the basis for the request. A rated company can also claim its profile and respond publicly to reviews.

7. No pay-to-play

This is the core promise of the standard: no payment, plan, or commercial relationship can change a score. Scores derive only from data and verified reviews. There is no path — for a rated party, an advertiser, or PROOF staff — to buy, edit, or suppress a published rating.

PROOF earns revenue from data, tools, analytics, and lead products — never from selling rank. Paid plans unlock analysis features; the number a company is rated is never gated by payment and never editable by payment.

8. Conflicts of interest

To avoid a conflict of the kind that has discredited other rating systems, any brand or business operated by PROOF's operator is clearly labeled as operator-affiliated and fenced outof any "best of" or ranking surface it would otherwise top. We would rather disclose and exclude than rank something we have a stake in.

Prototype notice. PROOF is currently a working prototype. Company and brand names shown in the product are real and used only to illustrate the model; all scores, momentum figures, reviews, prices, and deal data shown today are placeholder figures, not actual ratings or offers. We will not publish ratings on real companies until verified data and these guardrails are fully live.