01 · Capture the claim exactly as the creator states it.
Paraphrase drifts; quotes and captures do not. The article receipt pins what was promised and when, so later spec changes are visible instead of quietly absorbed.
Methodology · version 1.1
Campaign evidence before enthusiasm
Version 1.1 took effect September 22, 2026. This transparency revision adds the two-score policy; it is not a silent rewrite. It applies to every Crowdfunding review, guide and comparison.
02 · Separate prototype footage from delivery evidence.
A working prototype on camera is not a shipped product. Renders, animations and “expected” specifications are labeled as such, and continuous footage of the core function counts for more than fast cuts.
03 · Check the creator record, shipping plan and refund terms.
Past campaigns are the best predictor available: delivered late, delivered never, or delivered well. Refund and cancellation terms are read before the pledge button is ever discussed.
04 · Publish “Campaign — not yet shipping, not tested” until that state changes.
The label stays until units reach backers and independent evidence exists. Enthusiasm about a concept never upgrades into a verdict about a product.
What we check in crowdfunding
Prototype proof: continuous footage of the claimed function in one shot; we ask whether the object shown is the object being sold.
Manufacturing plan: named processes, suppliers, tooling status and certifications — “production ready” without detail is marketing copy.
Shipping math: does the pledge price plausibly cover parts, packaging, platform fees, freight, taxes and support — or does the math only work if nothing goes wrong?
Creator history: previous campaigns, communication quality during delays, and how past backers were treated when things slipped.
Spec-change tracking: comparing the current campaign page against its updates; unexplained specification changes are warnings, not footnotes.
Timeline realism: tooling, certification and freight take the time they take; magical timelines get flagged, not repeated.
Scoring gate
Hands-on lab-test score: Awarded only once the editorial team has physically tested the product and published the measurements behind the score.
Research-based editorial score: Editorial judgment built from published review consensus, specification analysis and third-party hands-on reports — never presented as a lab result. It tracks the article’s ranking and its cited pros, cons and verdict; products with no honest differentiating basis are marked Not scored.
Why versioned and dated
Stale methodology is worse than none — it lends authority to practices nobody follows anymore. That is why every page here carries a version number, an effective date and a review schedule, instead of quietly drifting out of date.
Our commitment is evidence honesty: every article shows the sources it reviewed and labels what came from the manufacturer. A methodology earns trust by revealing its workings, not by implying lab access or measurements that never happened.