How Rankings Work
Dewcode ranks products purely on community experience — not brand deals, gifted products, or paid placements. Here's exactly what goes into every score.
The Non-Negotiable Promise
✓ A product's rank cannot be bought, boosted, or suppressed
✓ Sponsored content lives in clearly labelled separate zones only
✓ No review is published without moderation
✓ User data is never sold to brands
✓ The same formula applies to every product, always
How a Score is Calculated
Every review scores a product across five dimensions. These are combined into a single score, then adjusted for review quality and quantity.
Step 1 — Weighted base score per review
Higher weight = matters more to real-world results
Effectiveness - 30%
Value for Money - 25%
Repurchase Intent - 20%
Texture - 15%
Packaging - 10%
Effectiveness carries the most weight because it answers the only question that matters: did it actually work? Packaging matters least — a product shouldn't rank lower just because the pump is inconvenient.
Step 2 — Trust signals adjust each review's weight
Recency
Recent reviews count more than old ones. A sunscreen reviewed last month reflects current formulation — one from three years ago may not.
Verified purchase
Reviews backed by a purchase receipt carry extra weight. Harder to fabricate, more likely to reflect genuine experience.
Reviewer credibility
Reviewers whose past reviews have been consistently marked "helpful" by the community carry more influence.
Step 3 — Bayesian confidence adjustment
A product with 1 five-star review should not outrank a product with 200 reviews averaging 4.7. Products with fewer than 10 reviews are gradually pulled toward the global average (3.0 / 5). As reviews accumulate, the score shifts to reflect actual community consensus. At 10+ reviews, a product's score stands entirely on its own.
A Worked Example
Let's say Minimalist SPF 50 Sunscreen has received two reviews this week:
Review A — Priya, oily skin, posted 2 weeks ago
5Value
4Repurchase
5Texture
4Packaging
3
Base score: (5×30%) + (4×25%) + (5×20%) + (4×15%) + (3×10%) = 4.40
Boosts applied: recent review + verified purchase + high credibility reviewer
→ This review counts with high weight
Review B — Anjali, dry skin, posted 5 months ago
3Value
3Repurchase
3Texture
4Packaging
4
Base score: (3×30%) + (3×25%) + (3×20%) + (4×15%) + (4×10%) = 3.25
Boosts applied: none (old review, unverified, newer reviewer)
→ This review counts with standard weight
Result after averaging both reviews
Raw average leans toward ~4.1(Priya's review carries more weight).
Only 2 reviews → Bayesian adjustment pulls the score toward the global average.
Published score: ~3.6 / 5 — honest, but will improve as more reviews come in.
How Filters Work
The global ranking is computed once per week across all reviewers. Filters don't recompute rankings — they narrow which products you see.
Skin type filter
Each product accumulates separate score data per skin type from reviewers who state their skin type. Selecting "Oily" surfaces products that performed well specifically among oily-skin reviewers. If a product doesn't have enough oily-skin reviews yet, it falls back to overall brand suitability claims.
Price bracket filter
Filters products by their listed price range. Helpful when you want the best product within a specific budget, not the best product overall.
Climate zone filter
As Dewcode grows, climate-specific performance data will be surfaced from reviewers who mention their city and climate. Currently in early data collection — more reviews improve this signal over time.
Update Cadence
Rankings are recalculated every Monday at 6:00 AM IST — never in real-time. This is intentional: it creates a weekly event, builds anticipation, and prevents rank positions from flickering every time a single review comes in. A product's rank reflects a full week of community sentiment, not a single moment.