What the Indian Serum Shelf Is Actually Worth

The shelf is not crowded. It is concentrated.

Thirteen brands and 637 products looks like a fight. It is not. Dot & Key and Chemist at Play together take 59% of the money on this shelf. Add Deconstruct and it is 74%. The remaining ten brands share ₹41.5 crore between them — and three of those ten are turning over under ₹2 crore a year on this channel.

That is the gap between a competitor count and a revenue distribution. Counting brands says thirteen. Counting money says three. If you are weighing entry into a category because it looks lightly populated, the brand count is the wrong number to be looking at.

The leaderboard

GMV is derived: Amazon's published units-per-month multiplied by current selling price, annualised. It is a floor, and the method section below explains why it understates in three separate directions.

Revenue per product

Treat this column with more caution than the others. It is the ratio of two measured numbers, but it moves with how completely a catalogue has been crawled — an earlier pass that read only each brand's storefront front page put the leader at ₹2.50 crore per SKU, three times what the full catalogue shows. These figures come from every product page, but the metric remains the softest one here.

Range is not a moat — and not a trap either

Fixderma lists 85 products and earns ₹10.8 crore. Deconstruct lists 35 and earns ₹23.9 crore. Two and a half times the catalogue, under half the revenue. On this shelf, revenue per product spans ₹0.78 crore at the top down to ₹0.01 crore at the bottom — a spread that catalogue size does not explain.

Note what that does not say. The three biggest brands here carry 65, 35 and 115 products. They are not lean-catalogue operators, and Chemist at Play has the widest range on the shelf while sitting second by revenue. Adding SKUs is neither the cause of winning nor a substitute for it. The brands in difficulty are the ones carrying range they have not made pay.

This correction matters because we nearly published the opposite. An earlier pass read only the front page of each brand's Amazon store — roughly the top ten products each — and on that data the leader looked thirteen times more productive per SKU than the widest-range brand. Crawling every product page instead of the storefront halved the gap and reversed the conclusion. The first measurement was capturing how much of each catalogue we had looked at, not how well each brand was selling.

Discount posture splits the shelf

Because we hold MRP alongside the selling price on every product, pricing strategy is directly visible. We rank on the median discount rather than the deepest cut: a single clearance SKU sets a maximum and tells you nothing about how a brand normally prices. Aqualogica touches 85% once across 40 discounted products; its typical product sits at 21%.

On that measure the shelf runs from 9% to 42%. Deconstruct's typical product carries a 9% discount; Conscious Chemist's carries 42%. A brand whose median product sits above 40% is not running promotions — that is its real price, and the MRP exists to make it look like a saving. Both are legitimate strategies. Only one survives a competitor willing to do it harder.

Reviews describe the past. Units describe the present.

Review counts are cumulative — they measure everything a brand has ever sold. Units-per-month measures what is moving now. The ratio between them is the most useful single number for reading a competitor, and it is one nobody publishes.

Dr. Sheth's sits at 4.5 reviews per current unit — indē wild at 0.3. The first describes a brand that has sold a great deal historically and is moving slowly now; the second, a brand whose sales are running well ahead of its accumulated reputation. Both are legitimate positions, and a competitor count cannot tell them apart. Neither can a review count on its own, which is why the leaderboards you usually see are misleading.

Review counts here are de-duplicated — Amazon repeats one parent product's figure across every size variant, and summing raw would have inflated this shelf by 1.8×.

Method, and what this is not

How it was measured. Price, MRP, discount, price per 100ml, star rating, review count, stock status and units-per-month were read from public Amazon product listings on 11–13 August 2026 — rendered page text, no estimation. Units-per-month is Amazon's own "bought in past month" figure.

GMV is derived, and it is a floor. Units × current price × 12. Amazon publishes velocity in buckets, so "10K+ bought in past month" is counted as exactly 10,000. Products with no published velocity contribute zero. Real figures are higher, not lower.

It is GMV, not revenue. This is what buyers pay on Amazon, before Amazon's commission, fulfilment and returns. No brand receives this number.

Amazon only. Several of these brands sell more through their own websites, Nykaa, Myntra and quick commerce. None of that is counted here.

Review counts are de-duplicated. Amazon reports one figure per parent product and repeats it on every size variant — one brand here showed the same 15,430 review count across six separate listings. Summing raw would have inflated this shelf by 1.8×. Each distinct value is counted once per brand.

Corrected on 13 August, and the correction was ours. This post first published at ₹147.5 crore with the top three at 78%. Both were wrong. Amazon renders its "bought in past month" figure on some brands' store pages and not others, and we had read store pages only — so we were measuring our own crawl surface rather than the market, and doing it unevenly: 96% coverage on the shelf leader, 14% on SkinQ, 0% on Mintree. We were understating hardest for the brands that already looked smallest, which flatters every concentration figure. Re-reading the missing velocity from search pages moved the shelf to ₹159.0 crore and the top three to 73.9%. Dr. Sheth's rose 129% and Fixderma 74%, while the leader moved 0% — which is the check that this was a fix and not a fresh error, since a uniform rise everywhere would have meant double-counting.

One shelf, one reading. Every figure above is reproducible from public product pages. If you believe one is wrong, we would genuinely like to know which.