How to Validate a Business Idea Before You Spend Money

Most new businesses that struggle don’t fail because the founder worked too little. They fail because the founder built something not enough people wanted. The good news is that you can validate a business idea cheaply, in a couple of weeks, before you put serious money into it.

This guide shows you how. It covers talking to customers the right way, sizing the market with simple maths, checking competitors, testing demand with a landing page or pre-orders, and building a tiny first version. It ends with a 14-day plan you can follow.

Validation doesn’t guarantee success, but it shifts the odds. It replaces “I think people want this” with evidence, and it tells you early whether to go ahead, change direction or stop.

Start by writing down your assumptions

Before you speak to anyone, write one sentence each on who your customer is, what problem they have, how they deal with it today and why they would pay you instead. Then circle the assumption that would sink the business if it turned out to be wrong. That riskiest assumption is what you test first.

Run problem interviews

Problem interviews are short one-on-one conversations, usually 20–30 minutes, with people you think have the problem. You are there to learn about their life and work, not to pitch. Find them through your network, LinkedIn, college alumni groups, WhatsApp communities, local associations or by simply walking into shops if your customers are small businesses.

As a rough rule of thumb, aim for 15–20 conversations with one customer type and keep going until you stop hearing anything new. Take notes, and record the exact words people use. Those words later become your marketing copy.

Good questions vs bad questions

The most common mistake is asking people to predict the future or judge your idea. People are polite, so they will say yes. Ask instead about what they have actually done. The book The Mom Test by Rob Fitzpatrick is a useful guide to this style of questioning.

Close every interview by asking “Who else should I talk to?” and “Can I show you something when it’s ready?” Warm referrals and people willing to try an early version are both early signals.

Size the market: TAM, SAM and SOM

Market sizing tells you whether the opportunity is big enough to be worth your time. Use three layers:

• TAM (total addressable market): the total yearly revenue if every possible customer bought from you.
• SAM (serviceable available market): the part of TAM you can realistically reach with your product, language, geography and channels.
• SOM (serviceable obtainable market): the share of SAM you could realistically win over the next few years.

A simple worked example (made-up numbers)

Note: every figure in this example is invented, in round numbers, purely to show the method. They are not real market data. Imagine an app that helps home tutors track fees and attendance, priced at ₹3,000 a year.

• TAM: suppose there are 10 lakh home tutors in India. 10,00,000 × ₹3,000 = ₹300 crore a year.
• SAM: suppose your app only supports languages used in five states, and only 2 lakh of those tutors use smartphones for work. 2,00,000 × ₹3,000 = ₹60 crore a year.
• SOM: suppose you could win 5% of SAM in three years. 10,000 tutors × ₹3,000 = ₹3 crore a year.

For a real estimate, replace every “suppose” with a number you can defend. Sources include government statistics, industry association reports, your own counts (for example, listings in your city on map apps) and your interview findings. Google Trends can show whether search interest in a problem is rising or falling. Investors and lenders trust a carefully built bottom-up number far more than a headline figure lifted from a report.

Scan your competitors

Search wherever your customers would look: Google, app stores, marketplaces, Instagram, LinkedIn and local directories. Remember the indirect competitors too, such as spreadsheets, WhatsApp groups, agents or simply doing nothing.

Make a simple table of eight to ten alternatives with their price, target customer, main strengths and the complaints that keep showing up in their reviews. One- to three-star reviews are especially useful, because they show exactly where customers feel let down. If you already have a brand name in mind, check it on the IP India trademark search before you print anything.

If you truly can’t find anyone solving the problem, be careful. Either the market is very new or very few people care enough to pay.

Test demand with a landing page or pre-orders

A landing page is a single web page that describes your offer as if it already exists. It should have a clear headline about the problem you solve, three or four benefits, a price and one call to action: join the waitlist, book a call or pre-order.

Send traffic to it through communities, WhatsApp groups, LinkedIn posts, the people you interviewed and, if you like, a small ad budget. Then measure how many visitors take the action. Payment is the strongest signal of all, even a small refundable deposit.

Be honest with anyone who pays early. Tell them it is a pre-order, give a realistic delivery date and refund them promptly if you decide not to go ahead. A “fake door” test, where a button for a feature leads to a “coming soon, want early access?” message, is also fine as long as you don’t take money for something you can’t deliver.

Choose the right type of MVP

A minimum viable product (MVP) is the smallest version that delivers the core value, so you can learn whether people use it and pay for it. It doesn’t have to be software.

Concierge MVP

You deliver the service by hand, and customers know a person is doing it. For example, rather than building a meal-planning app, you personally send weekly plans over WhatsApp to ten paying families. It’s slow and doesn’t scale, but you learn a lot about what customers value.

Wizard of Oz MVP

To the customer, the product looks automated, but behind the scenes people do the work. This tests whether customers like the experience before you spend months automating it. Never make false claims about how the service works, and handle customer data carefully.

No-code MVP

You stitch together existing tools: forms and spreadsheets, a WhatsApp Business catalogue, a simple website builder, payment links and no-code app builders. You can often have something working in days, and you only hire developers once demand is proven.

Go or pivot: reading the signals

Set your criteria before you start testing. For example, you might decide: “At least half the people I interview describe this as a top-three problem, and at least five people pay a deposit within two weeks.” Those numbers are only an illustration. Pick targets that suit your business.

Signals that point to “go”

• People describe the problem in detail without being prompted, and get animated about it.
• They already spend money or time on clumsy workarounds.
• Some pay, pre-order or commit time to a pilot.
• Early users come back and refer others without being asked.
• You can reach customers through a channel that doesn’t cost more than they are worth.

Signals that point to “pivot” (or stop)

• Lots of “sounds interesting” but nobody commits time or money.
• The problem is real but low priority, a “nice to have”.
• Only friends and family sign up.
• Customers like the solution but won’t pay a price that covers your costs.
• You keep changing who the target customer is to make the idea fit.

A pivot doesn’t mean failure. It usually means you change one thing, such as the customer segment, the problem you focus on or the business model, and test again.

A 14-day validation plan

Frequently asked questions

How many customer interviews do I need to validate a business idea?

A common rule of thumb is 15–20 interviews for each customer type, continuing until answers start to repeat. If different groups give very different answers, you may be mixing two segments and should treat them separately.

Can I validate an idea with an online survey instead?

Surveys are useful once you know what to ask, for example to check how common a problem is that interviews have already uncovered. They are a poor replacement for conversations, because you can’t ask follow-up questions or hear the stories behind the answers.

What if someone steals my idea while I am testing it?

At this stage execution matters far more than the idea. You can talk about the problem openly without revealing every detail of your solution. Most people you interview are potential customers, not competitors.

How much money should I spend on validation?

Very little. Many founders validate with only a few thousand rupees for a domain, simple tools and perhaps a small ad test. Treat that as a rough guide, not a rule. Your main investment is time.

Is a long waitlist enough proof to start building?

Not on its own. Free sign-ups cost people nothing, so they are weak evidence. Deposits, pre-orders, signed pilots and repeat usage are much stronger signals that people will pay.

Pick your riskiest assumption today and book your first five interviews this week. Two weeks of honest testing can save you months of building the wrong thing. This guide is general information, not financial advice.