Our Desk-Research Self-Tanner Scoring Method

The 9-dimension editorial scorecard

Only products with a safely matched product-level review sample receive a public rank. Each ranked self-tanner receives a 0-10 editorial score on the same dimensions, followed by the published weights below. Every review page names the source basis for each dimension. Unmatched products remain on an unranked evidence watchlist with N/A scores.

Reproducible formula: weighted score out of 100 = the sum of each 0-10 sub-score multiplied by its percentage weight, divided by 10. Tan List retains one decimal place; the weights total 100%.

Score anchors and coverage grades

• 9-10: among the strongest comparative signals in the evidence-qualified index.
• 7-8: competitive evidence with visible tradeoffs.
• 5-6: material limitations relative to the current index.
• N/A: no safe product-level match, so Tan List does not infer an experiential score.

Coverage grades describe corpus depth, not product quality or statistical certainty: B means at least 75 safely matched review URLs; C means 40-74; U means no safe match and therefore no rank. No product receives an A because the capture is rating-stratified, self-selected, and not a controlled or representative study.

Tan List editorial scoring dimensions and weights

Dimension	Weight	Why
Color naturalness	16%	Undertone and depth fit decide whether the result looks plausible on the intended skin-tone range.
Streak resistance	16%	Formula spread, set time, guide color, mitt choice, and technique all affect visible unevenness.
Smell tolerance	13%	DHA-development odor and added fragrance can matter for an hours-long at-home routine.
Transfer resistance	11%	Guide color and incomplete dry-down can mark sheets or clothing during development.
Fade quality	11%	Uneven fading around dry or high-friction areas changes how long the result looks intentional.
Conservative formula screen	9%	Fragrance, alcohol, and botanicals deserve a conservative screen; this dimension is not a safety claim.
Shade range	8%	A broader depth or undertone ladder gives shoppers more ways to avoid a mismatched result.
Retailer coverage	8%	Distribution and return access affect how easily shoppers can verify, buy, or replace a formula.
Value per application	8%	Snapshot price, bottle size, format, and likely routine frequency shape cost; verify live price before buying.

Why these weights

The weights are an editorial prioritization model. Color and streak control receive the most weight; smell, transfer, and fade follow; conservative formula screen, shade range, retailer coverage, and value keep the ranking from reducing every decision to color depth. Face use is excluded until exact label approval is verified.

The review corpus informed which failure modes deserve explicit coverage, but corpus frequency was not converted directly into the weights. The finder reweights the same dimensions for an individual use case instead of pretending one default weighting fits everyone.

Evidence sources and dates

• Source capture. 1,684 rows and 1,175 unique review URLs across 30 Amazon listings, collected May 3, 2026.
• Formula subset. After excluding one standalone applicator-kit listing (58 captured rows; 49 unique URLs), 1,626 captured rows and 1,126 unique formula-review URLs across 29 listings remained; 1,116 were marked verified.
• Sampling design. Of 1,126 deduplicated formula-review URLs, 866 appeared only in rating-filtered helpful slices, 92 appeared only in recent-all or validation slices, and 168 appeared in both. The mix supports issue discovery, not prevalence claims.
• Theme tags. Rule-based lexical matches for ease, streaks, smell, format, skin-tone fit, orange cast, transfer, fade, and related subjects. A mention is not automatically a complaint.
• Product-level samples. Each review page names the matched listing scope and unique-review count, or states when no safe match exists.
• Product and retailer surfaces. Price, rating, review total, shade, size, and availability snapshots captured May 2, 2026. They can change and should be verified before purchase.
• July editorial review. Titles, descriptions, evidence labels, internal links, and decision paths were reviewed July 10, 2026.

Known limitations

• No hands-on test panel. We cannot claim texture, wear time, scent strength, transfer, or fade from direct observation.
• No representative sample. Amazon reviews reflect self-selected reviewers, rating-stratified sampling, and the listings captured.
• Lexical tags are not sentiment. Phrases such as "not orange" can still trigger an orange-language tag.
• Four assessed products are unranked. Products without a safe product-level match publish N/A instead of experiential scores or a total.
• Snapshot fields age. Prices, retailer coverage, ratings, formulas, and review totals need live verification before purchase.

When a score changes

A score changes only after a material evidence update: a formula or shade change, a corrected product match, a refreshed retailer snapshot, a larger validated corpus, or documented original testing. A rebuild alone does not advance publication dates or create new evidence.

Disagree with our weights?

That is expected. The default score is a navigation aid, not a universal truth. Read the sub-scores, use the self-tanner finder, and choose the tradeoff that matches your routine.