You have decided to sell healthy snacks. Now comes the decision that quietly sets your margin, your shelf life and your legal exposure: who actually makes the food. Get this wrong and you ship stale makhana, or food made on a line with no valid FSSAI licence, and one bad batch or one marketplace audit ends the brand. Get it right and you launch a clean, compliant product for a fraction of a factory's cost, with your capital going into demand instead of concrete.
Here is the direct answer first. For your first two years, do not build a food plant. Find a licensed co-packer, a co-manufacturer that already holds an FSSAI manufacturing licence and runs roasting, baking and nitrogen-flush lines, and have them make your recipe to your pouch. Contract co-packing for snacks runs a per-batch or 100 to 500 kg MOQ per SKU, printed pouches carry a separate 3,000 to 10,000 unit minimum, and the whole compliance stack costs under ₹15,000. Before you send a single rupee, verify three documents: the co-packer's FSSAI manufacturing licence copy in their own name, a third-party nutrition lab report, and an accelerated shelf-life study. The rest of this guide is how you find the factory and prove all three.
India grows genuinely good snack raw material and has a dense network of co-packers who will make your product for you. The healthy snacks market sat near US$3.13 billion in 2025, and makhana is one of the fastest-growing snack categories, and India grows the vast majority of the world supply. Co-packers cluster in Delhi NCR, Mumbai and Pune, Gujarat and Bengaluru; makhana comes out of Bihar's Mithila belt, millets from Karnataka and Andhra. Three routes: private-label off a co-packer's stock recipe (cheapest, least yours), contract-manufacture your own recipe (the route for a real brand), or your own plant (₹15 to 40 lakh, only after proof). Co-pack MOQs run per-batch or 100 to 500 kg per SKU; printed pouches add a 3,000 to 10,000 unit minimum. The co-packer holds the FSSAI manufacturing licence; you verify the copy, a third-party nutrition report and an accelerated shelf-life study before you pay. Score every candidate on the Supplier Scorecard. Run a taste panel and accelerated stability test on samples, because shelf life is the number that kills snack brands. Red flags: a trader posing as a manufacturer, no FSSAI licence copy, no stability data, no allergen segregation.
This is the sourcing and manufacturing companion to the full category playbook. If you have not settled your format, pricing and business model yet, start with how to start a healthy snacks brand in India, which covers AOV, bundles, FSSAI tiers and the climb to ₹5 lakh a month. This page goes deep on the one thing that guide only touches: finding the factory, and proving it can be trusted with your name.
The manufacturing map: where snack co-packers actually are
Two facts shape the map. Raw material sits close to where it grows, and finished-snack co-packing clusters near food-processing belts and quick-commerce demand. Match your product to the right cluster and freight, freshness and price all improve at once.
| Cluster | What they make | Why it matters |
|---|---|---|
| Delhi NCR (Delhi, Noida, Sonipat) | Makhana, dry-fruit and trail mixes, roasted namkeen, protein snacks | The densest co-packer belt and the dry-fruit import trade hub; Farmley's home base |
| Mumbai and Pune (Maharashtra) | Baked and roasted snacks, bars, extruded millet snacks | Deep food-processing base, close to quick-commerce demand and premium buyers |
| Gujarat (Ahmedabad, Rajkot) | Namkeen, roasted and extruded snacks | The traditional low-cost snack heartland; strong for value price points |
| Bengaluru | Clean-label millet, protein and premium snacks | D2C-first, clean-label co-packers used to working with young brands |
| Bihar, Mithila belt (Darbhanga, Madhubani, Purnia) | Raw and roasted makhana at source | Makhana's origin; buy the raw material close to where it grows |
| Karnataka and Andhra Pradesh | Millet and ragi processing and snack lines | Millet raw material and the co-packers who handle it well |
If makhana is your hero product, the map starts in Bihar. India produces close to 80% of the world's makhana, and Bihar is the nerve centre, with the Mithila Makhana name carrying a GI tag since 2022. Buy roasted, graded makhana cheaper and fresher near the ponds than through a Delhi middleman, then have a co-packer flavour, pack and nitrogen-flush it. Dry fruits and nuts move through the Delhi and Mumbai import trade, where the mix-and-pack co-packers sit; the sibling dry fruits brand guide covers that lane.
The map also tells you about the model. Farmley scaled to ₹394 crore in FY25 revenue, up from ₹230 crore the year before, and is only now investing ₹40 to 50 crore in its own Noida factory, expected online by FY27. A brand crossed ₹394 crore before owning a plant. It grew on co-packers, and so can you, for a lot longer than you think.
Private label vs contract manufacturing vs co-packing: what the words mean
These terms get used loosely, and the confusion costs money, because each carries a different cost, MOQ and level of control over the one thing that matters, your recipe. Private label puts your brand on the co-packer's existing stock recipe; they own the formula, so a dozen brands can sell the same makhana under different pouches. Contract manufacturing makes your formulation to your spec, and if you fund the recipe development you can own it. Co-packing is the umbrella term for a licensed factory that manufactures and packs to your order, used loosely for both. So always ask which one a quote actually is: their recipe, or yours.
| Route | What it is | Upfront cost | MOQ | Control | Best for |
|---|---|---|---|---|---|
| Private label (stock recipe) | Your brand on the co-packer's existing formula | Lowest; per-unit only | 100 to 500 kg or per-batch per SKU | Low; shared recipe, others can sell it | Fast validation, no recipe yet |
| Contract manufacturing (your recipe) | They make your formulation to your spec | Recipe development ₹15,000 to ₹1 lakh + per-unit | Similar per SKU; higher for custom work | High; the recipe and story are yours | A brand you intend to defend |
| Your own plant | You build and licence a food-processing unit | ₹15 to 40 lakh plus months of setup | None, but fixed cost every month regardless | Total, and total risk | Proven brands with steady volume |
Almost every first-timer should be on route one to validate, then route two to build. The logic of picking between them is in white label vs private label vs OEM in India. Your own plant is a trap dressed as ambition: ₹15 to 40 lakh sunk before a customer has reordered, when the same money spent on demand would have told you whether the product works at all.
In my supply-chain years at Atomberg I learned to draw a hard line between what you own and what you rent, because you manage them differently. In food, the thing you rent from a co-packer is not just a machine. It is their FSSAI licence, their hygiene record and their shelf-life data, and all three become your brand's liability the second they pack under your name. So treat the licence copy the way a banker treats collateral: you do not take the borrower's word for it, you see the document and verify it on the portal. A co-packer slow to hand over that paper is telling you how the relationship goes when a batch fails at 2 a.m. and a marketplace is threatening to delist you. The cheapest quote in food is almost never the cheapest brand outcome.
The FSSAI manufacturing licence chain: the co-packer holds it, you verify it
Every packaged snack carries two licence numbers, and understanding why keeps you out of trouble. The unit that physically makes the food is the manufacturer, and its FSSAI licence and address print under "Manufactured by." Your brand company is the marketer, and your own FSSAI number prints under "Marketed by." Two numbers, two responsibilities, one pack.
Your own tier is set by your turnover. As of the reform effective 1 April 2026, Basic Registration covers turnover up to ₹1.5 crore, a State Licence ₹1.5 crore to ₹50 crore, and a Central Licence above ₹50 crore, with importers, exporters and multi-state operators also needing Central. Licences now carry perpetual validity, so the annual-renewal scramble is gone, per the revised thresholds and the FoSCoS portal. New brands start on Basic Registration; the full walkthrough is in the FSSAI licence guide for India.
Now the part that protects you. A genuine snack co-packer running real roasting and packing lines almost always holds a State or Central manufacturing licence, because a working factory's turnover and output push it past the Basic tier, and its Kind of Business on FoSCoS reads "Manufacturer." Pull up their licence number on FoSCoS yourself. If the Kind of Business reads "Trader," "Wholesaler," "Retailer" or "Repacker," that entity buys someone else's food and relabels it. They are not the manufacturer, cannot answer for the line your food runs on, and their licence does not cover making your product. That single check separates a real factory from a middleman.
How to vet a snack co-packer: the five documents
A co-packer is a supplier, and you score one on evidence, not on the warmth of the sales call. Five lines matter for food, and anything scoring below a weighted 7 out of 10 does not get your order or your recipe.
Supplier Scorecard™ for a snack co-packer, five weighted lines. FSSAI manufacturing licence and documentation (25%): a valid licence copy in their own name, Kind of Business reading Manufacturer, product category covering your snack. Third-party nutrition lab report (20%): a NABL-accredited lab COA for the exact product, because you cannot legally print a nutrition panel you have not tested. Shelf-life and accelerated stability study (25%): documented data proving the best-before date you will print, the single most important line in food. Allergen segregation and hygiene (15%): how they separate nut, gluten and dairy lines, backed by an audit, because a nut-free claim is worthless if the line also runs peanuts. Batch consistency and references (15%): two sample rounds that taste and weigh the same, plus two brand references you can actually call. The cheapest quote wins exactly one of these five lines.
Two of those lines are pure gates. The FSSAI manufacturing licence copy is non-negotiable: no valid copy in their own name, no deal. The accelerated shelf-life study is the line almost nobody asks for, and the one that decides whether your brand survives its first cohort. A co-packer who cannot show stability data is guessing at your best-before date, and you print it and ship it. The generic vetting method, the GST cross-check, the video walkthrough, the golden sample and reference calls, is in how to find manufacturers and suppliers in India. The food-specific additions are the nutrition report, the stability study and the allergen audit.
On the first call, ask four things and listen for hesitation. "Can you send your FSSAI manufacturing licence copy and FoSCoS number today?" "Will you share a third-party lab nutrition report and an accelerated shelf-life study for a similar product?" "How do you segregate your allergen lines?" "Will you run a 100 kg trial batch to my recipe, and a second identical batch four weeks later?" A real co-packer answers all four plainly. Then lock the approved sample as a golden reference in writing, with the recipe, gram weight, pouch structure and target shelf life, so batch three has something to be measured against when it drifts.
The sampling protocol: taste panels and accelerated shelf-life testing
This is where food splits hard from every other category. In skincare you sample once and eyeball the texture. In snacks you run two tests before you commit to a batch, and skipping either is how brands die quietly.
The taste panel. Get eight to twelve people from your actual target buyer to blind-taste your sample against the category leader. Not your family, who will be kind. A snack brand lives on repeat purchase, and repeat is decided by taste in the first bite. If your sample only ties the leader blind, you have no reason for anyone to switch.
Accelerated shelf-life testing. The co-packer or an independent lab holds the product at raised temperature and humidity, commonly around 40°C and 75% relative humidity, so months of ageing happen in weeks, then measures moisture pickup, rancidity (peroxide value) and crispness to project the true shelf life. Roasted and baked snacks typically hold 3 to 9 months, and without nitrogen flush and a barrier pouch, moisture and oxidation collapse that window fast. Shelf life is the killer variable: you print a best-before date, then the pouch sits in a warm dark store for weeks. If the real shelf life is shorter than the printed one, you ship rancid food to the exact early customers whose reorders were the whole point, and they never come back. The packaging that buys that shelf life is in product packaging design in India.
MOQ negotiation: how to start small in food
The MOQ trap in snacks is usually not the food. Contract co-packing runs at a per-batch or 100 to 500 kg per SKU minimum, which is survivable. The pouch is the problem: a custom-printed rotogravure pouch carries its own 3,000 to 10,000 unit minimum, and that printing order, not the makhana, forces you to commit to more inventory than the market has approved. The fix is boring and correct: use stock pouches with a printed label sticker, so your packaging MOQ collapses to almost nothing while you prove demand.
Watch the slab game on the food too. Every quote drops 15 to 25% at the next MOQ tier, and that discount is how founders end up with 5,000 units of a snack the market never wanted, now racing an expiry date. A per-unit discount you cannot sell through before the best-before date is not a saving, it is a write-off with a countdown on it. The scripts for holding your ground are in how to negotiate MOQ with suppliers.
If you are validating with no recipe and no story yet → private-label a stock recipe at the smallest batch, in stock pouches with a label, so you spend on learning demand not on inventory. If you have a defined format and want a brand you can defend → pay for recipe development and contract-manufacture your own formula. If a co-packer pushes a bigger slab "to save per unit" → hold your order at your proven monthly sell-through multiplied by the months of shelf life you can actually clear, never the full best-before window. If you cannot yet state your format, gram weight and target shelf life in one line → you are choosing a factory before you have chosen a product, which is backwards.
Red flags: walk away when you see these
- A trader posing as a manufacturer. Ask for the FSSAI licence and check the Kind of Business on FoSCoS. If it reads Trader, Wholesaler or Repacker, they buy someone else's food and relabel it. No control over the line, recipe or hygiene, and no one accountable when a batch fails.
- No FSSAI manufacturing licence copy, or "it is in process." Their compliance becomes your liability the moment they pack under your name, and marketplaces delist non-compliant listings fast. No valid copy in their own name, stop the conversation.
- No shelf-life or stability data. A co-packer who cannot produce accelerated stability data is guessing at your best-before date. You print that date and ship it, so their guess is your legal and reputational risk.
- No clear answer on allergen segregation. A vague reply on how nut, gluten and dairy lines are kept apart means your "nut-free" or "gluten-free" claim has nothing behind it, a recall waiting to happen.
- Happy to print any claim you like. A co-packer who agrees to print "high protein" or "sugar-free" without asking for a nutrition test does this for everyone. That is the batch that gets your listing pulled when a claim cannot be substantiated.
Trusting a directory listing and skipping the licence check. A founder finds a makhana "manufacturer" on a B2B directory, loves the ₹68 landed price, and orders 3,000 units on a call and a UPI advance. What they found was a trader: a middleman buying bulk roasted makhana from a village unit, repacking it, and relabelling for whichever brand pays. Because the founder never asked for the FSSAI licence copy, they never saw the Kind of Business read "Trader," not "Manufacturer," and that no stability study sat behind the best-before date. Two months in, a batch arrives soft and faintly rancid. The customers who had reordered, the entire point of a snack brand, taste it and quietly leave. The loss is ₹2 lakh of dead stock, plus a poisoned first cohort, plus a marketplace strike. The fix cost nothing but one question before payment: send me the FSSAI licence copy, and let me read the Kind of Business.
The FSSAI claim boundary: what you can and cannot print
The words that sell a healthy snack, healthy, high-protein, no-added-sugar, are exactly the words FSSAI and the marketplaces police. Under the FSSAI Labelling and Display Regulations and the Advertising and Claims Regulations, a front-of-pack nutrient claim is only legal when the product meets a defined threshold and you can prove it with a lab test. This is why the third-party nutrition report in your vetting checklist is not paperwork. It is your permission to make the claim that moves the product.
- "High protein" / "source of protein" map to nutrient thresholds (broadly, a defined share of energy from protein). You need the nutrition test, and the number has to clear the bar.
- "No added sugar" means genuinely no added sugars and no ingredient that itself contains added sugar; naturally present sugars still get declared.
- "Sugar-free" is a hard limit of under 0.5 g of sugar per 100 g or 100 ml, not a vibe.
- No health or disease claims you cannot substantiate, and no "immunity" shortcuts; those get listings pulled and invite penalties.
Print a claim you cannot back with the lab report, and you are one buyer complaint from a delisting and a compliance hit. Beyond the claim, the pack needs a Legal Metrology label: net quantity, MRP inclusive of taxes, month and year of manufacture, best-before, batch number, both FSSAI numbers, the nutrition panel, the veg or non-veg mark, the ingredient and allergen list, and a consumer-care contact. You also need GST registration from day one for any marketplace, most branded packaged snacks now sitting in the 5% slab after GST 2.0 (effective 22 September 2025), with a few specialty items higher, so verify per HSN, and it is worth filing a trademark in Class 30 before you print a single pouch.
Execution checklist
- Write a one-page spec before you contact anyone: format, gram weight, target landed cost, target shelf life, and the one claim that sells it.
- Shortlist co-packers by cluster: makhana near Bihar, millets in Karnataka or Andhra, finished-snack lines in Delhi NCR, Maharashtra, Gujarat or Bengaluru.
- Ask every candidate for the FSSAI manufacturing licence copy and verify the FoSCoS Kind of Business reads Manufacturer, not Trader.
- Demand a third-party NABL lab nutrition report and an accelerated shelf-life study before you print any nutrition panel or best-before date.
- Ask how allergen lines are segregated, and get it in writing if you plan a nut-free or gluten-free claim.
- Run a blind taste panel of 8 to 12 target buyers against the category leader; a tie is not a reason to switch brands.
- Score each co-packer on the Supplier Scorecard™ and reject anything below a weighted 7 out of 10.
- Start on stock pouches with a printed label; add custom rotogravure pouches only once demand justifies the 3,000 to 10,000 unit minimum.
- Cap your first order at proven sell-through times the shelf-life window you can clear, never the full best-before date.
- Register your own FSSAI and GST, file the Class 30 trademark, and build the Legal Metrology label before the first pack ships.
Your next action
Today, do two things. Write the one-page spec: format, gram weight, target landed cost, target shelf life, and the single claim that sells the product. Then message five co-packers across two clusters for samples, quotes at your MOQ, and the three documents, the FSSAI manufacturing licence copy, a nutrition lab report, and an accelerated shelf-life study. The samples cost little, arrive within a week or two, and turn this guide from reading into arithmetic and a product you can taste and test. The frameworks here come from Ravikant Tyagi's operating system for this journey, and when you are ready to turn a good pouch into a repeat-purchase business, the model is in the D2C subscription playbook, with returns handled in how to reduce RTO on COD orders.
If you'd like the complete execution system, calculators, SOPs, templates and operating frameworks behind this process, continue inside D2C Acquisition.Lab.
