Your first jewellery batch is in. Maybe 200 pairs of anti-tarnish studs from a Rajkot job-worker, a tray of oxidised pieces off Sadar Bazaar, or a small run of 925 silver. Now the question that sets your margin for the next two years: where do you actually sell them? Four real doors. Amazon and Flipkart for search demand, Myntra and Nykaa Fashion for a curated fashion audience, Meesho for cheap volume, and your own website for brand, margin and the repeat customer.
Short version, then the math. Your metal lane picks your channel, not your budget. Imitation and anti-tarnish under ₹1,000 clear fastest on marketplaces, because Amazon now charges zero referral fee below that price and the search demand already exists. Real 925 silver belongs on Instagram and your own store: it crosses the ₹1,000 fee line, needs a trust story a marketplace strips away, and sells as a gift. Build the owned base either way. And unlearn one thing from apparel. Returns are not the referee here. Nobody brackets three sizes of a necklace, and earrings are usually sold non-returnable on hygiene grounds. The deciders are the ₹1,000 fee cliff, plating quality that keeps reviews alive, and who owns the next order. Still picking your lane and supplier? Start with how to start a jewellery business in India and come back with stock in hand.
The ₹1,000 mark is the hinge. Since March 2026 Amazon charges zero referral fee on fashion jewellery under ₹1,000, so a sub-₹1,000 piece pays roughly ₹100 of closing and weight-handling fees plus 18% GST. That makes a marketplace a real first-sales engine here, unlike apparel. Flipkart runs heavier, up to about 25% by sub-category. Nykaa Fashion takes 20 to 30% plus GST and penalises low-value orders. Myntra stacks a category commission with a Growth Enablement Fee, but its Rising Stars programme now gives new homegrown D2C brands 0%. Meesho charges nothing and resets your price to ₹199 to ₹449, so it is a clearance lane, not a brand. Your own store keeps the full price minus about 2% gateway and a ₹120 to ₹180 Meta CAC, and it is the only place you own the phone number and the reorder in a gifting category. GST is 3% on jewellery, imitation and silver alike. The rule: under ₹1,000 go marketplace-first for discovery, 925 silver goes own-site and Instagram first, everyone builds owned from day one.
Why the channel call is different in jewellery
A phone-case seller picks whichever channel delivers the cheapest single order, because there is no second order and nothing comes back. Jewellery breaks both assumptions in your favour. It is a gifting and new-drop category, so a buyer who buys once buys again at Rakhi, at Valentine's, through wedding season, if you own her contact. Returns are milder than fashion too: no size bracketing on a pendant, and earrings routinely sold non-returnable, so the parcel that ships mostly stays sold.
Then it adds a fork nothing else has, and that fork should shape your price tags and your channel plan together. Below ₹1,000 Amazon's referral fee is zero and the marketplace is close to free discovery. At ₹1,001 it switches back on and stacks with closing fees, weight handling and 18% GST, so a ₹1,299 silver piece suddenly hands Amazon a real cut. That is why the lanes split cleanly. Imitation at ₹299 to ₹899 and anti-tarnish demi-fine at ₹499 to ₹1,499 sit under or near the line and are marketplace-friendly. Real 925 silver at ₹999 to ₹2,999 sits over it, pays full freight, and needs a trust story a product grid flattens. Its hallmark is an asset you choose to build, not one the platform hands you: silver hallmarking is still voluntary in 2026 while BIS studies making it mandatory, with a HUID required only on the pieces you do hallmark. Silver belongs where that story survives.
What each channel costs a jewellery brand in 2026
One anti-tarnish necklace at ₹899, with about ₹225 of product and gift packaging behind it, run through every door.
| Channel | Commission / take | Other real costs per ₹899 piece | Own the customer? |
|---|---|---|---|
| Your own store (Shopify + Razorpay + Shiprocket) | None; ~₹2,000/month platform + ~2% gateway (UPI 0%) | Shiprocket forward ~₹65 lowest slab, Meta CAC ₹120 to 180 | Yes: phone, email, gifting history |
| Amazon | 0% referral under ₹1,000 (since Mar 2026); resumes above ₹1,000 | ~₹100 closing + weight handling, + 18% GST on fees, ads ₹80 to 140 | No, buyer stays masked |
| Flipkart | Up to ~25% on fashion jewellery by sub-category; lighter in the low band, verify | Fixed fee ₹8 to 35, ~2% collection, weight shipping, + 18% GST | No |
| Myntra | Category commission + Growth Enablement Fee; Rising Stars gives new homegrown D2C brands 0% | Two-way logistics, + 18% GST on fees, 7 to 15 day gated onboarding | No |
| Nykaa Fashion | ~20 to 30% on jewellery + 18% GST | Extra fee on low-value orders, so sub-₹800 pieces fit badly; curated onboarding | No |
| Meesho | 0% commission | You still eat forward shipping on refused COD; economy AOV ₹199 to 449 | No |
Amazon and Flipkart: search demand, and Amazon's ₹1,000 gift
People search Amazon for "oxidised jhumka" and "anti tarnish necklace" with buying intent you would otherwise rent from Meta. Amazon made that shelf far cheaper in 2026: it dropped the referral fee to zero on products under ₹1,000, fashion jewellery included, so an ₹899 piece pays about ₹100 in fees plus GST instead of a quarter of its price. That is the most seller-friendly fact in this category. The catches are familiar: a new listing spends ₹80 to 140 an order on ads to get seen, the buyer's contact stays masked, and reviews decide everything. Flipkart carries similar traffic but takes more, up to roughly 25% on fashion jewellery above its low-price band, so check your rate card before you price. Method in how to sell on Amazon and how to sell on Flipkart.
Myntra and Nykaa Fashion: the curated floor for demi-fine
Both are gated on purpose, and the gate is the point: their buyers shop fashion with money, which suits demi-fine and silver better than the Amazon bargain aisle. Nykaa Fashion takes 20 to 30% on jewellery plus GST and adds a fee on low-value orders, so anything under ₹800 fits badly there. Myntra charges a category commission plus a Growth Enablement Fee and two-way logistics, with 7 to 15 day onboarding. One on-ramp is worth knowing: Myntra now offers zero commission to new homegrown D2C brands through its Rising Stars programme, a real subsidy for a label that could not otherwise carry a fashion take rate. Both are premium reach once your range has proof, not a day-one home. Steps in how to sell on Myntra and how to sell on Nykaa.
Meesho and your own store: cheapest volume, best margin
Meesho charges 0% commission, which sounds like a gift until you see the rest. Its average order sits at ₹199 to ₹449, so an ₹899 anti-tarnish piece either does not sell or gets relisted at half price against a ₹149 unbranded lookalike. There is no separate RTO fee, but you eat forward shipping on every refused COD order, and Meesho carries the highest RTO in the market. One honest use for a positioned brand: clearing dead designs at season end (how to sell on Meesho). Your own store is the mirror image. It keeps the full ₹899 minus roughly ₹2,000 a month of Shopify and about 2% on Razorpay, with UPI at zero MDR, and hands you the returns policy, the phone number and the reorder. This is the most Instagram-native category in D2C, so Reels-led reach buys a ₹120 to ₹180 CAC against the ₹180 to ₹260 apparel founders pay. Build details in the Shopify store setup guide, reach engine in Instagram marketing for D2C.
Margin Waterfall™: selling price minus COGS, packaging, shipping, commission or gateway, returns and RTO loss, then CAC, run separately for every channel. In jewellery the product and shipping lines are tiny, so the waterfall survives them easily. Two lines decide the winner: the channel take, near zero on Amazon under ₹1,000 and 20 to 30% on Nykaa, and CAC. Compare the bottom number, never the headline fee.
The same ₹899 necklace across four channels
| Line | Your own site | Amazon (under ₹1,000) | Nykaa Fashion | Meesho |
|---|---|---|---|---|
| Selling price | ₹899 | ₹899 | ₹899 | ₹449 (economy reset) |
| COGS + gift packaging | −₹225 | −₹225 | −₹225 | −₹225 |
| Channel take | ~2% gateway + platform | ₹0 referral + ~₹118 fees with GST | −₹265 (25% + GST) | ₹0 |
| Shipping | −₹65 | in fees above | −₹70 | −₹55 |
| Acquisition | −₹150 Meta CAC | −₹130 Amazon ads | −₹40 | −₹20 |
| Returns / RTO provision | −₹55 | −₹60 | −₹60 | −₹70 |
| Net per order | ~₹379 | ~₹366 | ~₹239 | ~₹79 |
| The reorder | Yours: WhatsApp gifting alerts | Masked | Nykaa's customer | Gone |
Read it like an operator. On a single order your own site and Amazon under ₹1,000 are nearly a tie, ₹379 against ₹366, which is a jewellery-specific surprise: the zero referral fee makes a marketplace genuinely cheap here, where in apparel it bleeds you. Nykaa's curation costs a quarter of the price. Meesho, at the price it actually clears, nets ₹79 and teaches you nothing about the buyer. But the single order is not the business. On your own site the second, third and fourth orders arrive through a WhatsApp gifting reminder at near-zero CAC and full margin, so the same customer is worth three or four times what she is worth anywhere else. On a marketplace every order is another masked, cold first order. And price silver over ₹1,000 and the referral fee snaps back, breaking even the first-order tie. So imitation and anti-tarnish under ₹1,000 can genuinely start on Amazon for speed, then migrate the buyer to owned channels. Silver starts owned.
In my supply chain years the number I trusted least was a channel's headline commission, and jewellery is where that instinct pays off. I watched a young anti-tarnish label celebrate a great Amazon month, all of it under ₹1,000 at zero referral, and conclude the marketplace had solved distribution. Then they launched an ₹1,899 silver line on the same listings and the maths inverted overnight: the referral fee came back, ad cost stayed high, and every buyer was still a stranger they could not reach at the next Rakhi. A marketplace order under ₹1,000 is cheap rent, and rent is fine while you are unknown. But jewellery makes its real money on the second and third gift order, and you only get those if you moved the buyer onto a channel you own. Start where sales come fastest, bank the cash, and spend it buying your customers off the platform.
Returns are not the referee here. Reviews are.
In apparel the return rate decides everything, because a third to a half comes back. Jewellery is milder. No bracketing on a pendant, and earrings, a big share of most catalogues, are routinely sold non-returnable on hygiene grounds, though that status is set at listing and sub-category level, so confirm yours before you model it.
Review risk replaces it. A non-returnable earring that arrives with cheap plating and greens a wrist does not come back as a return, it comes back as a one-star review, and on Amazon and Flipkart reviews are the whole game. Plating quality is the variable that makes or breaks a marketplace jewellery brand. Honest anti-tarnish protects your rating; a false claim on cheap plated stock destroys it inside a month, and a buried listing does not recover the way a refunded customer does. Prove every new coating batch on your own traffic first, where a bad batch costs you apologies and a supplier change instead of a rating. Close the COD leak early too, with the COD versus prepaid strategy.
Where the first sales come faster
Speed to first sale tracks your lane. For imitation and anti-tarnish under ₹1,000, a marketplace beats a cold Shopify store, because the search demand already exists and the zero referral fee means you are not paying to rent it. You can be taking "oxidised earrings" orders inside a week while a new store is still buying its first expensive Meta clicks. For 925 silver the opposite holds: a marketplace buyer will not trust an unknown brand at ₹1,999, so first sales come from Instagram and a store where the hallmark and the packaging do the convincing.
If your lane is imitation or anti-tarnish under ₹1,000 and nobody knows you → marketplace-first for discovery, Amazon for the zero-referral shelf, with your own site live from day one collecting WhatsApp opt-ins. If your lane is 925 silver above ₹1,000 → own site and Instagram first, because the fee cliff and the trust gap both work against you on a marketplace. If you do under 10 orders a day → let the marketplace carry discovery while you build the owned base quietly. If you do 10 to 30 a day → run hybrid, own site for margin and repeat, marketplace for reach, a parcel insert pulling buyers across. If you cross 30 a day → own site leads, Amazon becomes a search harvester, Meesho becomes clearance only. If your plating cannot survive an honest water-and-rub test → fix the supplier before you touch any channel, because reviews, not returns, kill a jewellery brand.
The hybrid sequence most jewellery brands should run
According to the Founder Decision Loop™, you add a channel when the last one hands you the proof that justifies it, not when ambition says so.
- Month 0 to 1: own store live, earrings marked non-returnable, WhatsApp opt-in on the order confirmation, prepaid nudge above ₹1,000, Reels of your top ten pieces. If you are imitation or anti-tarnish, list the same pieces on Amazon now to catch search demand under the zero-referral line.
- Month 1 to 2: reorder depth only on what sold. Put an insert card in every marketplace parcel offering a gift discount to reorder on your own site. That card is how rented reach becomes an owned customer.
- Month 2 to 3: with reviews and proof, apply to Myntra Rising Stars for the zero-commission window, or Nykaa Fashion if your pieces clear ₹800. List proven hero pieces only, never the whole tray.
- Month 3 onward: turn on WhatsApp drop and gifting alerts before Rakhi, Valentine's or wedding season, scale the channel with the best Margin Waterfall™ bottom line, and keep Meesho for clearing dead designs.
Pricing a silver line at ₹1,299 to look premium, then listing it on Amazon and wondering where the margin went. Under ₹1,000 the referral fee is zero; at ₹1,299 it switches back on, and with the closing fee, weight handling, 18% GST on all of it and ₹120 of ads to get seen, a piece that netted like a champion at ₹899 now barely breaks even. The founder blames the market. The real error was pricing across the ₹1,000 fee line without moving the channel with it. Silver over ₹1,000 belongs on your own store and Instagram, where you pay for traffic once and keep the customer, not on a marketplace that now takes a cut and still hands you a stranger. In jewellery, price and channel are one decision, never two.
Execution checklist
- Name your lane first, imitation, anti-tarnish or 925 silver, because it picks your channel before any fee table does.
- Run the Margin Waterfall™ per channel on your real numbers, single order and then net after your actual return and RTO rate.
- Price deliberately around the ₹1,000 Amazon line and keep marketplace pieces under it.
- Pass an honest water-and-rub plating test before you list anywhere, because a buried listing does not recover.
- Own store live from day one: WhatsApp opt-in, prepaid nudge on silver, gift-ready packaging.
- Imitation and anti-tarnish list on Amazon early for the zero-referral shelf; 925 silver leads with Instagram and your own store instead.
- Check Myntra Rising Stars eligibility before paying a fashion-marketplace commission you do not have to.
- Put a reorder-discount insert in every marketplace parcel to pull buyers onto owned channels.
- Verify your above-₹1,000 referral rate and your Flipkart, Nykaa and Myntra fees on the live seller dashboards, and confirm your 3% rate (GST for ecommerce sellers).
- Plan drops and stock two months ahead of Rakhi, Valentine's and wedding season, your three biggest revenue events.
Your next action
Open one sheet today. Four columns: own site, Amazon, Nykaa Fashion, Meesho. Two rows: what your hero piece nets on a single order, and what a repeat gifting customer is worth over a year at three or four orders on your own site versus zero owned reorders on a marketplace. Fill it with your real COGS, your real price and the fees above, then draw the ₹1,000 line across it and see which pieces sit on which side. That half hour turns this guide into a channel plan with real numbers, and it runs on the same operating frameworks Ravikant Tyagi uses with founders in this exact category.
If you'd like the complete execution system, calculators, SOPs, templates and operating frameworks behind this process, continue inside D2C Acquisition.Lab.
