Which Business Should You Start? Begin With the Customer, Then Test the Numbers

First, describe the customer problem in one paragraph

Complete these sentences for each idea:

• Our customer is a specific type of buyer in a defined market
• They currently obtain the product or service through a known alternative
• Their recurring difficulty is something we can describe and verify
• We can solve it through a capability we can actually deliver
• They have a reason to pay enough, and often enough, for the operation to survive

The difficulty might be inconsistent quality, small order availability, unreliable delivery, excessive lead time, insufficient storage, poor documentation or a service bottleneck. It does not have to be a completely new invention. But “there is a big market” needs to become a testable claim about a reachable customer.

Speak to the person who uses the product and the person who approves the purchase. They may care about different things. A production manager might value consistent specifications while procurement focuses on credit, supplier qualification and delivery terms. A good idea must survive both conversations.

Startup India's published discussion of early-stage businesses identifies market acceptance, customer trust and product validation as core challenges. The practical response is to gather buying evidence before building a large fixed-cost operation. Startup India: seed-stage business challenges

Turn conversations into a demand file

For each serious prospect, record:

1. What they buy and the specification they actually accept
2. Typical order size, purchase frequency and seasonal variation
3. How they approve a new supplier or service provider
4. What causes a rejection, return or dispute
5. Delivery, packaging, reporting or service requirements
6. Payment terms and who authorises payment
7. What would make them try an alternative supplier

Ask about recent behaviour. “What happened on the last delayed delivery?” is more useful than “Would you like a reliable supplier?” Ask what would prevent a trial, not only what sounds attractive about the idea.

Keep different strengths of evidence separate. A pleasant conversation is weaker than a documented trial requirement. A trial is weaker than repeat orders with payment received. A letter of intent may have conditions or be non-binding; read what it actually commits the buyer to do.

For a new venture, the aim is not to collect an impressive number of favourable opinions. It is to find out whether the proposed offer can pass a real buying process. Record negative findings as carefully as positive ones. If buyers consistently ask for a certification, delivery promise or credit period you cannot support, that is part of the answer.

Test a smaller version of the business where practical

Before purchasing a full production line, investigate whether a lawful pilot can test the key assumption. Possibilities include an appropriately licensed contract manufacturer, a job-work arrangement, a limited service engagement or a small batch made in a compliant facility.

Choose the pilot around what you need to learn:

• Can the product meet the buyer's specification consistently?
• Will buyers place repeat orders after the first trial?
• Can you deliver, handle complaints and collect payment?
• What operational work was missing from the original estimate?

A pilot has limitations. A small hand-managed batch does not prove factory-scale throughput. A customer buying once at a launch discount does not prove repeat demand at normal commercial terms. Document what the test demonstrates and what remains unproven.

Check the inputs as carefully as the market

For agro-processing, “raw material is available nearby” needs a calendar, specifications and supplier evidence. Investigate harvesting patterns, quality variation, storage needs, contamination or rejection risks and the practical procurement radius. Look for alternatives if a primary supplier or region becomes unavailable.

For manufacturing, include consumables, packaging, spares and maintenance support. A line can stop because of an inexpensive component that takes weeks to replace.

For services, the critical inputs may be skilled time, software, permissions, data quality or specialist review. For warehousing, the product being sold is usable, reliable storage and handling capacity. Occupancy assumptions, customer concentration and operating conditions matter more than a factory-style output estimate.

Build an input file that says what is required, who can supply it, how quality is checked, how long replenishment takes and what the fallback is. Obtain quotations for the actual specifications rather than using a generic project report's old numbers.

Choose a scale your team can operate

Equipment capacity is only one constraint. Compare the proposed operation with the team's experience in production, quality, maintenance, procurement, selling and collections. Decide who will own each responsibility on an ordinary working day.

Ask the equipment supplier what the stated capacity assumes: input quality, product mix, operating hours, staffing, utilities and downtime. Then investigate the bottleneck across the complete process. A fast machine does not create saleable output if cleaning, drying, packing, testing or dispatch cannot keep up.

Plan an achievable ramp-up. Separate technical capacity from the volume the business can realistically sell and collect payment for. Avoid using full utilisation from the first month simply because it makes the report look attractive.

Before buying land or signing a long lease, prepare the operating brief and check activity-specific approvals. The National Single Window System can help identify central and state approvals, but the relevant authority grants them. Manufacturing, food processing, storage and professional services should not be assigned the same compliance list. NSWS: approvals guidance

Read unit economics without confusing them with profit

Choose a meaningful unit: an accepted batch, a completed service engagement, an occupied storage position over time or another measure that matches how the customer buys. Then examine the revenue retained and the costs directly caused by delivering that unit.

Include normal wastage, rejected work, packaging, handling, delivery commitments, payment fees, warranty or rework where relevant. Do not treat a potential by-product sale as certain revenue without checking its market, quality and disposal requirements. For a service business, count the founder's delivery time and subcontractor effort instead of treating both as free.

Positive contribution still has to support salaries, rent, administration, maintenance and other fixed obligations. Financing, depreciation, tax and cash movements also need to be treated properly in the full financial model. Have an accountant distinguish operating results, accounting profit and cash availability.

Keep three cases:

• Base: Supported by quotations, pilot evidence and a defensible ramp-up
• Downside: Slower sales, lower utilisation, more rejection, cost increases or late collections
• Upside: Improvements that have a plausible operational explanation

Change related assumptions together. A weak sales month can mean lower utilisation, longer storage and slower payment at the same time. Testing only one favourable variable at a time understates the risk.

Working capital decides whether you can keep opening the doors

Trace the dates when money leaves and returns. A profitable order can still require cash for inputs, wages and delivery well before the customer pays.

Prepare a rolling weekly cash forecast during setup and early operations. Show committed payments, realistic receipt dates, loan repayments, statutory obligations and the cash needed to fulfil the next orders. Separate confirmed credit facilities from hoped-for borrowing and expected incentives from money already available.

Hypothetical timing example

Suppose a manufacturer pays for material on day 1, completes and dispatches the order on day 20, and receives customer payment on day 50. Ignore supplier credit and advance payments for this simple illustration. The first material payment is exposed for 49 elapsed days, while other operating payments occur along the way. If that receipt moves to day 80, the operation must fund another 30 days before that customer cash arrives.

No selling price or profit assumption is needed to see the problem. The business may already have started the next production cycle. Map overlapping orders rather than modelling every batch as if it finishes and pays before the next begins.

India has statutory delayed-payment provisions for eligible micro and small suppliers, including limits on agreed payment periods. Eligibility and the facts of acceptance matter. Those protections do not make late cash arrive automatically; obtain professional guidance where needed and maintain a collection plan. The Ministry of MSME's guidance explains the relevant provisions and eligibility limitations. Ministry of MSME: delayed payments to MSEs

Put subsidy analysis after commercial feasibility

Incentives deserve careful evaluation. They can influence financing and the choice between otherwise suitable projects. But an advertised percentage cannot answer whether the customer exists, whether the plant can run or whether the business can survive a delayed receipt.

Create a separate scheme note for every incentive under consideration:

• Which authority administers it, and is the relevant application window open?
• Is this applicant, activity, location and project size eligible?
• Which expenditure is eligible, and which is excluded?
• Must an application or approval occur before purchase, construction or production?
• Is support a grant, reimbursement, interest assistance, credit-linked adjustment or another mechanism?
• What must be funded upfront, and when could assistance actually become usable?
• Are there limits on combining schemes, inspection conditions, ongoing obligations or recovery provisions?

Read the operative notification and amendments, then resolve uncertainties with the administering authority and lender. Do not extend an old scheme document into a new financial year without checking its current status.

Run the commercial case without assuming an unapproved incentive will arrive on time. If the project requires assistance to proceed, make the dependency explicit and verify a lawful, financeable path before committing.

Compare opportunities using evidence gates

Use a short decision sheet for each idea. Avoid a single numerical score that lets a serious defect disappear into an average.

Scroll horizontally to read all table columns.

Classify each answer as verified, provisional or unsupported, and state what would change the assessment. Use a conditional decision where that is honest: “Proceed to a compliant pilot if the buyer confirms the trial specification,” or “Reassess after written utility feasibility.”

A hypothetical choice between three models

An entrepreneur is comparing a processing unit, a warehouse and a specialist service business. The processing unit has an attractive potential incentive but no buyer trials. The warehouse has interested tenants but uncertain access permissions. The service business has two paid pilots, yet delivery depends entirely on the founder.

There is no automatic winner. Each has a different next test. The processor needs customer and production evidence. The warehouse needs access verification and a realistic occupancy plan. The service business needs repeat-demand evidence and a deliverable staffing model. Approving the largest capital investment now would be premature; targeted tests can improve the decision before money becomes difficult to recover.

Finish with a decision you can explain

A useful business-selection note should fit the following summary:

“We know who the customer is, why the offer matters, how delivery will work and how the cash cycle will be funded. These assumptions are verified; these remain uncertain. This is the maximum commitment justified by the evidence today, and this is what must happen before the next commitment.”

That is a stronger starting point than choosing the project with the biggest machine, cheapest available plot or highest advertised subsidy. Build confidence through evidence, then increase the commitment as the business earns it.

Sources checked on 12 October 2026. This is general business-planning information, not individual investment, legal or scheme-eligibility advice. Verify live scheme terms and project-specific obligations before committing funds.