Your first decor order is packed and the courier invoice has landed: ₹390 on a ₹1,299 lamp, before any platform takes its cut. That is usually the moment a founder starts asking where to actually sell this stuff. Five doors are open: your own store, Amazon, Flipkart, Meesho, and Pepperfry if your pieces are big enough to matter there. In this category the answer is not decided by commission, and a founder who picks on commission alone picks wrong.
Short version, then the arithmetic. Two numbers settle it before any fee table does. The first is chargeable weight, because couriers and marketplaces bill on length times breadth times height divided by 5000 whenever that beats the scale reading, and a boxy ceramic vase routinely pays for three times its dead weight. The second is who eats the piece when it arrives cracked, which is the most expensive customer conversation in decor and works differently on every channel. Only then does the fee table matter, and it turns on one line: Amazon charges zero referral fee up to ₹1,000 and 17% on its Home Decor node above it, so your price tag and your channel are one decision, not two. Still picking pieces and suppliers? Start with how to start a home decor brand in India and come back with real packed box dimensions.
Sell from your own store from day one, and use marketplaces mainly below ₹1,000. Amazon dropped the referral fee to zero on items priced up to ₹1,000 from 16 March 2026, and charges 17% above that on Home Decor Products, 11% on Home Furnishing, 12.5% on Home Fragrance and Candles, plus a closing fee, a weight handling fee billed on volumetric weight, and 18% GST on all of it. Flipkart has charged 0% commission below ₹1,000 since November 2025 and runs roughly 8 to 15% above it depending on your sub-category rate card. Meesho takes 0% commission but resets your price into its economy band, so it is a clearance lane. Pepperfry suits larger, higher ticket pieces and its rate card is not public, so get it in writing. Freight is the swing factor: a 1.2 kg gift set in a 30 x 24 x 24 cm box ships as 3.5 kg. Damage, not returns, is the loss driver here, and every channel resolves a broken arrival differently. GST varies by material and HSN, from 5% on notified handcrafted lines to 18% on ordinary glassware, so verify per product.
The two numbers that decide this, and neither is the commission
Every channel comparison you have read was written for apparel or skincare, where the parcel is small and the return rate is the story. Decor breaks both assumptions. Nobody orders three vases to pick one, so change of mind returns stay low to moderate. What replaces them is freight and breakage, and both are physical facts about your product that no marketplace negotiation fixes.
Chargeable weight first. Couriers bill the higher of the scale reading and volumetric weight, which is length times breadth times height in centimetres divided by 5000, rounded up to the next slab. In decor the packed box is much bigger than the object, because the protection is the point. Here is what that does to four normal pieces.
| Piece | Dead weight | Packed box (L x B x H) | Volumetric (÷5000) | Chargeable slab | Multiple of dead weight |
|---|---|---|---|---|---|
| Brass urli bowl | 1.8 kg | 22 x 22 x 12 cm | 1.16 kg | 2.0 kg | 1.0x (weight wins) |
| Ceramic vase + brass tray gift set | 1.2 kg | 30 x 24 x 24 cm | 3.46 kg | 3.5 kg | 2.9x |
| Cushion cover set of 4 | 0.7 kg | 32 x 26 x 10 cm | 1.66 kg | 2.0 kg | 2.4x |
| Hanging metal and glass lantern | 0.6 kg | 35 x 35 x 40 cm | 9.80 kg | 10.0 kg | 16.3x |
Read the last two rows together. The lantern weighs a third of the urli and bills at five times its chargeable weight. This is why an assortment decision is a freight decision, and why a decor catalogue built from Instagram screenshots instead of a measuring tape loses money quietly for months. The mechanics, including slab jumps and how to win a wrong weight charge, are in volumetric weight and shipping costs.
The basket this page runs on. One representative order: the ceramic vase and brass tray gift set at ₹1,299, ₹390 of goods at cluster wholesale, ₹120 of fragile packaging, 1.2 kg dead weight, 3.5 kg chargeable. One unit per order, no bundle. That is the middle of the decor AOV band, not the ₹399 coaster and not the ₹2,499 statement lamp. Every number below is built on that one order, so you can swap in your own and recompute.
What each channel actually takes on a ₹1,299 piece
| Channel | Take on ₹1,299 | Other real costs | Who eats a broken arrival | Own the customer? |
|---|---|---|---|---|
| Your own store | None. About ₹2,000/month platform, roughly 2% gateway with UPI at zero MDR, ₹25 to ₹50 COD collection | Freight on chargeable weight, all of your own traffic cost | You, and you also decide the resolution | Yes: phone, email, room and style history |
| Amazon (Home Decor node) | 0% up to ₹1,000, 17% above | Closing fee, weight handling on volumetric weight, 18% GST on fees, ads to be seen | Buyer is refunded first, you claim back through SAFE-T | No |
| Flipkart | 0% below ₹1,000, roughly 8 to 15% above by sub-category | Fixed fee, about 2% collection, shipping fee, 18% GST on fees | You claim through the Seller Protection Fund inside 14 days | No |
| Meesho | 0% commission | System calculated shipping plus 18% GST, and a price reset into its economy band | You, with no cover worth planning on | No |
| Pepperfry | Not published. Reported 15 to 30% by category | Its own heavy goods logistics, studio assisted selling | Set by your seller agreement, so read it | No |
Amazon: the ₹1,000 cliff, and the node you sit in
Amazon expanded zero referral fees to products priced up to ₹1,000 across more than 1,800 categories from 16 March 2026, home decor and furnishings included. Below that line the marketplace is close to free shelf space. Above it, Amazon's published rate card charges 17% on Home Decor Products, 11% on Home Furnishing excluding curtains, and 12.5% on Home Fragrance and Candles, all for item prices above ₹1,000.
Those three rates matter more than founders expect, because a decor catalogue spans all of them. A cushion cover and a table runner are furnishing at 11%. A scented candle is fragrance at 12.5%. A vase is decor at 17%. On a ₹1,299 tag the gap between the 17% node and the 11% node is ₹78 of fee plus ₹14 of GST, so ₹92 an order. Classify honestly, then notice that your product mix is quietly setting your blended take rate. The same page confirms the freight rule: shipping fees are computed on volumetric or actual weight, whichever is higher, with the divisor at 5000. So a boxy piece pays the referral fee on its price and the handling fee on its air. Listing mechanics are in how to sell on Amazon in India.
One more line worth knowing before you build freight into your tag: the referral fee is computed on the total price the buyer pays. On a ₹1,299 tag carrying the ₹180 of Easy Ship weight handling used in every Amazon line on this page, about ₹31 of that referral fee is commission on your own courier bill.
Flipkart: cheaper above ₹1,000, if your rate card says so
Flipkart moved to a zero commission model on products listed under ₹1,000 in November 2025, and extended zero commission across all of fashion at any price in July 2026. Decor is not fashion, so above ₹1,000 you are back on the rate card, generally 8 to 15% depending on the exact sub-category. That band is wide enough to change the answer: at 8% Flipkart is the cheapest paid channel for a ₹1,299 piece, at 15% it is roughly level with Amazon. Fixed fee, collection fee, shipping and 18% GST sit on top either way. Do not model this from a table on the internet, including this one. Pull your own rate card from the seller hub before you list. Steps are in how to sell on Flipkart in India.
Meesho and Pepperfry: two very different edges
Meesho charges 0% commission and that is genuinely the whole commission story. The catch is price, not fee. Its buyer expects a decor piece at ₹199 to ₹649, so a ₹1,299 curated set either does not move or gets relisted at half price beside an unbranded lookalike. You still pay system calculated shipping plus GST, and the COD heavy, price led base puts it at the top of the RTO band. Use it to clear dead designs at season end, not to build a brand (how to sell on Meesho).
Pepperfry is the opposite problem. It has real home intent traffic, a studio network for assisted selling, and it handles big items as a matter of routine, which suits mirrors, large planters, consoles and lighting far better than a ₹399 impulse piece. Its commission is not published, and reported bands run 15 to 30% by category. That range is too wide to plan on, so treat the rate card as the first thing you ask for and the thing that decides whether you list at all.
Margin Waterfall™: selling price minus COGS, packaging, shipping, commission or gateway, RTO loss, then CAC, run once per channel. Decor changes two lines and adds a third. Shipping is billed on chargeable weight, not the scale reading. Packaging runs 8 to 15% of the selling price, because that spend is what buys the damage rate down. And breakage gets its own line, separate from returns, because it behaves nothing like a change of mind. According to the Margin Waterfall™ framework, contribution margin is calculated on volumetric freight and a damage provision before the ad budget is set.
Who pays when it arrives broken
Roughly 11% of unit loads arrive damaged somewhere in the chain, and fragile ceramics and glass sit at the wrong end of that. Engineered packaging is what pulls your realised rate down toward 3 to 5%. What changes by channel is not the physics, it is who carries the loss and how fast the customer gets their money back.
- Your own store. Entirely your loss, and entirely your call. You can replace once, refund, or send a partial credit, and you keep the customer either way. A replacement costs one extra unit, one extra pack and one extra freight leg, about ₹740 on this basket. A refund costs the whole order, about ₹1,080 including the burnt CAC. That gap is why a replace first policy is the cheaper policy.
- Amazon, Easy Ship or self ship. The buyer is refunded through Amazon's returns flow and you recover through a SAFE-T claim on evidence. Amazon stopped processing SAFE-T claims for product primary packaging or box damage and for broken or open seals from 1 August 2025. In a category sold in gift boxes, that is a real narrowing: a crushed outer box around an intact piece is now your loss.
- Amazon FBA. Fine for dense, fast moving goods, poor for decor. Storage is billed on cubic feet, so a boxy slow seller pays rent on air through a long off season, and inventory damaged inside the network reimburses on your manufacturing cost, not your selling price. At 30% COGS that returns about a third of the sale.
- Flipkart. Claim through the Seller Protection Fund, within 14 days of the returned product reaching you, with photographs of the damage and of the shipping label. Miss the window and there is nothing left to argue about.
- Meesho. You eat it. Price for that or do not list fragile pieces there.
The operational answer is the same on every channel and costs almost nothing: photograph each fragile parcel packed, sealed and labelled, with a timestamp, and keep the images for 60 days. Every claim on every platform turns on packing evidence, and the seller who has photos wins arguments the seller who does not never even opens. The packing spec itself is in shipping packaging and protection for ecommerce.
Pricing a marketplace piece at ₹1,199 to look premium. Run it on Amazon's Home Decor node. At ₹1,199 you pay 17% referral (₹204), a closing fee (about ₹50), weight handling on a 3.5 kg chargeable parcel (about ₹180) and 18% GST on all of it (₹78). Fees ₹512, so ₹687 reaches you. Tag the same piece at ₹999 and the referral fee is zero: closing about ₹40, handling ₹180, GST ₹40, fees ₹260, and ₹739 reaches you. You charged ₹200 less and banked ₹52 more, before the conversion lift a sub-₹1,000 price gives you. Solve for where the higher tag catches up and it lands near ₹1,265 on these assumptions. Everything you price between ₹1,001 and about ₹1,264 sits in a dead zone that pays the platform and costs you. Founders give away ₹40 to ₹60 an order in that band for months without seeing it, because the fee shows up in the settlement report, not on the sale.
The same order, run through every door
Per delivered order first, before breakage and RTO. Meesho gets its own column at the price it actually clears, on a cheaper piece, because pretending a ₹1,299 curated set sells there would make the table a lie.
| Line | Your own store | Amazon (Easy Ship) | Flipkart (12% mid) | Meesho (₹649 piece) |
|---|---|---|---|---|
| Selling price | ₹1,299 | ₹1,299 | ₹1,299 | ₹649 |
| COGS | −₹390 | −₹390 | −₹390 | −₹200 |
| Fragile packaging | −₹120 | −₹120 | −₹120 | −₹70 |
| Commission / referral | ₹0 | −₹221 (17%) | −₹156 (12%) | ₹0 |
| Closing / fixed / collection | −₹13 platform | −₹50 | −₹71 | ₹0 |
| Freight on chargeable weight (3.5 kg; Meesho 2 kg) | −₹230 | −₹180 | −₹200 | −₹150 |
| Gateway and COD collection / 18% GST on fees | −₹28 | −₹81 GST | −₹77 GST | −₹27 GST |
| Acquisition | −₹340 Meta | −₹140 ads | −₹110 ads | −₹25 |
| Net per delivered order | ₹178 | ₹117 | ₹175 | ₹177 |
Now the honest number, per kept order. Two separate deductions, never one blended reserve, because these are different failures with different fixes. Damage is a packaging and handling problem. RTO is a payment mode and address problem. The house rule on RTO: drag per delivered order equals the rate divided by one minus the rate, times the cost of one failed parcel, which here is forward freight plus reverse freight plus packaging plus burnt CAC. Meesho is the one exception below: its ₹245 is forward freight plus packaging plus burnt CAC, because a pure RTO there carries no reverse shipping deduction, unlike a customer return. Check your own settlement report before you trust that. At 8% blended RTO you are carrying 0.087 failed parcels for every delivered one, not 0.08.
| Channel | Net per delivered order | Damage provision | RTO drag | Net per kept order |
|---|---|---|---|---|
| Your own store | ₹178 | −₹35 (4% x ₹876) | −₹80 (8% blended, ₹920 per failed parcel) | ₹63 |
| Amazon (Home Decor node) | ₹117 | −₹42 (5% x ₹833) | −₹40 (6% blended, ₹620 per failed parcel) | ₹35 |
| Flipkart (12% mid) | ₹175 | −₹41 (5% x ₹826) | −₹47 (7% blended, ₹630 per failed parcel) | ₹87 |
| Meesho (₹649 piece) | ₹177 | −₹31 (7% x ₹440) | −₹69 (22% blended, ₹245 per failed parcel) | ₹77 |
Read it like an operator, and read the surprises. First, nothing here is fat. On a cold first order at ₹1,299 every channel lands between ₹35 and ₹87, which is 3 to 7% of the tag. That is what bulky and fragile costs, and it is why decor founders who chase revenue instead of AOV and repeat run out of cash around month eight. Second, Amazon comes last on this piece, and that is a verdict on the piece, not the platform. Third, Flipkart wins here only at 12%. At 15% its net per kept order drops to about ₹41, which is Amazon territory; at 8% it climbs to about ₹148. Fourth, Meesho's number looks respectable and is not comparable: different piece, lower price, nothing learned about the buyer, no second order.
Now change the piece instead of the platform, and watch the same marketplace behave differently. A compact brass diya set at ₹949, 0.9 kg dead in a 20 x 16 x 10 cm box so 1 kg chargeable, ₹290 of goods and ₹80 of packing. On Amazon it pays no referral fee at all, roughly ₹40 closing, ₹90 handling, ₹23 of GST on fees and ₹90 of ads. That nets ₹336 per delivered order and about ₹303 per kept order after a 2% damage provision and a 6% RTO drag. Same platform, same seller, nearly nine times the profit of the ₹1,299 vase set, and both levers did the work: under ₹1,000, and dense. That single comparison should decide your marketplace assortment.
The second order is the whole point. On your own store the same customer comes back for the matching tray, the second cushion, the festive set, at close to zero CAC and full margin. That is the only place ₹63 becomes ₹300 across a year. Decor repeat is not replenishment, since nobody runs out of a vase, so it runs on collection breadth and on you being able to reach her when the new drop lands. On a marketplace every order is another masked stranger. Build the owned base from day one (Shopify store setup guide for India), and get more out of each parcel by pushing carts to two pieces in one right sized box (how to increase average order value).
In my supply chain years the cheapest control we ever added was a camera. Not software, a camera. Every disputed load came down to what the packing looked like at handover, and the side with photographs won. Decor founders learn this the expensive way. I watched one argue a run of cracked ceramic with a courier for six weeks holding nothing but an angry customer's photo of a broken vase, which proves the piece is broken and proves nothing about whose fault it is. Two changes fixed it. A written pack spec, box size, wrap turns, void fill, corner protection, so every parcel leaves identical. And a timestamped photo of every fragile parcel sealed and labelled. Breakage went from just under 9% to about 4% inside a month, and the claims still worth filing started getting paid. Sequence matters here: fix the box before you argue about the channel, because a 9% damage rate makes every fee table on this page irrelevant.
GST in decor is not one rate, and the marketplace uses whatever you declare
There is no single decor GST rate, and any page that hands you one is guessing. Under GST 2.0 the Council cut a long list of handcrafted goods to 5% with effect from 22 September 2025, covering ornamental ceramic articles, glass artware, metal artware in brass, copper and iron, carved and inlaid wood, handcrafted candles and handmade carpets, per the GST Council's own press release on the 56th meeting. The mass produced twin of the same object is not handcrafted. Ordinary glassware for table or indoor decoration is commonly taxed at 18%, and lamps split: handcrafted lamps and traditional lanterns are 5%, ordinary lamps and lighting fittings 18%. GST 2.0 removed the 12% slab for goods on 22 September 2025, so no decor line is 12% any more, whatever an old invoice says. Two vases that look identical on your shelf can carry different HSNs and different rates.
On a marketplace you declare that HSN at listing and the platform collects tax on it. Get it wrong and you either overcharge and lose the price comparison, or undercharge and carry a liability that surfaces at assessment. Three things to settle with your accountant before you list, not after:
- The correct HSN and rate for every material line you sell, verified post GST 2.0 rather than copied off an old invoice.
- Registration. Ship to even one other state and GST is mandatory from day one whatever your turnover. Sell only inside your own state through a marketplace and Notification 34/2023 lets you run on a PAN based enrolment number instead, which is the lane Meesho opens for artisans. Either way the platform deducts TCS at 0.5% of net taxable supplies, which you claim back in your returns.
- The credit trap nobody mentions. If your output sits at 5% while the marketplace charges 18% GST on its commission and fees, that credit builds faster than you can use it, and the inverted duty refund route is computed on input goods, not input services. Ask your CA what that does to working capital before you assume the 18% comes back. Background in GST for ecommerce sellers in India.
If your hero pieces land under ₹1,000 → list on Amazon and Flipkart now, because the referral fee is zero and the search demand is already there. If your hero pieces sit at ₹1,200 to ₹2,500 → own store and Instagram lead, and test Flipkart before Amazon because the commission band is lower. If a piece is priced between ₹1,001 and about ₹1,264 → either bring it under ₹1,000 or push it past ₹1,300, because the middle pays the platform and not you. If a piece bills at more than three times its dead weight → redesign the box, bundle it with something dense, or drop the SKU. If your damage rate is above 6% after 100 orders → fix the pack spec before you add any channel. If your pieces are large or furniture adjacent → ask Pepperfry for the rate card and decide from that number, not from its traffic. If COD is more than half your orders → close that leak first with a COD versus prepaid strategy, because a bulky RTO parcel costs about ₹920 and often comes back unsellable.
The sequence that works
According to the Founder Decision Loop™, you add a channel when the last one hands you proof, not when ambition says so. In decor that sequence is unusually clean, because the ₹1,000 line does most of the sorting for you.
- Month 0 to 1. Own store live. Measure and photograph every packed box, set freight from chargeable weight, write the pack spec, and run 20 orders to get a real damage number. Prepaid nudge above ₹999.
- Month 1 to 2. Take the pieces that genuinely sit under ₹1,000, or that split into a single piece under ₹1,000, and list them on Amazon and Flipkart. Zero referral fee, real search demand, no reason not to. Put a reorder card in every parcel.
- Month 2 to 3. Keep the ₹1,200 plus statement pieces on your store, where the story and the box do the convincing and nobody takes 17%. If you want marketplace reach above ₹1,000, pull the Flipkart rate card first and list only if it comes in under about 12%.
- Month 3 onward. Approach Pepperfry only if your range has moved toward larger, higher ticket pieces. Keep Meesho for clearing dead designs on non fragile SKUs. Scale the channel with the best net per kept order, not the best headline fee, and start the festive build eight to ten weeks early.
- Measure the packed box, not the product, and run L x B x H ÷ 5000 on every SKU before you set a price.
- Split the catalogue at ₹1,000 deliberately: marketplace SKUs under it, statement pieces above it on your own store.
- Check which Amazon node each SKU truly belongs in, because furnishing at 11% and candles at 12.5% are not decor at 17%.
- Pull your Flipkart rate card in writing before listing anything above ₹1,000; 8% and 15% are two different businesses.
- Write a fixed pack spec, then photograph every fragile parcel sealed and labelled with a timestamp, kept for 60 days.
- Keep fragile pieces on Easy Ship or self ship so you control the pack, and keep bulky slow movers out of per cubic foot storage.
- Model damage and RTO as two separate lines with two separate fixes, never as one blended reserve.
- Verify the HSN and GST rate for every material you sell post GST 2.0, and ask your CA about 18% fee credit against a 5% output.
- Put a reorder card in every marketplace parcel, because collection breadth is what brings a decor buyer back.
- Recheck every fee on this page each quarter, and renegotiate courier rates once you cross 300 shipments a month.
Your next action
Today, take your three best selling pieces and build one sheet. Row one, the packed box dimensions and the chargeable weight. Row two, the tag price and which side of ₹1,000 it falls on. Row three, the net per kept order on your own store and on the one marketplace you are seriously considering, using your real COGS and the fee lines above. Then move any piece sitting between ₹1,001 and ₹1,264 to one side or the other. That half hour usually finds ₹40 to ₹90 an order that was going to a platform for no reason, and it runs on the same operating frameworks Ravikant Tyagi uses with founders in this category.
If you'd like the complete execution system, calculators, SOPs, templates and operating frameworks behind this process, continue inside D2C Acquisition.Lab.
