How we pick winners

Last updated April 29, 2026

This is a public methodology document. It is the rubric every editor at What's The Best Report uses to pick the single winner that goes in the yellow callout at the top of every article. The methodology is deliberately three steps. We have rejected proposals to expand it; one of the things this publication is built to do is keep verdicts uncluttered.

Step 1 — Category-defining criteria

Before we research a winner, the editor on the byline writes the criteria that define the category. The criteria are not tie-breakers. They are the question itself. For example, in "what's the best smart ring 2026," the criteria are: HRV measurement (the actual physiological signal a ring claims to measure), comfort over a 30-day continuous wear, battery life (real, not advertised), data export (do you own your readings), and the absence of a subscription paywall.

The criteria are written before the editor chooses a winner. This is methodologically important: it stops the rubric from being constructed to fit a pre-decided answer.

Step 2 — Head-to-head against the strongest contenders

Every article names the strongest credible alternatives the editor considered, and each one is run against the criteria from Step 1. The runner-up section ("Also considered and didn't win") is published in every article. If a runner-up is closer to the winner than the prose can defend, the editor escalates to Theodora and we either rewrite, re-test, or do not publish.

Two outcomes from Step 2 have, in practice, happened twice each since the publication launched. First, the article gets rewritten with a different winner — the editor's draft was wrong and the runner-up is the actual answer. Second, the article gets killed — there is no defensible single winner and we are not in the business of publishing "it depends." Articles 11 and 14 in our internal queue were killed for this reason.

Step 3 — Honest cons and the defensible single answer

The published article must contain the strongest argument against the winner. If a reader finishes one of our verdicts unable to articulate the case against the pick, we have failed. The "honest cons" section is a hard requirement, not a courtesy. It is the section we read most carefully in editorial review — both because it's the section the rest of the press tends to suppress, and because if it's weak, the verdict probably is too.

Editorial review chain

1. Author drafts. The category-owning editor writes the verdict, the criteria, the head-to-head, the cons, and the FAQ.
2. Cross-category gate. Sleep claims go through Yolanda. Nutrition claims go through Camille. Sensor and radio claims go through Bjorn. AI/coding claims go through Ravi. The cross-category gate exists even when the gating editor isn't the byline.
3. Editor-in-Chief sign-off. Theodora reads the final draft and signs off on the verdict before publication. Articles are not published without this signature.
4. Update cadence. Every published article is reviewed at minimum every 6 months. The reviewedDate appears on every article and is also written to the public update log.

What can move a verdict

A verdict moves when (a) the winner ships a substantive change that affects one of the Step 1 criteria, (b) a competitor ships something that genuinely closes the gap on more than one criterion, (c) we discover an error in our prior research, or (d) the price changes by more than 25%. We do not move verdicts because of news cycles, app-store rating drift, or marketing announcements that don't ship. Every move is logged in the public update log with a written reason.

Why this is short

A common pattern in app-review publications is a 4,000-word methodology document weighted across 17 criteria, each with sub-rubrics. We've read those documents. They are not in the service of the reader; they are in the service of looking thorough. Three steps, written clearly, signed by named editors, are sufficient to pick one winner and defend it. Anything longer is performance.

Conflicts of interest

Every editor on the masthead has a conflict-of-interest statement on their author page. Bjorn worked at Garmin from 2014 to 2019; the smartwatch-for-runners winner on this site is the Garmin Forerunner 265, and the article notes this prior employment. Theodora reviewed and signed off on that verdict before publication. Ravi owns Asana stock he received as an employee; Asana is not currently named in any winner verdict. Camille has occasionally received free trial accounts to enterprise tiers of nutrition apps for testing; this is disclosed in the relevant articles. We are not perfect at this, but we publish the disclosures so readers can audit them.

What we do not do

We do not maintain affiliate tracking links. We do not accept sponsored placements. We do not let vendors review or comment on drafts before publication. We do not gate methodology behind a paywall. We do not publish "best 17 ___" listicles. The single-winner format is the editorial product, not a stylistic choice — listicles were the failure mode we built this publication to fix.