Review Methodology

We distinguish testing from research

A product article should not claim hands-on testing unless the product was actually used or measured by the author or editorial team. Research-only comparisons should be labeled and written accordingly.

What hands-on testing should document

When we perform a hands-on review, the methodology note should identify the meaningful conditions of the test. Depending on the product, that can include device model and firmware, test duration, network environment, connected ecosystem, measurement tools, energy-monitoring setup, sample size, or any important limitation.

Core evaluation dimensions

• Setup and compatibility: installation requirements, wiring, hubs, bridges, operating systems, protocols, and ecosystem support.
• Reliability: routine performance, recovery from outages or restarts, connectivity, latency, and automation consistency.
• Local versus cloud operation: which functions work locally and what changes when an internet or vendor-cloud connection is unavailable.
• Privacy and account requirements: data flows that are visible from product documentation and actual setup behavior.
• Energy performance: relevant consumption, standby load, efficiency, or monitoring accuracy when the product category makes those measurements meaningful.
• Usability: app quality, controls, accessibility, household sharing, notifications, and manual fallback.
• Long-term value: software support, interoperability, replacement costs, subscriptions, repairability, and ecosystem lock-in.

Comparisons and recommendations

When an article recommends a product as “best” for a use case, it should state the decision criteria and explain why that product wins for that scenario. Strong alternatives and meaningful drawbacks should also be presented.

Manufacturer samples and sponsored access

If a manufacturer supplies a review unit, travel, early access, or another material benefit, that relationship should be disclosed. Receiving access does not guarantee coverage, a positive rating, or editorial approval.

Measurements and evidence

Original photographs, screenshots, measurements, logs, and test data are preferred when they materially demonstrate first-hand experience. Quantitative results should include enough methodology to help readers understand what was measured and what the numbers do—and do not—prove.

Corrections after publication

If subsequent testing reveals that a published measurement or conclusion is materially wrong, we update the article and follow our Corrections Policy.