Methodology

Evidence before enthusiasm.

Every label controls what the site can claim, what proof must be shown, and whether a retailer action is allowed.

01

Tested

We installed or used the exact model ourselves and recorded the date, setup conditions, installation result, observed limitation, and specific skip guidance. A date plus a short generic note does not qualify.

02

Researched

We checked a named model against linked manufacturer or manual evidence and a direct Amazon.ca product-detail destination. Each scorecard shows what each dated source supports. We do not claim firsthand performance.

03

Excluded

The product changed, became unavailable, failed a check, lacks a verified retailer detail page, or no longer deserves an active recommendation. The record stays visible with the exact reason and no purchase button.

Maintenance

Update rhythm

During validation, availability and merchant quality are checked every 30 days. Evidence and recommendation language are reviewed every 90 days.

Boundary

AI use

AI can organize research and help draft scripts. It cannot become a product tester, testimonial, or substitute for firsthand evidence.

Accountability

Corrections

Report a correctionthrough the channel maintained by the site owner.