Almost every dairy project we are asked to rescue has the same origin story. A promoter with genuine milk procurement strength gets a turnkey quotation from a machinery supplier, the number sounds achievable, land is bought, and eighteen months later there is a plant that can make eight products for a market that wants two.
This guide is the sequence we actually use, in the order we use it. It will not make the project cheaper by itself, but it will stop you spending capital on capacity you cannot fill or products you cannot sell.
Step 1: Decide what you will sell before deciding what you will build
Liquid milk runs at a 4–7% gross margin in most Indian markets. That is not enough to service a term loan and fund growth at the same time, which is why almost every viable dairy project includes value addition. The question is which products, and the honest answer is usually two or three, not eight.
Three constraints decide it. First, your procurement curve: a product that needs steady year-round volume is a poor fit for a supply base that swings 40% between flush and lean. Second, your cold chain reality: curd and flavoured milk need genuine unbroken chilled distribution, while paneer and ghee tolerate far more. Third, your market access: a product you cannot get onto a shelf is a product you cannot sell, however good your yield.
We model contribution per litre of milk for every candidate product against the actual month-by-month procurement curve. On one Kolhapur engagement this exercise reduced the proposed project from eight product lines to three, and the capital requirement from ₹24 crore to ₹18 crore, which was also the reason the bank finally sanctioned it.
Step 2: Choose between a chilling centre and a processing plant
If you do not yet control procurement, build a chilling centre first. At ₹60–90 lakh for a 10,000–20,000 LPD facility it buys you the farmer relationships, the quality data and the supply reliability that make a processing plant viable later. It also generates a trading record, which materially changes how a lender reads your next application.
If procurement is already secured — through a co-operative arrangement, a contract, or an existing collection network — go straight to processing. The failure mode to avoid is a processing plant with no assured milk, and we see at least two of those every year.
Step 3: Size the plant for the despatch pattern you will actually have
Tanker despatch and retail packs are different plants. A dairy selling bulk chilled milk to another processor needs chilling, storage and a tanker bay. A dairy selling 200ml pouches to 600 retail points needs packing capacity, crate logistics, a cold store sized for a day's despatch and a route structure designed around time-to-first-drop.
Plants designed for an average of both do neither well. Decide your channel split first, then build for it, with headroom only where headroom is cheap — usually in utilities and floor space, rarely in packing machinery.
Step 4: Understand what it will cost
At 2025–26 prices, liquid milk with chilling runs ₹9–16 lakh per 1,000 LPD of capacity, excluding land. A 50,000 LPD plant with a paneer and ghee line typically lands between ₹9 and ₹14 crore all-in.
The range is wide, and the variation is rarely in the dairy equipment. It sits in effluent treatment, power backup, the degree of automation, and whether the site needs significant levelling or approach road work. Dairy effluent is high-BOD and a compliant ETP is not optional; on constrained sites it can be 12–18% of project cost.
| Component | Share of project cost | Notes |
|---|---|---|
| Civil works and building | 26 – 34% | Hygienic flooring and drainage are non-negotiable |
| Processing equipment | 28 – 38% | Lower than most promoters expect |
| Utilities (boiler, chilling, DG) | 12 – 18% | Sized from the mass and energy balance |
| Effluent treatment | 6 – 14% | Higher on water-constrained or urban sites |
| Cold storage and despatch | 6 – 12% | Driven by the channel split |
| Pre-operative and contingency | 5 – 8% | Lenders expect this to be explicit |
Step 5: Get the licences in the right order
Eight approvals typically sit on the critical path, and they run in parallel, not in series. Company or firm registration and Udyam registration come first and are quick. Land use conversion and Consent to Establish from the state pollution control board are the long poles — start both immediately, because Consent to Establish alone routinely takes eight to twelve weeks.
Factory plan approval, the FSSAI central licence, fire NOC, boiler registration and finally Consent to Operate follow. Legal Metrology registration is required before you pack for retail sale. Losing four months here is normal for promoters who start the approvals process after the civil work.
- Udyam and company registration — 1 to 2 weeks
- Land use conversion / NA order — 6 to 10 weeks, start first
- Consent to Establish (state PCB) — 8 to 12 weeks, start first
- Factory plan approval and licence — 6 to 8 weeks
- FSSAI central licence — 6 to 10 weeks
- Fire NOC — 4 to 6 weeks, needs frozen drawings
- Boiler registration — 3 to 5 weeks, before commissioning
- Consent to Operate — 4 to 8 weeks, after construction
Step 6: Structure the funding, and file the subsidy first
A typical dairy project is funded with 20–25% promoter contribution, a term loan of 55–65%, and a capital subsidy or interest subvention covering the rest. The schemes that matter most are AHIDF for processing and value addition, PMKSY cold chain components, NPDD for chilling and testing infrastructure, and your state's dairy or industrial policy — which is frequently the largest single component and the one most often missed.
The single most expensive error in this section is sequence. Several schemes require the application to be filed before any machinery order is placed. We have seen ₹2.4 crore of eligible benefit forfeited because a promoter placed an advance on equipment three weeks before filing. Map entitlement before you spend anything material.
Most lost subsidy claims are not lost on eligibility. They are lost on sequence — an order placed three weeks too early.
Step 7: Commission against a checklist, not a handshake
Release the final payment to your equipment supplier against a documented commissioning protocol: water trials, product trials at rated capacity, a validated yield figure, and a defect log closed in writing. A performance guarantee you have not tested is a performance guarantee you do not have.
Baseline overall equipment effectiveness in the first month of commercial operation. Nineteen plants in, our commissioning baseline is typically 55–65% OEE; the plants that reach the mid-eighties within a year are the ones that measured from month one and worked a ranked loss list.
What this looks like on a timeline
Fourteen to twenty months from land in hand to trial production is realistic, assuming funding is sanctioned before civil work starts. Two months of pre-feasibility and design, three to four months to sanction, eight to twelve months of construction and installation, and two months of commissioning and stabilisation.
Promoters who compress this usually do it by starting construction before sanction, which is the most common way a viable dairy project ends up in distress. The bank's timeline is not the enemy of your timeline; it is a constraint to plan around.
Frequently asked questions
A note on the numbers in this article
Benchmarks come from engagements we have delivered and are indicative at 2025–26 prices. Scheme details, rates and eligibility change with each policy cycle — verify the current position before you commit capital. Nothing here is legal, tax or investment advice.