Your serum is made, the labels are compliant, and 800 units sit in the spare room. Now the question that decides your margin for the next two years: where do you sell it? Skincare founders in India face four doors, Nykaa, Amazon, Meesho and your own website, and most channel advice comes from people selling phone chargers. Skincare is a repeat-purchase, trust-heavy, positioning-sensitive category, and that changes the answer completely.
This guide puts the 2026 fees of all four channels side by side with skincare numbers, gives you the one rule that resolves the choice, your price point and positioning, and lays out the month-by-month sequence most skincare brands should run, with honest reads on how Minimalist and mCaffeine actually did it. If you are a step earlier, still picking the product and manufacturer, start with how to start a skincare brand in India and come back with a product in hand.
Skincare profit lives in the repeat order, and only your own website lets you own it. A serum empties in 45 to 60 days; on your own site the second bottle arrives at near-zero acquisition cost, while on a marketplace every reorder costs what the first one did. The 2026 fees: Nykaa takes 18 to 22% commission on skincare plus 18% GST on the commission itself, Amazon charges zero referral fee under ₹1,000 but beauty ads run ₹80 to ₹120 an order, Meesho takes 0% commission but its skincare buyer pays ₹99 to ₹299, and your own site keeps full margin while you buy every visitor at ₹150 to ₹250. The rule: under ₹300 mass pricing goes marketplace-first; ₹500+ ingredient-story brands go own-site first, Amazon at month 2 to 3, Nykaa at month 6 to 9. Serious skincare brands end up hybrid. What you are really choosing is the sequence.
Why "where do I sell" is a different question in skincare
A phone-case seller picks whichever channel delivers the cheapest order, because every buyer is a one-time buyer. Skincare is the opposite. A 30ml serum finishes in 45 to 60 days, and a customer who liked it needs another one. The brand that owns that reorder gets it almost free. The brand that does not pays full acquisition cost for its own customer, again and again.
Marketplaces sell your product and keep your customer. Amazon masks the buyer's phone and email. Nykaa's customer belongs to Nykaa. Meesho's buyer often meets your brand only when the parcel arrives. Only your own website hands you the contact that makes a day-45 refill message possible. So the real question is not which channel is cheapest this month. It is who gets paid on the second, third and fourth bottle.
What each channel costs a skincare brand in 2026
Run one product through all four doors and each channel's personality shows up. The table uses a ₹599 ingredient-led serum with ₹165 of product and packaging cost, the standard mid-premium skincare setup.
| Channel | Commission / take | Other real costs per order | ₹599 serum: net per order | Do you get the customer? |
|---|---|---|---|---|
| Nykaa (marketplace model) | 18 to 22% skincare commission + 18% GST on the commission, 21 to 26% effective | Partner fulfilment ~₹60, visibility ads ~₹70, settlement 30 to 45 days | ~₹150 | No |
| Amazon (FBA, beauty) | 0% referral under ₹1,000 since March 2026; closing fee ~₹35 | FBA fulfilment ~₹77, GST on fees ~₹20, beauty ads ₹80 to ₹120, returns ~₹10 | ~₹180 | No, contact details masked |
| Meesho | 0% commission; earns on shipping and ads | COD-heavy orders, RTO on you, buyer expects ₹99 to ₹299 | Not viable at ₹599; built for a sub-₹300 line | No |
| Your own website | No commission; ~₹2,000/month platform + ~2% gateway | Shipping + gateway ~₹86, RTO provision ~₹48, Meta CAC ₹150 to ₹250 | ~₹130 first order, ~₹330 on repeat | Yes: phone, email, full history |
Nykaa: the skincare-native channel you earn into
Nykaa is where Indian beauty buyers already shop with intent, which is exactly why you cannot just sign up. It is approval-based, and the category team filters out most small applicants. Skincare commission runs 18 to 22% by sub-category, with 18% GST charged on the commission itself, so the effective cut is 21 to 26% before visibility ads, and premium arrangements push toward 25% commission. Settlement lands 30 to 45 days after delivery, longer once skincare returns are processed. One more layer: beauty mostly runs on Nykaa's inventory model, where Nykaa buys your stock at a negotiated wholesale margin, so you model economics on the price Nykaa pays you, not your MRP.
The entry reality for a small brand: a filed trademark, clean Legal Metrology labels, a real brand story and margin room to share. Apply in month 6 to 9 with proof, not in launch week. The onboarding steps and both selling models are in how to sell on Nykaa.
Amazon beauty: cheap fees, expensive attention
The fee side turned friendly. Since March 2026, Amazon India charges zero referral fee on products under ₹1,000 across 1,800+ categories, beauty included. Your ₹599 serum pays no commission, just a closing fee near ₹35 plus fulfilment.
What remains is the expensive part. The review moat: results for "niacinamide serum" are held by brands with thousands of reviews each, and your zero-review listing starts beneath all of them. Ad dependence: beauty keywords are among the most contested on the platform, so budget ₹80 to ₹120 of ad cost per order for your first two quarters. Price pressure: deal events and ₹299 lookalikes grind at your MRP, and every discount becomes public history. Amazon is a strong second channel for skincare and a risky home base. The fee-by-fee arithmetic across categories is in Amazon vs Shopify in India.
Meesho: built for mass price points, wrong for positioning
Meesho's 0% commission is real; the platform earns on shipping and ads instead. But the buyer is price-first, COD-heavy, largely tier 2 and 3, and shops skincare at ₹99 to ₹299. A face wash, a soap range or a ₹199 cream can move serious volume here. A ₹599 ingredient serum sits beside ₹149 lookalikes and loses on the only axis this buyer compares: price. And marketplace discounting trains the market on what your label is worth. If you want Meesho volume, run a separate mass-price line and keep the premium label off it. Registration, fees and RTO detail are in how to sell on Meesho.
Your own website: the margin keeper that bills you for every visitor
Your store keeps the full ₹599 minus about ₹2,000 a month of platform cost and roughly 2% in gateway fees. You get bundles, subscriptions, the customer's phone number and the refill flows this category runs on. The bill arrives as traffic: a new skincare brand buys Meta visitors at a ₹150 to ₹250 CAC, and at ₹170 the serum nets about ₹130 on a first order. Thin next to Amazon's ₹180, until the second bottle enters the math.
The repeat-order math that settles the argument
Margin Waterfall™: selling price minus COGS, packaging, fulfilment, platform fees, RTO loss, then CAC or ad cost, run separately for every channel you are weighing. In skincare, run it twice per channel: once for the first order, once for the repeat order at that channel's real reorder cost. The first number shows where validation is cheap. The second shows where the business is.
Here is the same serum, first order against reorder.
| Order | Your own website | Amazon |
|---|---|---|
| First order, ₹599 serum | ~₹130 after a ₹170 CAC | ~₹180 after ads |
| Repeat order, day 45 to 60 | ~₹330 · CAC near zero, prepaid, negligible RTO | ~₹180 · the reorder is another cold search with rivals' ads above you |
| Customer record | Phone, email, order history: yours | Masked by the platform |
The own-site repeat arrives from a WhatsApp nudge at day 40, pays prepaid, almost never RTOs, and costs nearly nothing to win back: about ₹330 stays with you. On Amazon the reorder is another cold search under rivals' sponsored ads, at first-order economics, still with no customer record. At a 25% repeat rate, own-site profit per customer nearly doubles; marketplace profit per customer stays flat. In a category that empties every two months, that spread is the business. The machinery behind it, repeat-rate math, refill flows, LTV to CAC, lives in customer retention for D2C and WhatsApp marketing for D2C.
The decision rule: price band and positioning pick your first channel
The channel follows your AOV and your positioning claim, not the success story you read last week. A ₹249 face wash cannot fund a ₹170 Meta CAC; that is 68% of the price gone before product cost. A ₹699 barrier serum has no buyer on a platform where skincare tops out at ₹299.
If your skincare sells under ₹300 and competes on price (face wash, soaps, basic creams) → marketplace-first: Meesho for volume, Amazon for search, brand-building later. If you sit at ₹300 to ₹500 → Amazon-first inside the zero-referral window, own site live the same month to start collecting repeat buyers. If you sell at ₹500+ on an ingredient or audience story → own site first, Amazon at month 2 to 3 to catch search demand, Nykaa application at month 6 to 9 with trademark and traction in hand. If your contribution margin cannot survive a 26 to 30% all-in marketplace cut → stay own-site only until price or COGS fixes it, because a channel that eats the whole margin is not distribution, it is liquidation.
The hybrid sequence most skincare brands should run
According to the Founder Decision Loop™, channel expansion follows proof, not ambition: each channel gets added when the previous one produces the data that justifies it. For a ₹500+ skincare brand, the sequence looks like this.
- Month 0 to 1: own store live, WhatsApp opt-in on the order confirmation, prepaid share pushed past 60%. Every order builds the refill list.
- Month 2 to 3: Amazon listing aimed at one narrow search term you can actually win. A package insert in every marketplace order gives the buyer a reason to reorder on your site: bundle price, free shipping, a routine guide.
- Month 4 to 6: read the numbers. Scale the channel with the better Margin Waterfall™ result, and kill hero-SKU discounts on marketplaces. Meesho enters only if you launch a separate sub-₹300 line or need to clear stock, written down as exactly that.
- Month 6 to 9: Nykaa application with trademark, compliant labels and live sales proof; clean files typically clear review in about two weeks.
- Month 9 to 12: quick commerce and offline conversations once repeat rate crosses 20%, because shelf presence multiplies a brand that already retains.
The public examples map onto this. Minimalist launched in October 2020 on its own website and reached ₹100 crore in revenue within eight months, adding Amazon and Nykaa within three months of launch: the own-site start set its full-price, ingredient-first positioning, and marketplaces then multiplied reach it had already earned. mCaffeine went wide early, own site, Amazon, Flipkart, then offline shelves, and grew 51.8% to ₹205 crore in FY23, though growth stalled the following year. Useful reminder: being everywhere reduces dependency, but it does not by itself protect margin or momentum. Neither brand started on a marketplace. Both started where the customer data lands.
Listing the hero serum everywhere in launch week. A founder with 1,000 units puts the ₹599 serum on Amazon, Meesho and the website at once, panics at slow sales, and runs a ₹299 marketplace deal. It moves 300 units and resets the product's public price: own-site visitors now wait for the discount, the listing anchors to ₹299, and when a Nykaa category manager checks price history months later, the wholesale negotiation starts from the lowest number ever printed, not the MRP. The ₹90,000 of deal revenue cost about ₹300 of pricing power on each of the remaining 700 units, roughly ₹2.1 lakh, plus the premium story the brand was built to carry. Sequence your channels; never let a panic discount set your price.
As Ravikant Tyagi, who ran distribution at Eureka Forbes and led supply chain and operations at Atomberg before working with D2C founders as a fractional COO, puts it:
I've sat in a review where a skincare brand celebrated crossing 1,000 marketplace orders a month while its repeat rate sat at 4%. In a category where the product empties every 60 days, they were paying full acquisition cost a thousand times a month for customers they had already won once. Before any channel debate, I make founders write down one number per channel: profit on the second bottle. The channel that owns the reorder funds the brand. The rest rent you reach, and rent never builds equity.
Execution checklist
- Write one line: your AOV band, your positioning claim, and the first channel the decision rule assigns you.
- Run the Margin Waterfall™ for your hero SKU on every channel, first order and repeat order separately.
- Hold one public price everywhere; run offers as own-site bundles, never as MRP cuts on marketplaces.
- Own store: WhatsApp opt-in live and a day-40 refill flow built before you scale ad spend.
- Amazon: confirm your product sits under ₹1,000 in an eligible beauty category on the live rate card, and pick one winnable keyword.
- Print package inserts that give marketplace buyers a concrete reason to reorder on your site.
- Start the Nykaa file in month 4: trademark status, Legal Metrology labels, manufacturer license copy, margin sheet.
- Treat Meesho as a separate decision for a sub-₹300 line or stock clearance, never for the premium label.
- Plan cash flow around marketplace settlement gaps, 30 to 45 days on Nykaa, before committing the next production run.
Your next action
Open one sheet today. Four columns: Nykaa, Amazon, Meesho, own site. Two rows: what your hero SKU nets on a first order, and what it nets on a reorder. Fill it with your real COGS, your real price and the fee numbers above. The column with the strongest second row is your home base, and the decision rule gives you the order in which the others join. Thirty minutes of arithmetic ends the channel debate.
If you'd like the complete execution system, calculators, SOPs, templates and operating frameworks behind this process, continue inside D2C Acquisition.Lab.
