You have a tech pack, a fabric idea, and a WhatsApp list of ten "manufacturers" who are mostly traders with a nice catalogue. The gap between that list and a factory you can actually hold accountable is where most first-time clothing brands quietly lose money. Not on ads. On the 300-piece run where the XL was graded wrong, or the reorder that landed two shades off the approved sample while a wave of size-and-fit returns eats the profit of the batch that did sell. Apparel already carries the highest return rate in Indian ecommerce, so a sloppy manufacturer turns that from a problem into a wound.
This guide resolves one thing end to end: how to find and vet a clothing manufacturer in India you will not regret paying. The three real manufacturing models and which fits you, the cluster map of where knits, woven, winterwear and ethnic actually get made, how to reach real factories instead of trading offices, the vetting checklist that separates a maker from a middleman, the three-sample protocol before any bulk, the MOQ maths for a size set, and the label rules your unit must support. It comes from Ravikant Tyagi, who led supply chain and operations at Atomberg through its ₹400cr to ₹1,200cr phase and now works with early D2C brands as a fractional COO. The whole business side, the budgets, pricing, returns economics and revenue ladder, lives in the flagship on how to start a clothing brand in India. Read that for the business. Read this for the factory.
Clothing gets made three ways. Cut-make-trim (CMT), where you supply the fabric and the unit only stitches; private label, where you brand and tweak a maker's ready blanks or base; and full-package or FOB, where the unit sources everything and delivers finished goods to your tech pack. CMT gives the most control and lowest making cost but the most work; full-package is the least work and the highest MOQ. Tirupur is the hub for knits and T-shirts, Ludhiana for winterwear and knits, Delhi NCR for woven, womenswear and ethnic, Jaipur for block-print, Mumbai and Kolkata for woven and finishing. Find real factories through the Tiruppur Exporters Association and exporter directories, apparel trade bodies, a filtered IndiaMART search, a vetted sourcing agent, and referral chains, never a single Google result. Vet on GST plus factory registration, sample quality and consistency across a full size run, GSM, shrinkage and colour-fastness test reports, stitching SPI, and real buyer references. Sample in three stages, fit then size-set then pre-production, with a wash test and a measurement tolerance agreed in writing. Negotiate the MOQ per design across the size set and start narrow. The killer variable is size and fit returns, so the unit must hold measurements and shade batch to batch. On labels the maker must print fibre content, care, size, country of origin and MRP; GST on readymade garments is 5% up to ₹2,500 a piece and 18% above.
The three ways a garment gets made, and which one is yours
Every clothing brand is manufactured through one of three models. They differ in who buys the fabric, how much the making costs, the MOQ you must accept, and how much of the operation lands on your plate. Pick the wrong model for your stage and you either drown in fabric sourcing you were not ready for, or you overpay a full-package unit for control you did not need.
| Model | What it is | Who buys the fabric | Typical MOQ / design | Cost and control | Best for |
|---|---|---|---|---|---|
| CMT (cut-make-trim) | You hand over fabric and trims, or exact specs, and the unit only cuts, stitches and finishes | You do | 50 to 300 pcs knits; woven higher | Lowest making cost, most control, most work on your side | Founders who know fabric and want a specific hand-feel and fit |
| Private label (job-work) | You brand, print and lightly tweak the unit's ready blanks or standard base garment | The unit | Blanks 50 to 100 pcs; base runs higher | Fast, low hassle, least differentiation | First launch, testing designs before committing to fabric |
| Full-package / FOB | The unit sources fabric and trims and makes to your tech pack, delivering finished goods | The unit | 300 to 500+ pcs per style and colour | Highest per-piece cost, least operational load, hardest to vet | Funded brands scaling proven styles from a tech pack |
Read the honest version. CMT is where serious apparel brands end up, because fabric is the product and CMT is the only model where you own the fabric decision. It is also the most work: you have to source and inspect greige or finished fabric, manage dyeing and shrinkage, and carry that risk yourself. Private label is the softest start, a branded version of a base a hundred others also sell, fine for validating designs, weak as a moat. Full-package looks tempting because the unit does everything, but that is exactly why you must vet it hardest: you never see the fabric buy, so a unit cutting corners on GSM or shade hides it inside a finished garment. The full logic of these models across every category is in white label vs private label vs OEM in India.
Founder Decision Loop™: signal, smallest honest test, hard read of the numbers, then commit capital. Applied to the manufacturing model, the model follows your proof and your fabric knowledge, not your ambition. New brand with unproven designs and no fabric expertise starts private label or blanks-plus-print. A design that sells to strangers at an affordable CAC graduates to a small CMT run where you control the fabric. Full-package is a scaling tool for a proven style, not a starting tool. Committing to a 500-piece full-package order for a design the market has not approved is the most expensive way to learn nobody wanted it.
Where clothing is actually made: the cluster map
India's garment industry runs on clusters, and each cluster has a speciality built over decades: the yarn, the dyeing houses, the trims vendors, the stitching units and the labs all sit within a few kilometres of each other. Buy the right product from the wrong cluster and you pay 20 to 30% more and wait weeks longer. Here is the map that matters and what each place is genuinely good at.
| Cluster | Where | Known for |
|---|---|---|
| Tirupur | Tamil Nadu | The hub for knits: T-shirts, polos, oversized tees, sweatshirts, joggers. The full knit supply chain in one city, which is why a Tirupur order moves in weeks while a fragmented one takes months |
| Ludhiana | Punjab | Winterwear and knits: sweaters, sweatshirts, thermals, hosiery. The default for anything wool or warm |
| Delhi NCR | Delhi, Noida, Gurugram, Faridabad | Woven and womenswear: shirts, dresses, ethnic and fusion, plus small-run stitching units and export-surplus lots. The strongest base for woven and embroidered pieces |
| Jaipur | Rajasthan | Block-print, hand-work, cotton kurtas and co-ord sets, women's ethnic-fusion. The place for a print-led or craft-led brand |
| Mumbai | Maharashtra | Fast-fashion woven, finishing and sampling, close to design and agency ecosystems. Higher overheads reflected in quotes |
| Kolkata | West Bengal | Woven shirts, cotton, hosiery and lower-cost stitching labour for basics |
Tirupur earns its reputation. India's knitwear capital shipped about US$4.46 billion of garments in 2025-26, and the Tiruppur Exporters Association alone counts roughly 1,360 knitwear exporting units among its members, with several thousand more production units across the wider cluster. That density is your advantage as a small brand: one trip puts a dozen units that compete with each other in front of you, and the yarn, dyeing, printing, labels and packaging you need are all within the same city.
One distinction that saves confusion: if you are making sarees, that is a separate fabric-length trade in Surat and Varanasi with its own GST treatment, not this readymade knit-and-woven world. This guide is about stitched apparel. The other thing worth knowing is that a city address is not a factory. Plenty of listings are trading offices in Delhi or Mumbai that route your order to a Tirupur or Ludhiana unit and add a margin. Fine if you know it and price it, a problem if you are paying trader markup thinking you are factory-direct. The full method for reaching these units and reading them honestly is in how to find manufacturers and suppliers in India.
How to actually find them, not just IndiaMART roulette
Typing "tshirt manufacturer" into Google and messaging the top three is how founders end up with traders. Real factories are found through five channels, and using more than one is how you triangulate who is genuine.
- Exporter directories and trade associations. The Tiruppur Exporters Association and the Apparel Export Promotion Council publish member lists of registered manufacturing units. An exporter that survives foreign buyer audits usually runs a tighter shop than a domestic-only trader. Trade media like Apparel Resources also names real units and the state of each cluster.
- IndiaMART, filtered properly. IndiaMART is useful only if you filter for verified sellers, ask every listing point-blank "is this your own plant and where is it," and reject anyone who cannot send a live video from a shop floor. Treat the first quote as a starting screen, not a shortlist.
- Sourcing agents. A good agent in Tirupur or Delhi charges 5 to 10% and earns it for a first-timer by opening doors, checking samples and managing the local relationship in the local language. Vet the agent like a supplier: ask which brands they source for and call one.
- Referral chains. The fastest route to a good unit is a non-competing brand that already uses one. Fabric wholesalers and trims vendors also know exactly which stitching units deliver clean work and which do not, because they see every reorder. Ask them.
- Trade fairs. Events like the Tirupur and Delhi garment fairs put dozens of units in one hall where you can feel the fabric and read the operator in person, which a WhatsApp catalogue never lets you do.
The vetting checklist: the Supplier Scorecard
By now you should have three to five units that quoted your spec. Do not pick the cheapest quote or the friendliest sales rep. Apparel has failure modes that generic sourcing does not: a fabric that shrinks two sizes after one wash, a shade that drifts between batches, an XL graded so badly it poisons your reviews. The scorecard tests what actually breaks garments.
Supplier Scorecard™ for apparel: score every shortlisted unit 1 to 10 on six weighted criteria and refuse a purchase order below a weighted 7. Legitimacy (15%): valid GST, a factory registration or Udyam, and a bank account in the entity's legal name. Sample quality and grading (25%): does the fit sample match spec, and is the whole size run graded cleanly, not just the M. Fabric test reports (20%): GSM, shrinkage percentage and colour-fastness on paper, not on faith. Stitching quality (15%): SPI, seam strength and finish, checked on the sample. References (15%): two current buyers you can actually call. Communication (10%): speed and honesty on hard questions. The cheapest unit wins on exactly one line of this.
What each line means when you are holding the garment and the paperwork:
- Legitimacy. Ask for the GST certificate and confirm the legal name matches the bank account you will pay. A registered factory has a GST number, a registration or Udyam MSME record, and no problem showing them. A trader dodging tax will dodge accountability too.
- Sample quality across the size run. This is the apparel-specific check founders skip. Approve bulk off one M sample and you are gambling that the XS and XXL were graded well. They often are not, because grading is where cheap units cut corners. Insist on seeing the full size set stitched, then measure every size against the chart yourself.
- Fabric test reports. GSM tells you the fabric weight you are paying for; shrinkage tells you how many returns you are buying; colour-fastness tells you whether the shade survives washing and rubbing. A real unit produces these numbers, in-house or from a lab. "Trust me, the fabric is good" is not a test report.
- Stitching quality. Turn the sample inside out. Count the stitches per inch, pull the seams, check the finish at the collar and hem. Loose SPI and puckered seams are what customers photograph in one-star reviews.
- References and honesty. Ask for two brands they currently make for, and call one. Then ask the unit a hard question about a past defect and judge the honesty of the answer, because you will need this unit at 9 pm when a batch is late and your ads are live.
A quick reminder that fabric choice is a real moat, not a detail. Bummer, the Ahmedabad innerwear brand that hit ₹10.7 crore in FY25, built its whole pitch on one fabric decision, MicroModal, a beechwood-derived yarn softer than cotton. A distinctive fabric only works if the manufacturer can source it and hold it consistent across every reorder, which is exactly what the scorecard is checking for.
MOQ and the size-set trap
MOQ in apparel is not one number, it is one number multiplied by your size curve. A "300-piece" CMT order across five sizes and two colours is really ten stock-keeping lines, and the ends of the curve, XS and XXL, sell slowest, so a naive spread leaves 15 to 20% of the run near-dead on arrival. This is where founders sink cash chasing a per-piece discount on stock the market has not approved.
| Product | Typical CMT MOQ / design | Indicative making + landed cost | Typical MRP |
|---|---|---|---|
| Basic knit T-shirt, 180 GSM | 50 to 150 pcs | ₹120 to 200 landed | ₹399 to 699 |
| Heavyweight / oversized tee, 220 to 240 GSM | 100 to 300 pcs | ₹180 to 280 landed | ₹699 to 999 |
| Woven shirt | 150 to 300 pcs | ₹250 to 450 landed | ₹899 to 1,499 |
| Cotton kurti / ethnic co-ord | 50 to 200 pcs | ₹300 to 600 landed | ₹899 to 1,999 |
The move that protects your cash is to start narrow. Fewer sizes, one or two colours, one hero style, and negotiate the MOQ before you negotiate the rate. A unit that will run 100 instead of 300 for a new brand is worth more than one that shaves ₹15 off a batch you cannot sell, even if you pay ₹20 to 40 more per piece for the smaller run. That premium is cheaper than owning dead stock. The scripts for that conversation are in the MOQ negotiation guide, and where your ₹699 or ₹999 price actually comes from is in the product pricing guide.
In my supply chain years, the number I watched in every review was dead stock, and apparel is the category most eager to create it. The trap is always the size curve. A founder proudly tells me they got the per-piece cost down by ordering 400 instead of 150, then two months later I am looking at 70 unsold XS and XXL pieces that will move only at 50% off. The fix is unglamorous: order your first run in a narrow size band, sell it, read which sizes actually move for your customer, and only then widen the curve on the reorder. A slightly higher cost per piece on a run that sells clean beats a lower cost per piece on a run that clogs your godown. Cheap inventory is still dead inventory if it does not sell.
The sampling protocol: three samples before any bulk
Sampling is the cheapest insurance in apparel and the step excited founders rush. Done properly it is three samples in sequence, not one, and it costs ₹3,000 to 10,000 across your shortlist. Skipped, it costs you a whole run and a launch.
- Fit sample first. One garment in your base size to check the pattern, fit and hand-feel. This is where you decide the fabric and silhouette are right before anyone cuts a bulk lot.
- Size-set sample next. One piece in every size you plan to sell, so you can confirm the grading is clean end to end. Measure each against your chart. This single step prevents the badly-graded-XL disaster that quietly wrecks returns.
- Pre-production (PP) sample last. The final sign-off piece made from the actual bulk fabric and trims, on the actual line, before the run starts. Approve this, photograph it, and lock it as the golden reference in writing.
Test every sample the same way, and for apparel that means the lab-at-home routine: wash it twice per the care label and measure shrinkage, rub a damp white cloth on the fabric to check colour-fastness, and confirm each measurement sits inside an agreed tolerance, usually plus or minus half a centimetre to one centimetre per point. Agree that tolerance in writing before the order, because "close enough" is an argument you will lose after the batch is stitched.
Sample three units against one tech pack, never a vague "make me a nice tee." Run fit, then size-set, then pre-production, and at each stage wash twice to measure shrinkage, rub-test for colour-fastness, and check every size against the chart within an agreed tolerance. Get GSM, shrinkage and colour-fastness in a written report, not a verbal promise. Lock the approved PP sample as a golden reference in writing, with photos, fabric, GSM, shade code and measurements, before any advance. That message is your proof the day a reorder drifts and the unit says "this is normal."
Red flags: when to walk away from a garment unit
Some signals are not negotiation points, they are exits. Apparel has a few that are specific to the category, on top of the usual supplier red flags.
| Red flag | What it actually means |
|---|---|
| Will not send a live video from the shop floor | A trader routing your order to someone else's plant, not a factory you can hold accountable. |
| No GSM, shrinkage or colour-fastness reports | They do not test, or do not want you to see the numbers. Your returns become the test data. |
| Approves bulk off a single M sample | No sampling discipline. The size run will be graded by guesswork and your XL and XS will come back. |
| Shade or GSM varies between two samples | No process control. Your reorders will not match your first batch, and customers notice shade before anything else. |
| Quote 30 to 40% below every other unit | Lighter GSM, a cheaper blend passed off as your fabric, or a bait rate that climbs after your advance. |
| Payment demanded to a personal UPI or savings account | No registered entity behind the money. Pay only a current account matching the GST legal name. |
| Cannot name a single current buyer | Either brand new or hiding a churn of unhappy clients. Ask for two references and call one. |
The shade-and-GSM red flag deserves its own warning, because it is the one that turns a healthy brand into a returns machine. In apparel your customer judges the reorder against the first piece they loved. If batch three is a lighter fabric or a slightly bluer black, that is a return and a lost repeat customer, and repeat is where apparel money is actually made. A unit that cannot hold shade and weight batch to batch is not a supplier, it is a recurring liability. The full method for reading a supplier honestly, including the factory-video audit and the GST check, is in how to find manufacturers and suppliers in India.
The labels your manufacturer must be able to print
Your unit stitches the garment, but the label is your legal responsibility as the brand, and marketplaces delist non-compliant listings without warning. Textile labelling and the Legal Metrology Packaged Commodities Rules both apply. Confirm at sampling that the maker can print or attach all of this, because retrofitting labels on a finished run is a cost and a delay you set on fire for nothing.
- Fibre content. The honest composition, "100% cotton" or "60% cotton 40% polyester." Claiming pure cotton on a blend is the kind of shortcut that becomes a marketplace delisting.
- Care instructions. Wash, dry and iron guidance. This is not decoration, it directly reduces returns from shrunk or damaged garments.
- Size. Clearly marked, and matched to the size chart on your product page.
- Country of origin, MRP inclusive of taxes, net presence, and your name and address as the manufacturer or marketer, plus a consumer-care contact, all as required under Legal Metrology.
- Batch or lot reference so a quality issue can be traced to the run that caused it.
On tax, do not quote a single flat rate. GST on readymade garments after the September 2025 reform is 5% up to ₹2,500 a piece and 18% above ₹2,500, effective 22 September 2025. Price your premium co-ords and jackets with that cliff in mind, because a ₹2,600 tag and a ₹2,400 tag sit in different tax worlds. GST registration itself is mandatory to sell on marketplaces, covered in GST for ecommerce sellers in India, and you should file your brand-name trademark early, because working apparel names get copied within months.
Approving a 300-piece run off one perfect medium sample. The founder pays a 40% advance after the M fits beautifully, skips the size-set sample to save two weeks, and takes the unit's word that the rest of the curve is graded to match. The batch lands and the L is fine, but the XL is cut with the wrong armhole and the S runs short, because grading was the corner the cheap unit cut. Within three weeks the XL and S buyers are posting one-star reviews about fit and returning in bulk, and in apparel one badly graded size drags the returns rate and the review score of the entire style. The salvage is discounting the bad sizes to clear cash, so a ₹1 lakh order returns as ₹65,000 with a bruised listing. The fix cost ₹3,000 and one week: a full size-set sample, measured against the chart before the advance. That is the difference between a supplier you can hold accountable and a run you have to apologise for.
Returns are a manufacturing decision too
It is tempting to file returns under marketing, but in apparel they start on the cutting table. Fashion returns run 30 to 50% of orders when sizing is loose, the highest of any D2C category, and the two biggest causes, fit and fabric-not-as-shown, are both set by your manufacturer. A unit that grades cleanly, holds shade to the sample, and hits the GSM you paid for is directly lowering your return rate before a single ad runs. That is why the vetting above is not paperwork, it is profit. The operational side of returns, prepaid nudges, address checks and exchange-first flows, is in the RTO reduction guide and the returns and reverse logistics guide, and choosing the courier that handles your reverse pickups well is covered in Shiprocket vs NimbusPost vs Delhivery. For a clothing brand those are required reading, not optional.
Execution checklist
- Pick your manufacturing model honestly: private label or blanks-plus-print to validate designs, CMT once a design sells and you want fabric control, full-package only for a proven style at scale.
- Shortlist three to five real factories through the Tiruppur Exporters Association, apparel trade bodies, a filtered IndiaMART search, a vetted agent and referrals, matched to your cluster, Tirupur for knits, Ludhiana for winterwear, Delhi NCR for woven and ethnic.
- Ask every unit for a live shop-floor video, the GST certificate and a factory registration, and confirm the legal name matches the bank account.
- Get GSM, shrinkage and colour-fastness test reports in writing before any advance.
- Sample in three stages: fit, then a full size-set, then a pre-production piece from the actual bulk fabric.
- Wash every sample twice to check shrinkage, rub-test for colour-fastness, and measure every size against your chart within an agreed tolerance.
- Score each finalist on the Supplier Scorecard™: legitimacy, sample and grading quality, fabric reports, stitching SPI, references, communication. Below a weighted 7, no purchase order.
- Negotiate the MOQ before the rate; start with a narrow size band and one or two colours, and never buy a bigger run for a per-piece discount before sell-through is proven.
- Pay a 30 to 40% advance to a current account in the GST legal name, balance against dispatch proof, and lock the golden PP sample in writing with fabric, GSM, shade code and measurements.
- Confirm the unit can print compliant labels: fibre content, care, size, country of origin, MRP, batch, and your marketer details, and price premium pieces around the ₹2,500 GST cliff.
Your next action
Today, one concrete thing: message five units, most likely in Tirupur if you are starting with knits, and ask each for a quote on your one hero style at 100, 200 and 300 pieces, a live video from their floor, their GST number, and whether they run CMT or full-package. The quotes and videos are free and arrive inside 48 hours, and they turn this guide from reading into a real shortlist with a real cost per piece. The sampling, the scorecard and the labels all sequence behind those five messages. Once you have a maker you trust, the business side, the budgets, the pricing, the returns economics and the revenue ladder, lives in the flagship on how to start a clothing brand in India, and if your niche is ethnic wear the specifics are in how to start an ethnic wear brand in India.
If you'd like the complete execution system, calculators, SOPs, templates and operating frameworks behind this process, continue inside D2C Acquisition.Lab.
