How to validate a business idea before you spend money

To validate a business idea, run a short series of tests in order of what they cost: talk to ten people who have the problem, find out what they use and pay for now, say your price out loud, put up a one-page offer, then ask for money with a pre-order. Write down what would count as a pass before each test runs. Only the last test proves demand, because it’s the only one where people pay; the others are cheap ways to find the broken part before you get there. Most of it costs hours rather than dollars, and all of it comes before the stock, the software, the logo or the LLC.

What counts as proof, and what doesn’t

A test proves something when people give up something they’d rather keep: money first, then their time, then their name on a list. Everything else is opinion, and opinions about a new idea are mostly polite.

These don’t count:

• friends and family saying they’d buy it (they’re being kind, and they’re rarely the group);
• a yes to “would you buy this?”, which costs nothing to say;
• likes, follows and comments on a post about the idea;
• a survey where most people tick “very interested”;
• a rising line on Google Trends: its figures are scaled from 0 to 100 against all searches in that place and period, so they show relative interest, not how many people search, and nothing about whether anyone would pay;
• competitors existing, which shows people pay for something like it, but not that they’ll pay you.

The tests below run from cheapest to most expensive, and each proves a little more than the last:

Test	What it costs	What it proves	Example pass mark
Ten conversations	A week of evenings	The problem is real, recent and costs them	Six of ten describe it before you mention the idea
What they use now	An afternoon	What you’re up against, and what they pay	Three alternatives named, with their price and flaw
The price, aloud	Nothing	The price doesn’t send the group away	Most accept it, and a few say it’s too much
A one-page offer	A free plan and a few hours	Strangers act on the offer, not just the problem	20 sign-ups from the 200 people you sent there
A pre-sale	Card fees on the money you take in	People pay	Five pre-orders from people you didn’t know before

The pass marks are examples to show the shape. Set yours from what the business needs: if it has to sell 30 a month to cover its fixed costs, five pre-orders is a start, not a launch.

Tip: write the pass mark down before the test runs, and tell someone what it is. A pass mark set after the results come in always passes.

Name the group it’s for

An idea for “anyone” can’t be tested, because there’s nobody in particular to test it on. Name one group in a line you could search for: a trade, a place, a job title, a situation. “Dog groomers in Columbus” (an example) can be found, called and priced for; “pet owners” can’t.

The check is simple: could you find ten of them this week? If not, narrow the group until you could. It can widen later; the first customers come from one group.

If the business is local, the US Census Bureau’s Census Business Builder maps the people and businesses in an area by type, so you can see whether enough of your group live or work within reach before you print a flyer or sign a lease.

Talk to ten people who have the problem

The first test is the cheapest and catches the most broken ideas. Have ten short conversations with strangers in the group about the last time the problem came up. Don’t pitch the idea: people are polite about ideas and honest about last Tuesday.

1. Find ten strangers in the group, through introductions and the places they already gather.
2. Ask for fifteen minutes about how they handle the problem, not about your idea.
3. Ask about the last time it happened, step by step, and what it cost them in money and hours.
4. Write down their words and any numbers as soon as the conversation ends.
5. Count how many raised the problem without prompting, and how many already spend money or time on it.

The question script and tally sheet for customer interviews gives you the questions to ask, the ones to avoid and a way to turn ten conversations into a decision.

If most of the ten don’t have the problem, or have it and shrug, don’t build anything yet. Change the group or the problem and talk to ten more.

Find out what they use and pay for now

The real competitor is whatever your group does today, and it’s rarely nothing. List the three things they use now, counting a spreadsheet, a friend who helps, a local business and putting up with it, with what each costs them and what they dislike about it. Your conversations will have given you most of this already.

Then look at the businesses already selling a fix:

• their prices, written down;
• what their customers complain about in reviews and forums, which is where a new business gets in;
• who they’re built for, and who they leave out.

The SBA’s guide to market research and competitive analysis lists what to compare (market share, strengths and weaknesses, barriers to entry) and the free federal data for sizing a market.

“There’s nothing like it” usually means nobody pays to solve this, or nobody has looked hard enough. A crowded field is less worrying than an empty one, because it shows people pay. The question is what you’d do better, and for which group.

Say the price out loud

Before you build a page, put a number on the idea and say it to people in the group: “It’s $60 a month,” or “$150 a job,” to take two examples. Don’t ask what they’d pay; that gets a guess. Say the price and watch what happens.

• Most accept it and a few say it’s too much: the price is in range.
• Everyone says yes straight away: it’s probably too low.
• Most say no, even at the lowest price that covers your costs: the idea doesn’t run at this price, and a discount won’t fix it. A different group, product or way of delivering it might.

Work the number out first, from both ends: what the alternatives cost the customer, and what one sale costs you, your time included. How to price a product works both ends through with an example.

Put up a one-page offer

Next, test whether strangers act on the offer itself. One page is enough:

1. One sentence: what it is, who it’s for and what it changes for them.
2. The price.
3. When it will be ready.
4. One button: join the waitlist, request a spot or pre-order.

Be plain on the page that it isn’t ready yet. A waitlist with a stated date is honest; a “Buy now” button that leads to “sold out” is not.

Put the page in front of the group: send it to the people you spoke to and ask them to pass it on, post it where the group gathers (within each place’s rules), or spend a small, capped ad budget aimed at the group. Count visits and sign-ups, not likes. There’s no standard sign-up rate to aim for, so set the number that would make you go ahead before the page goes up. One-page site builders and form tools with free plans are in the Toolbox.

Ask for money: the pre-sale

An email address costs nothing to give. Money is the test that counts. Go back to the people who signed up, and to the ones from your conversations who had the problem worst, and ask for a pre-order, a deposit or a first job at full price, with:

• a clear delivery date;
• the full price, or the deposit and what it counts toward;
• a full refund if you don’t go ahead or can’t deliver.

Federal rules cover shipping dates, delays and refunds on goods ordered online, by mail or by phone, and they apply to pre-orders. How to pre-sell a product before you make it covers them, along with when to ask for full price, a deposit or only a place on the waitlist.

This is also the first test that takes anyone’s money, so it’s the point where people ask whether they need an LLC first. LLC vs sole proprietorship: which a new business needs goes through it.

If nobody will pay, don’t build it to find out why. Change the offer or the group and ask again.

Read the results: go, change or drop

Put each result next to the pass mark you wrote down before the test.

• Go: the pre-sale met its mark. Build the smallest version that delivers what you sold, by the date you promised. The setup spending (the LLC, the website, the stock) now has a reason.
• Change: the earlier tests passed and one didn’t. Change the part that failed and rerun that test only: a narrower group if the conversations were mixed, a different offer if the page drew visits but no sign-ups, a smaller first step, such as a deposit, if people signed up but wouldn’t pay.
• Drop: the problem wasn’t there in the conversations, or nobody would pay a price that covers your costs, even after one honest change. Dropping an idea after a few weeks and a few dollars is the test working. Keep your notes: the group and its problem may suit a different idea.

For a quick read of where you stand, the four demand questions in the Idea Inspection ask exactly this: who it’s for, whether you’ve talked to people, what they use now and whether anyone has put money down. The Validation section has the guides for each test.