Tell readers what we actually did

Each review identifies its evidence basis: hands-on tested, research-based comparison, or first-look analysis. A desk-researched article must never imply that a product was purchased, handled or independently tested.

Scope and selection

State the intended user, market, budget, product version and inclusion criteria. Document why alternatives were included or excluded. Obtain products independently where possible and disclose loans, gifts, travel and manufacturer access.

Evaluation criteria

Products: task performance, usability, build quality, accessibility, compatibility and total cost. Services: coverage, terms, onboarding, support and cancellation. Software: task accuracy, privacy controls, accessibility, export options and limitations. Select criteria relevant to the category before testing.

Testing process

Write a repeatable protocol with tasks, equipment, conditions, dates, versions and sample size. Run comparable tasks under consistent conditions, record failures and retain original notes or measurements. Separate measured results from impressions. Publish enough detail for the reader to understand the limits; do not generalize one sample to every unit.

Research-only comparisons

Use current provider documentation and identifiable independent evidence. Record prices with currency, market and check date. Label unverified claims as provider claims. Explain missing data and avoid performance rankings where comparable evidence is unavailable.

Ratings

Use a rating only when evidence supports it. Proposed five-point scale: 1 poor, 2 below expectations, 3 adequate, 4 strong, 5 excellent for the stated use. Default weighting: performance 35%, usability and accessibility 25%, reliability and support 20%, value 20%. Calculate the weighted mean to one decimal. Disclose any category-specific changes before testing; do not turn unknown results into zero scores.

Transparency and updates

Show who evaluated the item, what was tested, when, and what could not be assessed. Disclose commercial relationships beside the verdict. Recheck after material updates; archive superseded findings. Our four launch drafts are educational guides, not hands-on product reviews, and carry no product ratings.