How we test, rank, and write about gear.

Editorial principles

We use it before we publish. Every first-party review on Vanguard Loadout is written from real use. Length of use varies by category — wearables resolve faster than a sauna — but the standard is the same: enough time to have an opinion that will not change next week.

Research-driven coverage is labeled. When we cover a product or brand we have not personally tested — as part of a buying guide, or because the category requires breadth — we label it research-driven and explain what we used to evaluate it.

We disclose every commercial relationship. Affiliate links are disclosed at the top of every review and guide, before the first link. Free units, loaned units, and any other non-arm's-length arrangement are disclosed inside the piece. See /disclosure/ for the full statement.

Commission structure does not influence ranking. This is the central editorial commitment. Brands with higher commission rates do not get better placement, more flattering language, or shorter caveat lists. If a budget product outperforms a premium product for the use case in question, we say so. If a premium product is worth the spend, we say that too.

We re-evaluate when products materially change. Significant firmware updates, sensor changes, subscription model shifts, or design revisions trigger a re-look. Pieces that need updating are updated; pieces that need pulling are pulled.

Sponsored content, if accepted, is quarantined. We do not currently accept sponsored content. If that ever changes, sponsored placements will be labeled, separated from review and guide content, and excluded from "best of" rankings.

What you will see in a review

A clear verdict near the top. A statement of who the product is for and who it is not. What we tested, for how long, in what conditions. What it does well, where it falls short. Specs. Alternatives we considered and how they compare. A final recommendation with price context.

Long-term reviews carry an After-Action designation. The standards are the same as any other review; the designation signals enough accumulated use to draw conclusions a first-impressions review cannot. The threshold for that designation varies by category, because what counts as long-term differs meaningfully between a wearable, a cold plunge, and a topical product.

What you will see in a buying guide

A category-level evaluation with a top recommendation, a premium pick, a value pick, and use-case-specific picks where the category warrants them. The first-party-versus-research-driven labeling applies — we tell you which is which.

When something is removed from a guide — because it has been superseded, or because longer use changed our view — we say so plainly. We call that a drop, and the reasoning is part of the update note.

Evidence hierarchy

Where claims on Vanguard Loadout are evidence-grounded, we distinguish clearly between:

  • Randomized controlled trial data — highest weight

  • Observational and registry data

  • Mechanistic and animal-model evidence

  • Manufacturer studies and self-reported user data

  • Anecdote and influencer claims — lowest weight, often noted to flag hype

A product that performs in a well-designed RCT gets weighted differently from one that performs in a manufacturer-sponsored case series. Both can be useful. They are not equivalent.

A note on uncertainty

The optimization space sometimes gets ahead of the science. Where the evidence is solid — randomized trial data, replicated outcomes, established mechanism — we say it is solid. Where the data is observational, animal-only, or thin, we say that. Where a category is genuinely promising but unsettled, we say that too. The point is to help readers calibrate, not to oversell.

Questions

If you have a question about a specific recommendation or a methodology call we made, write to us at field-notes@vanguardloadout.com.