A detailed project report is the evidence behind the numbers in your loan application. It sets out what you intend to manufacture, what it costs to set up, where the demand is, and how the money gets repaid — with each figure traceable to a quotation, a survey or a comparable business rather than to an estimate.
It is not a business plan, and treating it as one is why so many are returned. A business plan persuades. A DPR evidences.
What is the bank actually reading it for?
Three questions, in this order.
Is the project cost real? Meaning: does the machinery cost what you say, has anyone quoted it, and have you included the things first-time promoters forget — installation, electrical work, pre-operative expenses, and the working capital needed before the first customer pays.
Will anybody buy it? Not whether the market is large, but whether this specific unit can sell its output at the price assumed, against the competitors already doing so.
Does it repay? Whether the cash the project generates covers principal and interest with margin to spare, under assumptions that survive a sceptical reading.
Everything in the report should serve one of those three. Sections that serve none of them are why DPRs run to a hundred pages nobody reads.
What goes in it?
| Section | What it must actually establish |
|---|---|
| Promoter background | That someone involved has done something relevant before |
| Product and process | What is made, how, and to what specification or standard |
| Market | Who buys it, at what price, and who they buy from today |
| Location and infrastructure | Land, power, water, effluent, access — with status, not intent |
| Plant and machinery | Itemised, with supplier quotations attached |
| Project cost | Every component, including the ones usually missed |
| Means of finance | Promoter contribution, term loan, working capital |
| Manpower | Numbers, skills, and whether they are available locally |
| Financial projections | Revenue, costs, profitability, DSCR, break-even |
| Implementation schedule | What happens in which month, to commissioning |
The location row is worth care. “Land identified” is not the same as “land acquired”, and “power applied for” is not “power sanctioned”. An officer reads those distinctions closely, because a project that stalls waiting for a connection is a loan servicing nothing.
Where do project costs get understated?
Consistently, in the same places — and understating is worse than overstating, because you run out of money halfway through and have to go back.
- Installation and commissioning, treated as included in the machinery price when it is not.
- Electrical infrastructure — transformer, cabling, connection charges, which for a manufacturing unit is rarely trivial.
- Pre-operative expenses — approvals, registrations, consultants, trial production.
- Margin money for working capital, forgotten because it is not a fixed asset.
- Contingency, which experienced officers expect to see and are suspicious when they do not.
- Interest during construction, payable before anything is produced.
A cost sheet with no contingency line reads as a first attempt rather than a considered one.
What makes the market section credible?
Specificity about buyers, not size of market.
“The Indian market for this product is worth thousands of crores and growing” is in almost every DPR and establishes nothing — a market being large does not mean a new unit can enter it. What establishes something is naming the customer type, the geography you will serve, the price you can achieve, who currently supplies them, and why anyone would switch.
Where you have letters of intent, supply enquiries or an existing trading relationship in the same product, include them. Evidence that someone specific will buy is worth more than any amount of industry data.
What about the first year?
Under-plan it, deliberately.
The commonest error in new-project files is first-year revenue at or near full capacity. Real plants commission late, run trials, produce rejects, and take months to find their customers. A ramp-up across three years reads as considered; a first year at ninety per cent utilisation reads as though nobody involved has commissioned a factory before — and it undermines the credibility of every other number.
We have written separately about what makes projections believable in financial projections a bank will believe.
How does this connect to the CMA?
The DPR justifies the assumptions; the CMA presents them in the bank’s format. They must agree exactly.
Two documents prepared separately — often by different people, at different times — is a reliable way to produce a file where the sales figure in one does not match the other. If your DPR says capacity utilisation reaches 65% in year two, your CMA operating statement must reflect the same revenue.
Build the financial model once, and generate both from it. What a CMA report is covers the second half of that.
Last verified: 22 August 2026. General guidance, not financial advice. Requirements vary by bank, scheme and state — confirm before relying on it.
Common questions
How long should a DPR be?
As long as the evidence requires. A focused report with quotations and a defensible market section beats a hundred pages of industry background, most of which the officer will skip.
Do we need a consultant to prepare it?
Not necessarily, though most promoters use one for the financial sections. The parts a consultant cannot do for you are the market evidence and the supplier quotations, and those are the parts that carry the file.
What if the machinery quotation expires?
Refresh it. Sanction processes take time and a stale quotation is a common reason for a cost revision midway, which is disruptive and avoidable.
Can one DPR be used for a subsidy application as well?
Often the same underlying report serves both, though scheme applications usually have their own format and additional requirements. Check the specific scheme’s format before assuming — requirements vary by state and change.
More on this: projections a bank will believe and how to choose what to manufacture. See also StartYourIndustry.
Webmaster
Keep reading
More on Starting a Business
Starting a Business
How to choose what to manufacture
Most people choose the product first and check the economics afterwards. That is the wrong order, and it is expensive.
· 4 min read
Starting a Business
Financial projections a bank will actually believe
Credit officers read hundreds of these. They are not checking whether you are optimistic - they are checking whether the optimism is costed.
· 4 min read
Starting a Business
Why CMA files get sent back
Usually not because a number was wrong. Because two forms disagreed and nobody noticed.
· 5 min read
