About & Methodology

What EvidencePick is

EvidencePick is a new publication produced by a small independent team. It is not a store, and it does not sell products. Our work is analytical: we take a buying decision, gather the available documentation and market information, and explain what the evidence supports.

Our current editorial focus is portable power and home-backup equipment. The publication is intentionally broader than any single category, and coverage will extend to other emerging consumer technology and gear as research is completed.

Desk research versus hands-on testing

These are two different things, and we keep them clearly separated.

• Desk research means analysis of documentation, published specifications, standards information, market data and other material we can read and cite. It can establish what a product claims, how those claims compare, and whether a configuration is compatible or plausible. It cannot establish real-world performance.
• Hands-on testing means we physically used or measured a unit ourselves. When we do this, the article says what we tested, how, and for how long.

When a product has not been physically tested by us, the article will say so clearly and in plain language. We do not describe unverified figures as measured results, and we do not imply hands-on experience we do not have.

Sources we use

• Manufacturer specifications, manuals and technical documentation
• Regulatory, certification or standards information where relevant to the category
• Retailer data such as configurations, availability and listed pricing
• Public technical resources, datasheets and component documentation
• Direct testing and measurement when a unit is available to us

Where a claim depends on a specific source, articles link to it so readers can check the original.

How we form conclusions

We normalise claims into comparable figures where possible — units, capacities, ratings, duty cycles — and note where a manufacturer measures something differently or does not publish it at all. Missing or ambiguous data is treated as missing data, not as a defect and not as a strength. If the evidence is insufficient to recommend anything, we say that instead of picking a winner.

Conclusions can change

Recommendations may change when products are revised, pricing shifts, firmware changes, new evidence becomes available, or market conditions change. Articles carry a publish date and, where applicable, an updated date, so readers can judge how current the analysis is.

Commercial relationships

EvidencePick uses affiliate links and may earn a commission when readers purchase through qualifying links, at no additional cost to the reader. Those relationships fund the research, but they do not determine our conclusions: no partner reviews or approves our analysis before publication, participation in a program does not guarantee coverage, and a product being commissionable is never a reason to rate it more favourably. Full details are on our affiliate disclosure page.

Corrections

If you believe something we published is inaccurate, tell us and we will check it. Material corrections are noted in the article. You can reach us through our contact page.