The first production run usually starts with a referral: a friend of a friend runs a stitching unit in Tirupur, or a filling line in Ahmedabad, or a casting workshop in Jaipur. You send a sample, agree a price over chai, and three weeks later boxes arrive. Sometimes they're perfect. Sometimes 40 of 300 units have a crooked label, the invoice doesn't match the PO, and the unit owner has stopped replying because Diwali orders came in.
Contract manufacturing in India is how most D2C brands get built — the ecosystem is deep, MOQs are humane compared to China, and being in the same time zone as your factory is worth more than founders expect. But the gap between a good and bad manufacturing relationship is rarely about the unit's skill. It's about process: how you vet, how you order, how you inspect, and how you communicate. This guide covers each, with the numbers.
Key takeaways
- Vet with paid samples and reference calls, not factory tours alone — and test capacity honesty by asking what they'd refuse.
- Nearly every MOQ is negotiable; the lever is reducing the unit's changeover cost, not pleading.
- Budget lead times by category — 15–30 days for apparel, 30–45 for beauty — and measure actuals, because quoted and delivered are different numbers.
- Inspect inline, not just at delivery; a simple AQL sampling plan beats checking every unit or none.
- Move orders out of WhatsApp into written POs and work orders — most "quality problems" are actually specification problems.
Finding and vetting a unit
Directories, trade fairs (IIGF for apparel, Cosmoprof India for beauty), and founder referrals will each surface candidates. The vetting is what separates outcomes.
Pay for samples — plural. One sample tells you what the unit's best worker can do once. Ask for three to five units of the same spec, then measure the variation between them. Consistency across samples predicts production quality far better than a single showpiece does.
Call two current clients. Ask three questions: Do deliveries land on the promised date? What happened the last time a batch had defects? Have prices moved mid-order? A unit that won't share references is answering the question a different way.
Test capacity honesty. Every unit says yes to volume. Instead ask: "What order size would you turn down?" and "Who are your biggest three clients and what share of your line do they take?" A unit running at 90% capacity for a large export client will build your order in the gaps — and your delivery date will be the first casualty of their busy season.
MOQs, and how to actually negotiate them
Minimum order quantities exist because changeovers cost the unit real money: line setup, material sourcing below wholesale breaks, and the opportunity cost of pausing a bigger client's run. Understand that and negotiation becomes engineering rather than haggling.
- Consolidate variants. An MOQ of 300 is often per fabric, not per style. Three styles cut from the same cloth may qualify together.
- Accept their stock inputs. Using the unit's in-stock fabric, bottles, or findings can halve the minimum, because their sourcing risk disappears.
- Trade commitment for size. "150 units now, and a standing reorder every 6 weeks for two quarters" is more valuable to a unit than a one-off 500 — and they'll price it that way.
- Pay a small-batch premium knowingly. A 10–15% surcharge on a half-MOQ trial run is cheap insurance against 500 units of an unproven product.
Contract manufacturing lead times in India, by category
Quoted lead times are opening bids. Typical ranges, assuming materials are available and specs are frozen:
| Category | Typical lead time | What stretches it |
|---|---|---|
| Apparel (cut & sew) | 15–30 days | Fabric dyeing, embroidery/print jobs, festival season |
| Beauty & personal care | 30–45 days | Component MOQs, filling slots, stability and compliance checks |
| Jewellery | 20–35 days | Casting queues, plating, stone availability |
| Footwear | 30–45 days | Last and mould setup, sole outsourcing |
| Home & lifestyle | 25–40 days | Artisan capacity, kiln/finish cycles, monsoon drying delays |
| Food & beverage | 20–30 days | Ingredient seasonality, FSSAI batch testing, packaging print runs |
Treat these as planning ranges, not promises — then measure your own actuals. After three or four orders you'll know that your unit's quoted 21 days is really 27, and 27 is the number your reorder math should use. Tracking quoted-versus-actual per supplier is exactly what supplier lead time and OTD tracking automates, and it feeds directly into avoiding stockouts: a lead time you underestimate by six days is six days of lost sales, every cycle.
Quality control: inline beats final
There are two places to inspect: during production (inline) and before dispatch (final). Final-only inspection finds problems when they're most expensive — the whole batch is made, your launch date is committed, and your only options are accept, rework, or reject. Inline inspection — a check after the first 10% of units are complete — catches systematic errors while the unit can still correct course cheaply. If you can only do one, do inline on new products and final on repeats.
You don't need to inspect every unit; you need a sampling plan. AQL (Acceptable Quality Limit) is the standard way to do this: for a 500-unit order at general inspection level II, you'd inspect a random sample of 50 and, at AQL 2.5 for major defects, accept the batch if 3 or fewer fail and reject at 4 or more. The specific numbers come from published AQL tables; the point is that sample size and pass/fail thresholds are agreed with the unit before production, in writing, so "acceptable quality" stops being a mood and becomes a number. Third-party inspection services will run an AQL check in most manufacturing clusters for a modest fee — worth it for the first few orders with any new unit.
Payment terms
The near-universal norm with small and mid-size Indian units is an advance to start and the balance against delivery — commonly 30–50% up front, with 50/50 typical for new relationships. The advance covers the unit's material purchases; expecting zero advance from a small unit is expecting them to finance your inventory.
What improves with trust is the back half. After a few clean cycles, push the balance from "against delivery" to 15 or 30 days after delivery — this aligns cash outflow with your sales inflow. Two cautions: never pay 100% up front regardless of discount offered, and always pay through banking channels against a proper invoice. UPI transfers to a personal number save nobody money once GST credit is factored in — more on that below.
The communication problem
Here is where most relationships actually fail. The order lives in a WhatsApp voice note. The revised quantity lives three days deeper in the chat. The approved shade is a photo taken in bad light. When the delivery is wrong, both sides scroll furiously, and both sides are right.
The fix costs nothing: every order becomes a written purchase order with SKU codes, quantities per variant, unit prices, delivery date, and the spec sheet attached. Every change gets a revised PO, not a message. WhatsApp stays — it's where India does business — but it becomes the channel for conversation, while the PO remains the single source of truth. Structured production orders and material POs that generate from your actual demand make this the default rather than a discipline; pair them with the run-sizing math in our production planning guide and the quantity on the PO stops being a guess too.
GST and documentation hygiene
Boring, and worth real money. Verify the unit's GSTIN before the first order (the government portal takes a minute). Insist on proper tax invoices — on an 18% GST category, the input credit on a ₹5,00,000 order is ₹90,000 you can offset against the GST you collect; a kaccha bill makes that credit vanish. Ensure e-way bills accompany shipments above the threshold, and confirm your invoices appear in your GSTR-2B so the credit actually lands. If you supply materials to the unit for processing, use job-work challans (ITC-04 applies) rather than informal handovers. Five documents, one habit: PO, invoice, e-way bill, QC report, goods receipt — filed per order. Your accountant, and any future due-diligence process, will thank you.
When to dual-source
One good unit is efficiency; one only unit is fragility. The signals that it's time to qualify a second source: your hero SKU has a single point of failure, your orders consume more than 60–70% of the unit's capacity, lead times have slipped on two consecutive orders, or your seasonal peak exceeds what one line can produce in time.
Dual-sourcing done well is not a threat to the first unit — keep them as primary with the volume that earns their priority, and give the second unit a small, regular run so the relationship is warm when you need it. The overhead is real (two spec transfers, two QC baselines, slightly worse pricing on the split volume), which is why the wrong time to set it up is mid-crisis. Qualify the backup while things are fine.
Frequently asked questions
It varies by category, but as broad starting points: 100–300 pieces per style in apparel, 500–1,000 units per SKU in beauty (driven by packaging MOQs more than filling), and 50–100 pieces in jewellery or handcrafted goods. Almost every quoted MOQ is negotiable if you consolidate colourways, accept standard packaging, or commit to a repeat schedule.
Yes — even a simple two-page agreement covering specifications, tolerances, delivery dates, defect thresholds, payment terms, and IP ownership of your designs. Many small units work on trust and POs alone, which is fine until a dispute. A written agreement mostly matters because writing it forces both sides to agree on specifics up front.
Inspect within 48 hours of receipt, document defects with photos against your spec sheet, and quantify the failure rate before the conversation. Reasonable outcomes are rework at the unit's cost, replacement in the next run, or a credit note. Units respond far better to "7% of 400 units fail on stitching spec clause 3" than to "quality is bad."
When a single unit's failure would break your business: your hero SKU depends on one factory, your volume exceeds 60–70% of their capacity, or you've had two consecutive late deliveries. Qualify the second unit with small paid runs before you need it — dual-sourcing set up during a crisis costs double.