You have ₹1 lakh and a saved folder full of brass, ceramic and carved wood. Here is the direct answer before anything else. At this tier the money buys about 144 pieces across 12 to 16 designs from two sourcing clusters, an engineered fragile pack, day-one compliance, and roughly ₹22,000 of ring-fenced ad money. Not four clusters. Not six designs stocked deep. Wide on designs, narrow on clusters.
That is the opposite of the usual advice about going deep on a few winners, and the reason is specific to this category. Nobody runs out of a vase. In skincare or coffee the bottle empties, so depth pays. In decor a customer comes back only because you have something she has not seen, for a room or an occasion she has not bought for yet. Your catalogue is your retention plan. So breadth earns its place here in a way it never does in a consumable category, and this page is about where breadth is cheap and where it quietly costs you ₹3,000 you did not budget.
₹1 lakh in home decor splits roughly: ₹36,000 of stock ex-factory, ₹4,000 inbound freight, ₹9,000 packaging, ₹8,000 photography, ₹7,000 store, ₹5,500 compliance, ₹4,000 samples, ₹22,000 ring-fenced ads, ₹4,500 reserve. That buys 12 to 16 designs at the cluster MOQ of 12 pieces a design, from two clusters and not more, because every extra cluster is a separate part-load consignment with its own minimum charge, its own packing standard and its own lead time. This category is decided by freight and breakage, not by returns. Couriers bill on volumetric weight, length x breadth x height in cm divided by 5000, so a 0.6 kg ceramic vase in a 24 x 24 x 30 cm box ships as 3.46 kg. Fragile packaging runs 8 to 15% of the tag and is what buys your damage rate down from the roughly 11% industry baseline for fragile loads to 3 to 4%. GST follows the HSN of the exact article, not the word decor, and the gap between 5% and 18% on a ₹1,199 tag is ₹126 an order. Launch settings on a ₹1,199 blended basket lose ₹49 an order. Month three, with a wider basket, more prepaid and a ₹280 CAC, makes ₹93 an order and about ₹2,100 of owner profit on ₹72,820 collected. At ₹1 lakh, the capital's job is data, not profit.
What ₹1 lakh buys in home decor that ₹50,000 cannot
At ₹50,000 you can only afford one material. About ₹18,000 of stock is six designs at the cluster minimum of 12 pieces, from one cluster, in one material. That is a brass shop with a logo, and it caps your basket at one piece an order.
₹1 lakh buys the second cluster, and the second cluster is the whole point. Brass alone is a product. Brass plus ceramic is a styled table, which is what makes a set, which is what lifts the basket. And the freight math rewards the set hard. On the courier card used through this page, a single-piece parcel bills at about ₹141 and a two-piece parcel at about ₹189. The second piece adds ₹48 of freight and ₹625 of order value. That single line is why breadth in materials pays for itself in decor, and it is not available to you at ₹50,000. The ₹50,000 version of this plan sits in the home decor category guide.
| Line | ₹50,000 tier | ₹1 lakh tier |
|---|---|---|
| Clusters | 1 | 2 |
| Designs | 6 to 8 | 12 to 16 |
| Pieces at MOQ 12 | 72 to 96 | 144 to 192 |
| Packaging | Plain 3-ply box plus wrap, about ₹55 | Engineered 5-ply pack, ₹110 to ₹160 |
| Typical basket | One piece, ₹749 to ₹949 | Singles and pairs, ₹1,199 blended |
| Ring-fenced ad money | ₹8,000 to ₹10,000 | ₹22,000 |
| What it can prove | That one material sells | That a collection sells, and at what damage rate |
Breadth or depth: the call this budget actually makes
Depth means fewer designs with more pieces behind each. Breadth means more designs with the minimum behind each. Most category advice says depth, because most categories replenish. Decor does not, so the honest answer here is breadth, with three hard limits.
Breadth in designs is cheap. Breadth in clusters is not. Adding a thirteenth design to an existing Moradabad order costs you 12 pieces of cash on a shelf, one more product shoot, and nothing else. Adding a third cluster costs you a separate consignment, a separate minimum freight charge, a separate lead time, a separate packing standard and a supplier you will never have time to call. Those are different orders of expense that founders treat as the same decision.
Breadth in designs is only cheap if the designs share a box. Two designs that need two different box sizes cost you twice in packaging inventory, twice in the drop testing, and twice in the freight learning. Pick one box that fits about 80% of the range with 4 to 5 cm of cushion on every face, and one larger box for the outliers. Designs that will not fit either box do not get bought this quarter, however good they look.
Depth is bought later, with data. You will find three or four winners out of 12 to 16. Those get restocked deep from month three, funded by revenue, not by the launch ₹1 lakh. That is the only depth worth paying for, because it is the only depth you have evidence for.
Inventory Confidence Model™: order depth is a function of demand certainty, never of ambition. At ₹1 lakh with zero sales history, your confidence on any single design is close to zero, so depth is a bet and breadth is a test. Buy the supplier minimum on every design, 12 pieces, and let 60 days of sell-through tell you which three deserve 60 pieces. Being wrong about a design at 12 pieces costs you ₹3,000 of dead stock. Being wrong about the same design at 60 pieces costs you ₹15,000 and the working capital you needed for ads.
Two clusters, not four: what splitting the order really costs
India gives a decor founder an unfair advantage: real manufacturing clusters that sell wholesale to anyone, at minimums a small brand can actually meet. Moradabad for brass, copper and aluminium. Khurja for ceramics. Jodhpur and Saharanpur for wood. Firozabad for glass. Jaipur for block-print textiles and blue pottery. Channapatna for lacquered wood. Twelve to fifty pieces a design is a normal artisan or wholesale run, and sampling is cheap.
The trap is that it all looks equally easy on IndiaMART, so a founder places three ₹12,000 orders in three cities and feels efficient. Here is the bill.
A ₹12,000 buy at about ₹250 a piece is roughly 48 pieces, which is somewhere near 35 to 45 kg packed. Part-load surface transport is minimum-charge driven at that size, so a 35 kg consignment costs close to what a 70 kg one costs. Landed at a metro from western Uttar Pradesh, expect ₹800 to ₹2,500 per consignment including local cartage, depending on distance, whether you take door delivery or collect from the transport nagar godown, and whether the goods are booked with fragile handling. Jodhpur is roughly 600 km from Delhi against Khurja's 90 and Moradabad's 160, so the same size consignment costs more from there.
So three clusters is three minimums instead of two, on smaller consignments each. On a ₹36,000 stock buy that is ₹2,000 to ₹3,000 of pure loss, which is a full product shoot or a week of ads. And freight is the cheap part of the mistake. The expensive part is that each cluster has its own packing standard, and the packing standard you accept at the supplier's gate is your damage rate at the customer's door.
The one exception worth taking. Corridor clusters can consolidate. Khurja and Moradabad both feed the Delhi corridor, so a single transporter can often collect both and bill you as one consignment. Jodhpur, Jaipur or Channapatna are separate corridors and separate bills, every time. Ask the transporter before you assume it. The full supplier hunt and vetting method sits in how to find manufacturers and suppliers in India, and the minimum-order conversation itself is in how to negotiate MOQ with suppliers.
In my supply chain years the number that surprised new planners was never the per-kg rate. It was the minimum. Freight is a step function, not a slope, and small consignments sit on the expensive step. I have watched founders win a ₹15 per piece saving at the workshop and hand all of it back at the transport booking counter, because they split a buy that should have been one consignment. My rule at this budget: consolidate first, negotiate second. Get the order shape right, then argue about price. And insist the supplier packs for a truck journey, not for a handcart across the lane, because you cannot repack 144 pieces in your living room after the fact.
The exact ₹1,00,000 allocation
Three principles hold this together. Stock and its inbound freight are one number, not two, because unlanded cost is a fiction. The ad budget is ring-fenced and no other line may raid it, because it is the only line that produces an answer. And packaging plus photography together are allowed to sit close to half the stock value, because in a category sold on how something looks and judged on whether it arrived whole, those two lines are the product.
| Line item | Allocation | Notes |
|---|---|---|
| Samples from 6 to 8 suppliers across 2 clusters, plus their courier | ₹4,000 | Three of each shortlisted design, because one goes into the drop test |
| Stock, cluster A (metal, Moradabad): 7 designs x 12 pieces | ₹21,000 | 84 pieces at about ₹250 ex-factory |
| Stock, cluster B (ceramic, Khurja): 5 designs x 12 pieces | ₹15,000 | 60 pieces at about ₹250 ex-factory |
| Inbound freight and cartage, 2 consignments | ₹4,000 | About ₹2,000 each landed at a metro; more if a cluster sits off-corridor |
| Fragile packaging system | ₹9,000 | Two box sizes, moulded or corrugated inserts, wrap, void fill, edge protectors, tape, labels |
| Product photography and styling props | ₹8,000 | One surface set, two backdrops, phone plus one styled shoot day |
| Domain plus Shopify, 3 months | ₹7,000 | The store gets 60 days to earn month three's rent |
| GST registration plus trademark | ₹5,500 | ₹4,500 government fee per class for individuals and small enterprises. The class follows the material: Class 6 for brass and metal artware, Class 21 for ceramic and glass, Class 20 for wood. File the class your biggest line sits in, before you print boxes |
| Ad and validation budget, ring-fenced | ₹22,000 | Meta plus one creator seeding test; untouchable by every line above |
| Damage and restock reserve | ₹4,500 | Replacements, one reshoot, the first broken arrivals |
| Total | ₹1,00,000 |
That is 12 designs and 144 pieces, ₹36,000 ex-factory plus ₹4,000 inbound, so ₹278 landed per piece. Every number after this on the page is built on that ₹278. If your average ex-factory lands nearer ₹190 a piece, the same ₹36,000 buys 16 designs instead of 12, which is the version of this plan you want if you can find it.
What is not worth a rupee at this tier: custom moulds, exclusive designs, a designer logo, a third cluster, a studio shoot, or any SKU that needs its own box size. Those are ₹5 lakh problems. The pricing method that turns landed cost into a tag is in how to price a product in India.
Volumetric weight decides which SKUs you are allowed to sell
Volumetric weight is the parcel's size expressed as a weight. Indian couriers compute it as length x breadth x height in centimetres divided by 5000, and they bill you on the higher of that or the actual weight, because a truck fills on volume before it fills on weight. Shiprocket publishes the same formula and the same rule. Bulky decor routinely pays two to four times its dead weight, which means the courier's ruler, not your scale, decides what you are allowed to stock.
Every freight figure on this page uses one card: ₹45 for the first 500 g and ₹24 for every additional 500 g, surface, on a zone-local to regional aggregator card. Your card will differ. National zones run ₹65 to ₹85 for that first 500 g alone, which makes every row below worse, and aggregator rates only come down once you cross a few hundred shipments a month. Swap your numbers in. The shape does not change.
| SKU | Dead weight | Packed box (cm) | Volumetric (÷5000) | Billed | Freight |
|---|---|---|---|---|---|
| Brass diya set, 4 pieces | 1.1 kg | 20 x 18 x 12 | 0.86 kg | 1.1 kg | ₹93 |
| Brass planter, small | 2.2 kg | 20 x 18 x 16 | 1.15 kg | 2.2 kg | ₹141 |
| Wooden tray plus 2 coasters | 1.4 kg | 36 x 28 x 10 | 2.02 kg | 2.02 kg | ₹141 |
| Ceramic vase, 20 cm | 0.6 kg | 24 x 24 x 30 | 3.46 kg | 3.46 kg | ₹189 |
| Metal lantern, large | 0.9 kg | 26 x 26 x 34 | 4.60 kg | 4.60 kg | ₹261 |
Read the last two rows together, because that pair is the category in one line. The 0.9 kg lantern costs ₹261 to ship. The 2.2 kg brass planter, more than twice its weight, costs ₹141. Light and boxy is expensive. Heavy and compact is cheap. If you price on the scale reading you will lose money on exactly the SKUs that photograph best.
Three consequences. Run every design through the formula before you place the order, not before you price it, because the answer is sometimes do not buy this design. Weight the assortment toward dense goods, metal, wood and textiles, and be deliberate about how many large ceramic and glass pieces you carry. And right-size the box, but not below the cushioning you need, which is the tension in the next section. Weight slabs, zones and re-weigh disputes are covered in volumetric weight and shipping costs in India.
Buying one box size bigger than needed, for safety. It feels prudent. Take the 24 x 24 x 30 cm ceramic vase pack: 3.46 kg volumetric, ₹189 on our card. Go up 5 cm on every side, to 29 x 29 x 35 cm, and volumetric jumps to 5.89 kg, which bills as 6.0 kg and costs ₹309. That is ₹120 an order for empty air, every order, or ₹6,600 across 55 orders a month, which is almost the entire ₹7,425 those same 55 orders spend on packaging material. The fix is not a smaller box, because you still need 4 to 5 cm of cushion on all six faces. The fix is a box built for the cushion instead of a box you found. Order two custom sizes from a corrugation unit once your range is set; minimums run 500 to 1,000 boxes and the per-box price usually lands below stock sizes anyway.
Packaging engineering, and the drop test you run before the bulk order
Fragile packaging costs 8 to 15% of the selling price. On a ₹1,199 basket that is ₹96 to ₹180, and it is a real line, not a rounding error. Here is what the money buys, per parcel:
- Five-ply double-wall outer box, right-sized: about ₹38
- Moulded or corrugated insert that holds the piece off every wall: about ₹30
- Bubble wrap, two layers on the item, plus air pillows or paper void fill: about ₹25
- Reinforced tape and corner or edge protection: about ₹12
- Tissue, thank-you card, fragile and this-way-up labels: about ₹25
That is ₹130, or roughly 11% of the tag. Now the honest part, because the usual pitch for good packaging overclaims. Industry damage data puts roughly 11% of unit loads arriving damaged; take that as your unmanaged case for fragile goods. A properly engineered pack should hold you at 3 to 4% on ceramics and glass, 1 to 2% on metal and wood. On a ₹1,142 net order, moving the damage line from 11% to 4% saves about ₹80. The pack that gets you there costs ₹130 against maybe ₹55 for a plain box and a wrap, so the extra ₹75 roughly pays for itself on the damage line alone. In rupees it is close to break-even.
So why do it? Because the damage line is not the whole cost of a broken arrival. A cracked vase is a refund conversation, a replacement to ship, a customer who does not come back, and a one-star review with a photo that sells against you for years in a category bought on looks. Fragile packaging is self-funding on the arithmetic and profitable on everything the arithmetic misses. The material-by-material stack sits in shipping packaging that survives Indian couriers.
Order three pieces of every shortlisted design. Pack one exactly as you intend to ship it. Drop it onto concrete from waist height, about 75 cm: once on a bottom corner, once on the longest edge, once flat on each of the two largest faces. If the piece survives but the insert has collapsed, the pack fails, because a real parcel takes more handling than four drops. Then do the part founders skip: ship one sample to another zone by COD and have the recipient refuse it, so it returns as an RTO and you see what two full journeys do to your pack. The lab version is ISTA 3A, the recognised parcel test for packages up to about 70 kg, and it is worth paying for later. At ₹1 lakh, a concrete floor and one refused parcel tell you nearly everything.
GST by HSN: there is no home decor rate
This is the compliance line that costs real money, and most decor pages get it wrong by quoting a single rate. There is no decor rate. The rate follows the HSN classification of the exact article, and two objects that look identical on a shelf can sit in different lines.
The direction of travel helped you. Under GST 2.0, the 56th GST Council recommended a reduction from 12% to 5% on labour intensive goods including handicrafts, effective 22 September 2025, and CBIC notified the handicraft lines in Notification 13/2025-Central Tax (Rate) dated 17 September 2025. In practice, handmade pottery and ceramic artware, brass and copper artware, carved wood articles and handmade candles largely sit at 5%. Decor that plugs in is a different story: lamps and lighting fittings under HSN 9405 commonly sit at 18%, as do many plastic and machine-made articles.
Here is what the difference is worth. Tag a piece at ₹1,199 inclusive. At 5% your net revenue is ₹1,142. At 18% it is ₹1,016. That is ₹126 an order, about ₹6,900 across 55 orders in a month, and it is the difference between the month three P&L below working and not working.
Two rules. Get the eight-digit HSN in writing from every supplier, on the invoice, before you buy, then verify it yourself against your own article, because their classification is for their sale and is usually but not always yours. And be careful with gift sets. Under Section 8 of the CGST Act, a mixed supply, two or more goods sold together for a single price that are not naturally bundled, is taxed at the highest rate among them. Put a 5% brass diya and an 18% LED string light in one box at one price and the whole box can go to 18%. The general ecommerce tax picture is in GST for ecommerce sellers in India.
The rest of the compliance stack is light. No single licence governs general decor and there is no mandatory BIS certification for non-electrical pieces, though anything with wiring pulls you into electrical safety rules. What does apply is Legal Metrology: a packaged retail pack must declare your entity name and address, the common or generic name of the article, net quantity or piece count, MRP inclusive of all taxes, month and year of packing, and a consumer care contact. Country of origin goes on the pack only for imported goods under Rule 6, but every marketplace listing asks for it anyway, so keep it ready. Those declarations have to appear on your listings too.
The P&L, on a stated blended basket
Every number below runs on one basket, stated up front so you can rebuild it with your own. Launch basket ₹1,199 collected: 60% of orders are one piece at an average ₹949, 40% are two pieces at an average ₹1,574, so 1.4 pieces an order. COGS is ₹278 landed per piece. Freight is the blended parcel from the table above. Payment gateway is 2% plus GST on the fee. RTO is modelled on the house convention, drag per delivered order = rate ÷ (1 minus rate) x (forward + reverse + packaging + burnt CAC), at a blended 11% (45% COD orders at a 22% COD RTO rate, 55% prepaid at 2%). Damage sits on its own line at 4% of net revenue, because in this category a write-off is not a return.
Minus ₹49. That is not a typo and it is not pessimism, it is what launch settings look like before anything is tuned. Contribution before CAC is ₹291, which means ₹291 is your breakeven CAC on day one. Note also that COGS at ₹389 is 34% of net revenue, sitting inside the category's normal 25 to 40% band. That is a cost line. It is not your margin, and reading it as margin is how founders end up here without noticing.
Three levers close the gap, and the honest thing about them is that no single one is enough.
| Lever | What changes | Contribution after | Worth |
|---|---|---|---|
| Launch settings | Basket ₹1,199, COD 45%, CAC ₹340 | −₹49 | Starting point |
| 1. Basket mix | Two-piece share 40% to 60%, basket ₹1,324 | −₹11 | +₹38 |
| 2. Prepaid share | COD 45% to 25%, blended RTO 11% to 7% | +₹29 | +₹39 |
| 3. CAC | ₹340 to ₹280 as one creative starts working | +₹93 | +₹64 |
Lever 1 works because the second piece adds ₹48 of freight and ₹625 of order value. Lever 2 is the cheapest of the three and the one founders postpone; prepaid nudges, partial COD advance and address verification are covered in the COD vs prepaid strategy guide. Lever 3 looks like it should be worth ₹60 and is worth ₹64, because CAC is burnt twice: once on the sale and again inside every RTO. According to the Margin Waterfall™ framework, contribution is calculated on volumetric freight and a separate damage line before the ad budget is set, and that is exactly the calculation that turns a ₹340 CAC from acceptable into fatal.
Month three, honestly
Fifty-five delivered orders in month three is a fair result for a ₹1 lakh launch with ₹22,000 of learning money behind it. Basket at ₹1,324, contribution ₹93 an order.
| Line | Month 3 |
|---|---|
| Delivered orders | 55 |
| Collections (55 x ₹1,324) | ₹72,820 |
| GST out at 5% | −₹3,465 |
| Net revenue | ₹69,355 |
| COGS (55 x ₹445) | −₹24,475 |
| Packaging (55 x ₹135) | −₹7,425 |
| Freight (55 x ₹170) | −₹9,350 |
| Payment gateway | −₹1,705 |
| RTO drag (7% blended) | −₹3,135 |
| Damage write-off (4%) | −₹2,750 |
| Ad spend (55 x ₹280) | −₹15,400 |
| Contribution | ₹5,115 |
| Fixed costs (store, apps, domain) | −₹3,000 |
| Owner profit | ₹2,115 |
₹2,115 on ₹72,820 collected. Look at it clearly. Month three of a ₹1 lakh decor brand pays for lunch, not rent, and any page telling you otherwise is selling something. What it does buy is the only thing that matters at this stage: a landed cost you trust, a damage rate you measured, a basket you know how to lift, and three designs with real sell-through.
One cash-flow warning. Your 144 pieces at 1.6 pieces an order is about 90 orders of stock. At 55 orders in month three you run out during month four, and that restock comes out of month three's cash, not out of the launch ₹1 lakh, which is gone. That is what the ₹4,500 reserve is protecting, and why the restock decision lands in week 12 rather than week 20.
Where the first 55 orders come from
Own store first, because in a freight-heavy category you cannot afford to give away margin, and because sets are your basket lever and marketplaces sell singles. Instagram and Pinterest are where people find you, and decor is bought with the eyes, so product photography is not a vanity line in that allocation table. It is the ad account.
Marketplaces come second, and the fee position moved in your favour. Amazon India expanded zero referral fees from products under ₹300 to products priced up to ₹1,000, across more than 1,800 categories including home decor and furnishings, effective 16 March 2026, and cut referral fees on high-demand categories above ₹1,000. Check your exact node before you plan around it, because the rate above ₹1,000 still varies by product type. The practical read for a decor founder: single pieces priced at or under ₹1,000 are unusually cheap to list there, which makes Amazon a decent place to move slow designs without discounting on your own store. The gifting season overlay, which is where this category makes its year, is in how to start a gifting brand in India.
The 90-day sequence
Weeks 1 to 3: shortlist 6 to 8 suppliers across two corridor-compatible clusters, order three samples of every candidate design, run the drop test, and get eight-digit HSN codes in writing. Weeks 3 to 5: place both orders at MOQ 12, book one consolidated consignment where the corridor allows, register GST, file the trademark, order two box sizes. Weeks 5 to 7: goods land, count and inspect every piece, shoot the whole range in one styled day, build the store and the Legal Metrology compliant listings. Weeks 7 to 13: sell. Spend the ₹22,000 in controlled increments, watch three numbers only, damage rate, prepaid share and CAC, and restock the three winners in week 12.
Spending the ad money on stock because the supplier offered a better per-piece rate at 25 pieces instead of 12. The maths looks great: ₹250 becomes ₹225, and ₹36,000 now buys 160 pieces instead of 144. What actually happened is that ₹10,000 came out of the ring-fenced ₹22,000, and the founder now has more inventory and no way to find out if anyone wants it. Three months later there is ₹34,000 of stock on a shelf, ₹12,000 of ad learning that ran out before any creative worked, and no answer to the only question ₹1 lakh was supposed to answer. A ₹25 saving on a piece you cannot sell is not a saving. At this tier stock is the cheap part and demand data is the expensive part, so the budget should reflect that.
Execution checklist
- Pick two clusters, and check with a transporter whether they sit on one corridor before you place both orders.
- Buy 12 to 16 designs at the supplier minimum of 12 pieces each. Depth is bought in month three, with sell-through data.
- Run every candidate design through length x breadth x height ÷ 5000 before you order it, and drop the ones that bill at two to four times their weight without earning it.
- Choose two box sizes that cover about 80% of the range with 4 to 5 cm of cushion on all six faces, then reject designs that fit neither.
- Drop test three samples of every design onto concrete from 75 cm, and ship one to another zone as a refused COD so you see a two-journey pack.
- Get the eight-digit HSN in writing on every supplier invoice and verify it against your own article. Never assume one decor rate.
- Check the mixed supply rule before you build a gift set, or your 5% pieces get taxed at your 18% piece's rate.
- Ring-fence the ₹22,000 ad line. No supplier discount is allowed to touch it.
- Model the P&L with a separate damage line at 4% of net revenue, and RTO on the house formula, never one blended fudge.
- Track exactly three numbers weekly for 60 days: damage rate, prepaid share, CAC.
Your next action today
Take the three designs you are most excited about. Get their packed box dimensions from the supplier, in centimetres, including the protection. Run each through the divisor, then subtract that freight number and ₹130 of packaging from the price you were planning to charge. If two of the three still work, you have a range to build. If none of them do, you just saved ₹36,000 and four months, and you now know to hunt for denser pieces instead of prettier ones. Then message six suppliers across two corridor-compatible clusters for samples and eight-digit HSN codes. The frameworks and calculators used through this page come from Ravikant Tyagi's operating system for exactly this stage.
If you'd like the complete execution system, calculators, SOPs, templates and operating frameworks behind this process, continue inside D2C Acquisition.Lab.
