You have a design, a ₹15,600 sampling round behind you, a folder of block-print references, and eleven WhatsApp numbers that all say "manufacturer" in the display name. Nine of them are traders. Ethnic wear punishes that confusion harder than any other apparel category, because your product is not one thing that one factory makes. It is fabric from one city, print or embroidery from a second, stitching in a third, and the person quoting you usually controls only one of the three.
So here is the direct answer first. Five routes make almost all of India's ethnic wear and they are not interchangeable: Delhi NCR for womenswear cut-make-trim and machine embroidery, Jaipur for small-batch cotton and Sanganer block-print job work, Surat for man-made fabric and value kurtis, Lucknow for chikankari hand embroidery, and Varanasi and Kolkata for fabric only. Budget 100 to 300 pieces per design across a size set, ₹3,000 to ₹10,000 for sampling, and a landed cost of 30 to 45% of MRP. Two things then decide whether the money works: how cleanly your size set is graded, and how far ahead you place the order, because festive and wedding demand lands between September and February and hand work does not compress. The business side of the category, budgets, pricing and the revenue ladder, sits in how to start an ethnic wear brand in India. This page is about the factory.
Ethnic wear is made through five routes, not one. Delhi NCR runs womenswear CMT and machine embroidery at 150 to 300 pieces a design, with the deepest tailoring bench for suit sets and festive pieces. Jaipur runs small-batch cotton stitching and Sanganer block-print job work, and will take 100 to 150 when Delhi will not. Surat sells fabric and ready value kurtis in catalogue lots. Lucknow does chikankari by hand through home-based artisans, on 6 to 12 week lead times that move with the season. Varanasi and Kolkata sell cloth, not garments. A CMT rate covers cutting, stitching, finishing and pressing. It almost never covers fabric, lining, the dupatta, trims, print, embroidery, wastage or your labels, so ask for a landed cost per piece and not a make rate. Sample three times, fit then size set then pre-production, at ₹3,000 to ₹10,000 total. Grade the size set properly: eight extra points of returns costs about ₹52 per delivered order on a ₹1,299 basket, ₹15,600 a month at 300 orders. GST on stitched ethnic wear is 5% up to ₹2,500 a piece and 18% above. Fibre, care, size and MRP labelling is mandatory, and Handloom Mark or a GI name goes on the tag only when it is real and documented.
The five routes, and what each one actually gives you
Before you message anyone, decide which route your product belongs to. Buying a printed cotton kurta from a Delhi export house and a georgette suit set from a Jaipur cotton unit are both ways to pay 25% too much and wait three weeks longer than you needed to.
| Route | What it really is | MOQ per design | Realistic lead time | Who buys the fabric |
|---|---|---|---|---|
| Delhi NCR (Okhla, Noida, Gurugram, Faridabad) | Womenswear CMT plus machine embroidery, and the tailoring depth for suit sets, anarkalis and festive pieces | 150 to 300 pcs across a size set | 4 to 7 weeks | You, in most CMT deals |
| Jaipur, with Sanganer and Bagru | Small-batch cotton stitching, plus hand block-print job work done on cloth you supply | 100 to 150 pcs; print job work from 50 to 100 metres a colour | 5 to 8 weeks including printing and washing | You. The print house prints, it does not sell you cloth |
| Surat | Man-made fabric (rayon, georgette, crepe, chiffon) and ready value kurtis in catalogue lots | Fabric from 50 to 100 metres a design; ready kurtis in sets of 4 to 12 | 1 to 3 weeks | Fabric is the product here |
| Lucknow | Chikankari hand embroidery, distributed to home-based artisans through a karkhandar | 50 to 150 pcs, lower on fine work | 6 to 12 weeks, seasonal | You supply cut panels or fabric |
| Varanasi and Kolkata | Brocade, silk, handloom cotton and jacquard. A cloth route, not a garment route | Loom-dependent, often 20 to 50 metres a design | 3 to 8 weeks | You are buying cloth, nothing else |
Most brands end up using two of these, not one. Cotton fabric or block print from Jaipur, stitching in Jaipur for the first 150 pieces, then Delhi NCR once volumes and embroidery demands grow. If you sell sarees alongside stitched pieces, that is a separate fabric-length trade with its own vendors and its own tax treatment, covered in saree wholesale sourcing in India.
One warning that saves money on day one. A Delhi or Mumbai address is not a factory. A large share of listings are trading offices that route your order to a unit in Jaipur or Surat and add 15 to 25%. That is fine if you know it and price it. It is expensive if you think you are factory-direct. The general method for telling a maker from a middleman, the shop-floor video, the GST check, the reference calls, is in how to find manufacturers and suppliers in India.
What a CMT rate includes, and what it quietly excludes
Cut-make-trim means the unit cuts your fabric, stitches the garment, finishes it and hands it back. You own everything else. That single sentence is where most first orders go wrong, because founders hear a make rate and think they have heard a cost.
A CMT quote almost always covers cutting, stitching, thread, overlock and finishing, pressing and poly bagging. It almost never covers: fabric, lining, the dupatta, buttons, hooks, tassels and latkans, your woven label, the care label, the hang tag, printing, embroidery, fabric wastage, inward and outward freight, and job-work GST. On kurta sets the dupatta is the classic surprise, because plenty of stitching units subcontract dupatta cutting and tasselling and quote it separately if you do not ask.
Ask every unit for a landed cost per piece with those lines shown separately. Then settle the one number nobody volunteers: fabric shortfall. Cutting waste is real, and on a directional print it is worse. Agree in writing what percentage of shortfall the unit absorbs before it starts finishing your run with its own cloth, usually 3 to 5% on plain fabric and up to 8% on a large repeat. Put that percentage in the purchase order, because after the fabric is cut it is your word against theirs.
At an MRP of ₹1,299 that landed cost is 34.6% of price, inside the 30 to 45% band this category runs on. Notice that the make rate, the number everyone negotiates hardest, is ₹115 of a ₹450 cost. Shaving ₹10 off it moves your COGS by 2%. Getting the fabric, the print and the wastage right moves it by 20%. Negotiate the MOQ and the fabric terms before you negotiate the make rate; the scripts are in MOQ negotiation with suppliers.
Fabric: whoever buys it carries the risk
There are only two arrangements. You buy the fabric and the unit stitches, or the unit buys everything and quotes a finished piece. Both are fine. Pretending the risk moved when it did not is what hurts.
Buy it yourself and you own shrinkage, shade drift between dye lots, and the day the processor delivers 90 metres of a 100 metre booking. Let the unit buy and you never see the GSM you paid for, so a lighter cloth or a cheaper blend hides inside a finished garment until your customers wash it. For a first or second run, buying your own fabric is usually right, because fabric is the product in this category and you learn more in one trip to a Surat or Jaipur market than in six months of quotes.
Two practical numbers. Mills and processors want 50 to 100 metres per colour on a dye or print run, so a four-colour story is a 200 to 400 metre commitment before a single garment exists. And input costs moved in your favour recently: the September 2025 GST reform cut man-made fibre from 18% to 5% and MMF yarn from 12% to 5%, and raised the 5% readymade garment threshold from ₹1,000 to ₹2,500 a piece, per the Ministry of Textiles. If your quote for Surat rayon or georgette has not moved since 2025, ask why.
Block-print job work: the numbers behind the aesthetic
Hand block printing is job work priced per metre, not per piece, and it has a tooling cost most founders find out about after they have fallen in love with a design.
- Printing rate. Roughly ₹15 to ₹45 a metre for pigment printing at two to four colours. Natural dye work at Bagru costs more because it runs through multiple wash, dry and sun cycles. More colours means more passes, and each pass is a chance to misregister.
- Blocks are tooling. A carved teak block runs about ₹800 to ₹2,500 depending on size and detail, and you need one block per colour per motif. Three blocks is ₹2,400 to ₹7,500 spent before any cloth is printed.
- Amortise over the design's life, not the first order. ₹3,600 of blocks over 200 pieces is ₹18 each. Over 60 pieces it is ₹60 each. That is why a tiny first run in hand block looks expensive per unit, and why founders wrongly blame the printer.
- Variation is the craft, and it is also a return. Slight misregistration, a visible join where the block met, uneven dye pickup. Put one honest line about it on the product page. A customer who was told expects character. A customer who was not files a defect claim.
Chikankari and the lead time you cannot compress
Chikankari is hand embroidery worked by home-based artisans, mostly women, coordinated by a karkhandar who distributes cut panels and collects them back. It is piece work spread across a network, not a shift on a factory line. Two hundred pieces of medium-density work is dozens of hands over six to twelve weeks, and the clock moves with harvest season, exams, weddings in the artisans' own families, and whoever placed a bigger order the week before you.
Plan for that instead of arguing with it. Agree a split delivery in writing, a first tranche and a balance, each with a date. Sample the stitch density and motif coverage, not just the drawing, because "chikankari kurta" covers everything from 40 motifs to 400 and the price gap is three times. And be precise in your copy. Lucknow Chikan Craft is a registered geographical indication, application 119, on the register since December 2008. Machine-embroidered chikan from a Delhi unit is a perfectly good product sold honestly as machine chikan. Calling it Lucknowi chikankari when it was never near Lucknow is a false claim on a protected name.
The festive calendar, worked backwards
September to February makes the year in this category. Everything upstream has to be scheduled against that, and the founder who buys stock in March is financing it for six months before the demand shows up.
| Milestone | Weeks before the drop | What actually happens |
|---|---|---|
| Design freeze, blocks carved | 22 to 24 | Styles locked, blocks or embroidery artwork commissioned, fabric shortlisted |
| Fit sample | 20 to 22 | One garment in your base size. Pattern, drape and hand-feel decided here |
| Fabric booked and delivered | 16 to 18 | Cash leaves. Dye and print runs booked per colour, shrinkage tested |
| Size-set sample measured | 14 to 16 | One piece in every size, measured against your chart before anything scales |
| PP sample locked, bulk starts | 12 to 14 | Golden reference photographed and agreed. Chikankari needs the full window from here |
| Bulk delivered and inspected | 4 to 6 | Full-lot QC, measurement audit, rejects returned while the balance payment is still unpaid |
| Shoot, size charts, listings live | 2 | Fit-on-body content, published measurements, inventory loaded |
Printed cotton can be squeezed into 14 to 16 weeks. Hand chikankari and heavy embroidery cannot, and a unit that promises otherwise in July is either lying or planning to deprioritise you the moment a bigger order lands. The cash consequence deserves a line in your model: money leaves in May for a September drop and returns in October and November. Six months of working capital on a category people casually call fast fashion. Build that into your cash flow model, plan the drop itself with the festive sale playbook, and hold the size curve honestly using inventory management for D2C.
Inventory Confidence Model™: buy depth only where you have evidence, and buy the calendar backwards from the demand window, not forwards from your enthusiasm. For ethnic wear that means one hero style in real depth, two supporting styles at minimum quantity, and the size curve weighted to what your last 100 orders actually bought rather than an even split across S to XXL. Anything you cannot defend with a number from last season goes in at MOQ, not at a discount-driven quantity.
Qualifying a unit before you pay anything
Generic supplier vetting applies here, and it is written up properly on the apparel manufacturing page. These are the ethnic-specific questions that separate a real unit from a confident one.
- "What did you make last season, and for whom?" Ask for a style they ran in volume, then ask to see it in three sizes. A unit that can produce the same garment in S, L and XXL from stock is telling you it grades properly.
- "How many embroidered pieces a week, at this density?" Capacity in pieces per week at your specification, not "we can do anything." Anyone who answers instantly and without asking about density is guessing.
- "Who finishes the dupatta?" If it is subcontracted, you have a second vendor in your chain that you have never met and cannot chase.
- "What happens if the fabric falls short?" The honest answer names a percentage and an escalation call. The bad answer is "don't worry."
- "What else is on your line in August?" Peak season is when small orders get pushed back. A unit that admits it is loaded and offers you an earlier slot is worth more than one that says yes to everything.
Supplier Scorecard™, weighted for ethnic wear: legitimacy (15%), GST certificate, Udyam or factory registration, and a current account in the same legal name. Grading across the full size run (30%), the heaviest weight in this category because a bad grade multiplies returns across every size. Fabric and finish evidence (20%), shrinkage, colour fastness and GSM in writing. Hand-work quality (15%), stitch density, embroidery pull, print registration. Calendar honesty (10%), what they are running in peak season and whether they will commit to dates. References you can call (10%). No purchase order below a weighted 7.
The four QC tests that decide your return rate
Run these on the pre-production sample and again on a random 10 pieces from the bulk lot, while the balance payment is still in your hands.
- Colour fastness. Rub a damp white cloth across the fabric, hard, and look at the cloth. Then wash the sample with a white swatch in the same bucket. Dark indigo and natural-dye Bagru work bleed the most, and a customer whose kurta ruined a white dupatta is a refund plus a public review.
- Shrinkage. Wash twice per the care label and measure. Cotton mulmul and cambric can move 3 to 7% if the fabric was not pre-shrunk. Either buy pre-shrunk cloth or make the unit cut with a shrinkage allowance in the pattern. Get which one, in writing, before cutting.
- Stitching density. Count stitches per inch at the armhole and the side seam, the two places a kurta splits. Cotton ethnic wear wants roughly 10 to 12 SPI. Loose stitching photographs badly in a one-star review.
- Embroidery pull. Hold the panel to the light. Puckering around a motif means thread tension is dragging the fabric. Wash the sample and look again, because pull often appears only after the first wash, which is exactly when your customer sees it.
Add two lot-level checks: dupatta length measured across ten pieces, because it drifts, and print placement at the yoke, because a motif landing in the wrong place turns a hero style into a discount rack.
The size set is where returns are decided
This category carries the apparel returns band, 30 to 50% unmanaged, and grading is the manufacturing decision that moves it. Approve bulk off one perfect medium and you are gambling on sizes nobody measured.
Put a number on that gamble. House convention: RTO and return drag per delivered order is rate divided by one minus rate, multiplied by forward shipping plus reverse plus packaging plus the burnt CAC. On a ₹1,299 blended basket, one piece per order, with ₹70 forward, ₹70 reverse, ₹35 packaging and ₹200 of burnt acquisition cost, each failed parcel costs ₹375. At a managed 20% rate you carry 0.25 failed parcels per delivered order, which is ₹94 of drag. Let a badly graded XL push you to 28% and you carry 0.389 failed parcels, which is ₹146. That gap is ₹52 on every delivered order, ₹15,600 a month at 300 orders, on a style whose entire print bill was ₹91 a piece.
A full size-set sample costs ₹3,000 to ₹10,000 once. It pays for itself inside the first week of a decent month. The operational half of this, prepaid nudges, address checks and confirmation calls, sits in reducing RTO on COD orders.
In my supply chain years I learned to distrust any vendor conversation that stayed friendly. Ethnic wear is worse than most, because the relationship is warm, the samples are beautiful, and the hard questions feel rude. So I ask them early and in writing, when nobody has money at stake yet. What is the delivery date, what happens on a split delivery, what percentage of fabric shortfall do you absorb, and who pays to redo a lot that fails the wash test. A unit that answers all four without flinching has run real orders. A unit that says we will manage is telling you it plans to manage using your money. Ask in June, not in September. ₹52 of block-print job work on a ₹1,299 kurta is the cheapest differentiation available in apparel, and the one traders will quietly skip.
What to get in writing before the advance
Before any advance leaves your account, one message must carry all of it: approved PP sample photo with fabric, shade code and full measurement chart, the size-wise quantity split, the tolerance per measurement point (0.5 to 1 cm), the delivery date with the split-delivery tranches named, the fabric shortfall percentage the unit absorbs, who pays rework on a failed wash or fastness test, and the payment split, 30 to 40% advance with the balance against dispatch proof, paid to a current account matching the GST legal name. Screenshot it. That message is the only thing you can hold up in November.
Labels, Handloom Mark and GI claims
Your unit stitches the garment. The tag is your legal responsibility, and marketplaces delist listings that get it wrong.
- Mandatory on every piece. Honest fibre composition, care instructions, size, MRP inclusive of taxes, and your name and address as manufacturer or marketer with a consumer-care contact under Legal Metrology. Confirm at sampling that the unit can attach all of it, because retrofitting labels onto finished goods costs money and weeks.
- GST. 5% up to ₹2,500 per piece, 18% above, effective 22 September 2025. That cliff is real in this category because occasion wear crosses it. A ₹2,499 kurta set and a ₹2,600 one are thirteen tax points apart, so price your festive hero deliberately rather than by accident. Registration and filing basics are in GST for ecommerce sellers.
- Handloom Mark. It certifies a product was actually woven on a handloom, is implemented by the Textiles Committee, and needs registration with an onsite check before you get labels, which cost 50 paise each, per the Textiles Committee scheme. Right now the scheme is voluntary. That is set to change on 1 October 2026, the entry-into-force date India gave the WTO for the Handloom Mark Standards (Labelling and Display) Regulations, 2025, which make the label compulsory on anything sold as handloom. Read who the rule binds, because founders get it backwards. The weaver or producer registers and buys the labels, not you. Your job as the brand is to make sure every handloom piece you ship already carries a valid label and traces back to a loom on paper. That date lands mid-season, so settle it with your supplier before the festive buy, and check the final gazette notification before you print tags.
- GI names. Sanganeri hand block print and Lucknow Chikan Craft are registered geographical indications. They describe where and how something was made, not a style you can borrow. Use them only when your supplier can document origin, and never as a design adjective.
Paying a 50% advance in July for a festive chikankari run with no written delivery date, no split-delivery clause and no size-set sample. It happens because the samples are gorgeous and the karkhandar is likeable. The founder sends ₹93,000 against 300 pieces landing at ₹620 each, plans to sell the lot at ₹1,999 through October and November, and hears "do not worry, madam" every time she asks about dates. On 20 September, 120 pieces arrive. The balance of 180 lands on 12 December, well after the window she bought for. Those 180 clear in January at ₹1,199, so ₹3.6 lakh of planned revenue becomes ₹2.16 lakh. ₹1.44 lakh gone, on stock that was already paid for and already correct. Two free decisions caused it. She started a chikankari run in July, when the calendar above puts bulk start at 12 to 14 weeks, meaning early June for a September drop, and she never got a date in writing. So the fix is both halves: place the embroidery order by early June, then send one WhatsApp message before the advance naming tranches, dates and who eats the rework. It costs nothing and it is the single highest-return thing you will do all season.
Execution checklist
- Pick your route before you message anyone: Delhi NCR for embroidery and suit sets, Jaipur for small-batch cotton and block print, Surat for man-made fabric and value kurtis, Lucknow for real chikankari.
- Ask every unit for a landed cost per piece with fabric, print or embroidery, trims, labels, wastage and freight shown as separate lines, not a make rate.
- Agree the fabric shortfall percentage the unit absorbs, 3 to 5% on plain cloth, before you send a single metre.
- Sample three times, fit then full size set then pre-production from the bulk fabric, and budget ₹3,000 to ₹10,000 for it.
- Measure every size against your own chart within a written tolerance of 0.5 to 1 cm per point. Never approve bulk off a medium.
- Run all four QC tests on the PP sample and on 10 random bulk pieces: rub and wash fastness, twice-washed shrinkage, stitches per inch at the armhole, embroidery pull held to the light.
- Count your festive drop backwards: 22 to 24 weeks for embroidered pieces, 14 to 16 for printed cotton, and put the fabric payment date in your cash flow.
- Get the delivery date, split-delivery tranches, tolerance, shortfall rule and rework liability in one written message before any advance, paid to a current account in the GST legal name.
- Confirm the unit can attach compliant labels: fibre, care, size, MRP and your marketer details, plus a batch reference so a bad lot is traceable.
- Use a GI name only when your supplier can document origin. Handloom is its own rule now: the label comes from the registered weaver or producer, and from 1 October 2026 you have to make sure every handloom piece you sell already carries one. Price the festive hero deliberately around the ₹2,500 GST line.
Your next action
Today, send one message to five units, two in Jaipur, two in Delhi NCR, one in Lucknow if hand work is in your plan. Ask each for four things: a landed cost per piece on your one hero style at 100, 200 and 300 pieces with every line shown separately, their MOQ across a full size set, a live video from their floor, and the earliest date they will commit to in writing for a September delivery. The quotes cost nothing and arrive inside 48 hours, and the gap between the five replies will teach you more about this category than another week of scrolling catalogues. The frameworks here, the Supplier Scorecard™ and the Inventory Confidence Model™, come from Ravikant Tyagi's operating system for founders doing exactly this.
If you'd like the complete execution system, calculators, SOPs, templates and operating frameworks behind this process, continue inside D2C Acquisition.Lab.
