The standard way to test a business idea in India is to build the product, launch it, and find out. It is also the most expensive way, because the answer arrives after the money is spent.
Validation is the practice of buying that answer earlier and cheaper. It is not market research and it is not asking friends. It is a sequence of small tests, each designed so that failing is informative rather than fatal.
The four questions, in order
Ideas fail at four distinct points, and testing them out of order wastes money. Work down this list and stop as soon as one fails.
- Does the problem exist and does it hurt? Not “would this be nice” but “what are you doing about this today, and what does that cost you”.
- Will someone pay to solve it? Interest is free; payment is a signal. These are not the same measurement.
- Can you reach those people at a cost that leaves a margin? Most Indian consumer ideas die here, not at the product.
- Can you deliver it repeatedly? Doing it once by hand proves demand. Doing it a hundred times proves a business.
Question one: talk to people badly, then well
The mistake is describing your idea and asking whether they like it. Everyone says yes, because saying no to a person is awkward.
Ask about the past instead of the future. “When did you last face this? What did you do? How long did it take? What did you spend?” Past behaviour is data. Future intention is politeness.
Twenty conversations with people who are not your friends is enough to know whether the problem is real. If you cannot find twenty people with the problem, that is itself the finding.
Question two: test payment, not interest
There are cheap ways to test willingness to pay before anything is built.
| Test | What it costs | What it proves |
|---|---|---|
| A one-page site with a real price and a payment button | A weekend | Whether the offer converts. Refund anyone who pays, and tell them why |
| Pre-orders or a paid waitlist | Almost nothing | Commitment, which is a stronger signal than a signup |
| Delivering the service manually to five customers | Your time | Everything — demand, price, delivery cost and the objections you did not anticipate |
| Selling through an existing marketplace before building your own | A commission | Whether demand exists without you having to solve distribution first |
The manual delivery test is the most underrated. Doing the job by hand for a handful of paying customers teaches you more in two weeks than three months of building.
Question three: the number that kills most Indian consumer ideas
Work out what it costs to acquire one paying customer, and compare it to what that customer is worth over their lifetime. In the Indian consumer market, the first number is often higher than the second, and no amount of product quality fixes that.
Test it cheaply: run a small paid campaign to your one-page site, spend a modest fixed amount, and count actual conversions rather than clicks. If it takes ₹900 of advertising to acquire a customer who pays ₹299 once, you do not have a marketing problem — you have a pricing or a channel problem, and you now know it before you built anything.
Cheaper channels worth testing in parallel: WhatsApp and community groups, regional-language content, local partnerships, and referral built into the product rather than bolted on later.
Price in the Indian market, deliberately
Three patterns are worth knowing before you set a price:
- Free-to-paid conversion is low and the free tier is expensive to serve. Decide what the free tier is for. If it is not producing paying customers or genuine word of mouth, it is a cost centre.
- Annual pricing outperforms monthly for many Indian software products, because monthly small-ticket recurring payments have higher failure and churn.
- The B2B ceiling is higher than founders assume and the B2C ceiling is lower. If your idea can be aimed at businesses, test that version first — it is usually easier to validate because businesses can quantify what the problem costs them.
Question four: the boring compliance layer
Validation is not only about demand. Find out early whether the idea is legal to run at scale and what it costs to run properly, because these change the model.
- Business structure. Proprietorship is fast and cheap; a private limited company is what investors and many B2B customers require. Choose based on who you need to sell to
- GST registration thresholds and whether your category has special rules
- Payment collection. Which payment aggregator will onboard your category — some are restricted, and finding out after launch is painful. Our UPI operating guide covers the merchant setup
- Data handling if you collect personal information, which almost every digital business does
- Sector licences if you are anywhere near finance, health, food or education
A one-hour conversation with a chartered accountant at this stage is one of the cheapest inputs available and routinely saves months.
Set the kill criteria before you start
Decide in advance what result means stop. For example: fewer than three of twenty interviews describe the problem unprompted; fewer than five paid pre-orders from two hundred visitors; acquisition cost above a stated ceiling after a fixed spend.
Write these down before running the test. Written in advance, they are a decision. Decided afterwards, they become a negotiation you will win against yourself.
A six-week validation plan
- Week 1: twenty problem interviews with strangers
- Week 2: one-page site with a real price and payment
- Week 3: small paid traffic test, measure cost per paying customer
- Week 4: deliver manually to the first paying customers
- Week 5: ask those customers what nearly stopped them buying, and what they expected that they did not get
- Week 6: compare results against the kill criteria, and decide honestly
Six weeks and a modest budget buys an answer that most founders pay a year and their savings for.
Frequently asked
Is it worth validating if the idea already exists?
Existing competitors are the best possible validation that the problem is real. What you are testing then is not demand but differentiation and acquisition cost.
Should I build an MVP first?
Usually not. Manual delivery answers the same questions faster and cheaper, and tells you what to build rather than making you guess.
How much money should validation cost?
Less than a month of what building would cost. If a validation plan needs serious funding, it is a build plan wearing a disguise.
The short version
Ask about the past, test payment rather than interest, measure acquisition cost before you build, deliver by hand first, and write your kill criteria down in advance. Once you have a validated idea, our guide to automation for Indian SMEs covers running it without drowning in manual work. More business coverage on Techleez com.
Smith Leopard
A contributor to The Techleez editorial desk, focused on practical, clearly explained digital guidance for readers in India.
More from this author →
