Editorial standard

Make the evidence visible.

Our job is to make practical tool research easier to inspect. We show where material claims come from, when a source was checked, what was actually tested, and what remains uncertain. Inventory records, vendor claims, observations, and editorial assessments use different labels so one kind of evidence is not mistaken for another.

Methodology version 2.0 · 13 September 2026

01 / Selection

Why a tool is covered

Coverage starts with a clear user problem and something we can evaluate. Payment, sponsorship, or affiliate availability does not determine inclusion or order.

02 / Sources

What supports a claim

Material product claims are checked against dated primary sources. Secondary sources may add context. They are identified and are not treated as product documentation or direct observation.

03 / Testing

What we actually observed

A hands-on label requires a documented task, access level, environment, date, observed result, and limitation. A public form test does not establish account-level usability or data accuracy.

04 / Comparisons

How products are compared

Both products use the same criteria, source date, and evidence tier. Source-only comparisons may explain scope and fit, but cannot score performance or assign a winner. Unknown evidence remains unknown.

05 / Commercial

How relationships are disclosed

Affiliate, referral, sponsorship, access, and other material relationships are disclosed near the affected content. Paying us does not improve a product's position.

06 / Corrections

How research changes

Material corrections identify what changed and when. Product facts are rechecked on a dated schedule based on how quickly they can change.

Publication gates

What must be true before a profile or comparison goes live?

  1. 01

    Profile source gate

    Use at least two current primary sources, map each material capability to a source, and state the access we did not have.

  2. 02

    Hands-on gate

    Record one bounded task, environment, access tier, result, duration when relevant, and every limitation. Do not generalize beyond that task.

  3. 03

    Comparison gate

    Use matched criteria and evidence dates. If account access is not equivalent, compare only documented scope and label the page source-checked.

  4. 04

    Recommendation gate

    No rating, performance winner, or value claim without equivalent hands-on trials, repeated observations where outputs vary, and current commercial terms.

Evidence labels

A compact language for research status

Source checked

An official source supports the statement and the check date is visible.

Hands-on test

We describe the tested task, conditions, date, and observed result.

Vendor claim

The statement comes from the vendor and remains clearly attributed.

Editorial assessment

The conclusion is ours and is kept separate from sourced facts.

Not tested

We did not test this capability and will not imply that we did.