The 5 factors, explained
Every Chayan Score is the weighted sum of 5 publicly-verifiable signals. We don't physically test products — we score them using cross-marketplace data and category-specific feature rubrics. Here's what each factor measures and where the data comes from.
1. value_for_money
Weight 30%What it measures: How much you get per rupee compared to direct alternatives in the same category and price band.
How we calculate it: We pull the top-10 products by review count in the same category × ₹500 price band. We compute the median feature score (see factor 3) per rupee. The candidate's value-for-money score = its (feature_score ÷ price) ÷ band_median.
Data sources: Amazon listing prices (manual + EarnKaro feed), Flipkart Affiliate API, Croma + Reliance Digital scraped weekly via legal partner sources.
2. user_reviews_signal
Weight 25%What it measures: Cross-marketplace verified-purchase review aggregate. We don't trust a single source.
How we calculate it: Weighted average of 4 sub-metrics across Amazon, Flipkart, Croma, Reliance:
- Average rating (40% of this factor) — across all platforms where review count > 100
- Review velocity (25%) — recent 30 days vs all-time average. Rising = positive signal.
- Verified-purchase ratio (20%) — only Amazon publishes this; we extrapolate
- Review count threshold (15%) — sigmoid scaled to discount low-count flukes
Anti-manipulation: We discount review spikes around new-launch promotional periods.
3. feature_completeness
Weight 20%What it measures: How many category-relevant must-have + nice-to-have features the product has.
How we calculate it: Every category has a public feature schema (e.g., for wireless earbuds under ₹2000: ANC, IPX rating, battery life ≥30hr, BT 5.0+, gaming mode, mic noise reduction, custom EQ, fast charging). Score = weighted sum of features present.
Where to inspect: Every Best Under X article links to its category feature schema at the bottom. We publish the schema so you can challenge it.
4. popularity_trend
Weight 15%What it measures: Bestseller Rank movement over the last 30 days. A product climbing is a meaningful signal.
How we calculate it: Logarithm of BSR change, clamped between -1 and +1. A product that went from BSR 200 to BSR 50 gets the same boost as 5,000 to 1,000.
Why this matters: Catches products gaining real momentum (vs paid launch spikes which decay fast and we filter out).
5. reliability_record
Weight 10%What it measures: Brand-category-level signal of reliability based on past product performance.
How we calculate it: Brand-category track record. Boldfit in fitness has a different reliability score than Boldfit in supplements. Computed from prior products' user ratings + warranty fulfillment complaints scraped from public review threads.
Caveat: This is the most subjective factor. Lowest weight intentionally.
Score tiers — what the number means
Every product gets a 0-100 Chayan Score. Three tiers determine where it shows up on the site:
Pushed everywhere. Shows in articles (TL;DR top picks), homepage spotlight, category top results, comparison defaults. The product passed all five sub-scores at acceptable level + at least one at strong level.
Listed but not pushed. Visible in category browsing and comparison if you specifically pick it, but doesn't appear in articles, homepage spotlight, or "best of" picks. These products are mediocre — not bad enough to reject, not strong enough to recommend.
On the rejection list. Failed multiple sub-scores. We publish these openly with the specific reason — value, reliability, after-sales, or feature gap — so you can see exactly why we passed.
Why the rejection threshold is 50, not 65: at 65 the cutoff catches half of any real-world catalog — most products are simply mediocre, not actively bad. Calling that many "rejected" devalues the signal. At 50, only products that fail 2+ sub-scores show up — when we say "we wouldn't buy this," we mean it.
Editorial boosts TRANSPARENT
We believe a small number of categories deliver value that the universal 5-factor formula systematically under-weights. Where that's the case, we apply a fixed editorial boost to the Chayan Score — and we disclose it on every affected product card as a labelled pill (e.g. + 9 Spiritual). Hover the pill on any card to see the rationale inline.
Total boosted score is always capped at 100. We expose the pre-boost score in the product detail page so you can audit how much of the final number is base vs editorial.
What's NOT a boost: We do not apply boosts for brand, region, price tier, sponsorship, advertising spend, or anything a merchant could buy. Only category-level convictions we'd defend in print. Every boost added is logged in the changelog below.
Where the data comes from
How we don't cheat
- Zero sponsored placements. We do not accept money to include, rank higher, or remove a product. Ever. Full disclosure.
- Authors named on every article. If you disagree with a ranking, you can email the author directly.
- Public rejection list. We publish products we considered and refused to recommend, with specific reasons. See it →
- Methodology changelog. Every weight adjustment is logged with date and reason (below).
- No "best of" categories you've never heard of. We don't invent categories to inflate winner counts.
- Affiliate links are disclosed, not hidden. Every product link is marked as an affiliate link with the merchant name visible.
Calibration schedule + changelog
Weights are reviewed every quarter. When we change one, we publish it here. If we drift, you should be able to see exactly when and why.
FAQ
Why don't you physically test products?
Physical testing requires buying every product (capex ~₹3 lakh/quarter) + storage + a lab + writers who can compare. We chose a different bet: cross-marketplace verified data + category schemas + radical transparency. Wirecutter physically tests; we methodologically score. Different model, equally defensible — as long as we're honest about it. We are.
How do you handle paid affiliate relationships?
Every product link is an affiliate link. We earn a commission when you buy through it. The commission rate does NOT affect the Chayan Score. If we recommended only the highest-paying products, you'd notice — and so would Google. Our reputation is the asset; the affiliate commission is the rev-share. Full disclosure on the Disclosure page.
What's the difference between Chayan Score and Amazon's star ratings?
Amazon stars measure user satisfaction at one marketplace. Chayan Score measures relative value, feature completeness, momentum, and reliability across four marketplaces. A product can have 4.5 stars on Amazon and a Chayan Score of 62 if it's overpriced for its features compared to alternatives in the same band.
Can I challenge a score?
Yes. Email the author named on the article. If your challenge identifies a data error or factor we missed, we update the score (and credit you in the changelog if you want). We have already adjusted 14 scores from reader feedback in the past year.
How often are scores updated?
Prices update weekly. Reviews update weekly. BSR updates daily. Full score recomputed weekly (every Monday). Major calibration quarterly.
What if a product hits an 80+ score but you still don't recommend it?
A high Chayan Score makes a product eligible for our recommendation list — it does not guarantee inclusion. We also look at the rejection list considerations: reliability complaints, after-sales-service track record, return rate signals from public threads. Some 85+ scoring products are on the rejection list. Read why →