CompactSetup Lab compares products for a specific physical constraint rather than trying to name a universal “best.” A recommendation must first pass an evidence gate. We look for an exact model identity, a verifiable product page, at least two distinct sources, at least one primary source, and enough dimensional information to judge the stated use case.
Eligibility comes before scoring
A high score cannot rescue weak evidence. A product is held out of a fit-based comparison when its footprint is unknown, its specifications conflict across sources, it cannot meet the desk or rental constraint, or its source set is too thin. Unknown information is shown as unknown rather than estimated.
Scoring weights
- Physical fit: 30%
- Evidence quality: 22%
- Match to the reader’s intent: 16%
- Cable cleanliness: 10%
- Portability: 8%
- Renter friendliness: 8%
- Setup ease: 6%
Physical measurements and source quality dominate the score. Analytical scores such as cable cleanliness are labeled as editorial assessments, not manufacturer specifications.
AI assistance and quality control
AI helps find, structure, and explain public information. Deterministic code calculates fit and ranking. A separate quality pass checks the draft against the source-backed product record. Unsupported claims, hands-on language, fixed Amazon prices, missing disclosures, and serious conflicts block publication. Passing content can still contain errors, so every product section links to its sources and shows the access date.
Commercial relationships
Affiliate availability never makes a product eligible and is not part of the ranking score. Commercial links are added only after ranking, from an approved link registry.