Startup Idea Validation Checklist for 2027

On this page

1. How to use this checklist
2. Stage 1: The problem is real and keeps coming back
3. Stage 2: You know exactly who it’s for
4. Stage 3: The offer makes sense to them
5. Stage 4: Somebody actually did something
6. Stage 5: The rough numbers don’t break
7. Where a simulated test fits in
8. Your next step

Nobody skips validation on purpose. It just slides. You mention the idea to a few people, one of them says it sounds great, a friend offers to do the logo for free, and three weeks later you’re comparing fonts without having asked a single stranger whether they’d pay.

A checklist is a dull fix for that, and dull is fine here. It drags the awkward questions into the open, in an order that stops you doing the expensive work first.

There are five stages below. Each has a handful of yes/no checks and a note on what counts as proof. Print it, stick it somewhere you’ll see it while you plan for 2027, and don’t move on until the current stage is mostly ticked.

How to use this checklist

Every box is a claim, and a claim needs something you saw or heard behind it. Belief doesn’t tick anything. “People hate doing their expenses” is a belief. “Four of the six freelancers I asked do their expenses once a quarter, in a panic, the night before the deadline” is evidence, and you can hear the difference just by reading the two out loud.

A “no” is simply the next item on your to-do list. Plenty of ideas will stall at stage one (which is the cheapest place an idea can possibly stall).

Go in order, even when jumping to pricing or a landing page feels more productive. A great price for a problem nobody has is still worth nothing.

Stage 1: The problem is real and keeps coming back

What you’re checking here is that people already feel the problem without your help.

• You can describe the problem in one sentence without mentioning your product.
• Several people have described it in their own words, before you brought it up.
• It happens weekly or daily, not once a year.
• People already spend time or money working around it.
• You can say what it costs them when it goes wrong.

Give the workaround the most weight. It might be a spreadsheet someone built, a phone alarm that goes off every Friday at 4pm, or a part-timer who comes in two afternoons a week mostly to deal with it. Workarounds are proof of pain. When nobody has bothered with one, the problem may well be real but not annoying enough to pay to fix.

Writing the problem as a sentence you can actually check takes some practice, and the article on testing a business idea walks through it.

Stage 2: You know exactly who it’s for

Picture two founders with the same idea, a tool that chases unpaid invoices. One says it’s for “small business owners aged 30 to 50.” The other says it’s for “people who run a two-person bakery and do the books on Sunday nights.” Only the second one knows where to go on Monday morning, and only the second one will get answers sharp enough to act on.

The customer

• Have you described them by situation rather than by age and income?
• Do you know who pays, and is that the same person who uses it?
• Could you name the group who feels the problem worst, so you can start there?

The reach

• Do you know where these people already gather, online or in person?
• Have you actually reached a few of them, or just assumed you could?

That last one quietly kills a lot of good ideas. Buyers who are hard to reach now, while you’re small and keen, get harder to reach once you need hundreds of them.

Stage 3: The offer makes sense to them

Only now does your solution come into it.

• Someone in your audience can explain back what you do after hearing it once.
• You know what they use today instead, including “nothing.”
• You can say why yours beats that alternative, in their words rather than yours.
• You’ve heard the main objections and have an answer for the top two.
• There’s a first version you could deliver by hand, with no code and no factory.

The explain-it-back test deserves a proper go. Pitch someone in a sentence or two, then ask them to describe it to an imaginary friend who missed the conversation. Say your pitch was a booking app for mobile dog groomers, and what comes back is “some kind of pet thing, like an app for vets?” Your message is fuzzy. Honestly, that’s good news, because a fuzzy sentence costs nothing to fix today and quite a lot to fix after you’ve paid for ads nobody understood.

For objections, the free Customer Objection Finder gives you a starter list to carry into real conversations. Read each item as a question to put to someone, not as an answer.

Stage 4: Somebody actually did something

Interest is cheap, so this stage only counts actions.

• Did a few people give you something real, like an email address, a booked call, a deposit or a signed letter of intent?
• Did any of them come from outside your friends and family?
• Did they see a price before they acted?
• Did you write down your pass mark before you started?

Almost everyone skips the pass mark. Before any test, write a sentence such as “If at least three of the twenty people I pitch pay a deposit within two weeks, I keep going,” and put a date next to it. Goalposts are much harder to move quietly once they’re on paper.

What doesn’t count

Compliments, “I’d totally use that,” likes on a post, a waitlist made of your cousins. They feel lovely. None of them tells you whether a stranger will pay.

Stage 5: The rough numbers don’t break

No financial model yet. You only need to know the idea survives simple arithmetic.

• You know roughly what one customer pays you per month or per order.
• You know roughly what it costs to deliver to them.
• You have a guess at the cost of winning one customer, and it’s clearly lower than what they’ll pay you over time.
• Your first revenue goal needs only as many customers as you can actually picture finding.

Imagine Priya, a nurse, planning a meal-prep service for colleagues on night shifts. Each customer pays a weekly fee, and after ingredients and those pricey compostable tubs she keeps a modest slice. Now say finding one customer takes three evenings of handing out flyers in the hospital parking garage, and the average customer stays about a month. The arithmetic says stop and rethink. Maybe the channel is wrong, or the price is, or she should be selling to the hospital rather than to each nurse. The checklist won’t pick for her, but it shows her exactly where to look.

Taxes, permits and food-safety rules vary a lot from place to place. For those, spend an hour with an accountant or your local government’s business pages before any money changes hands.

Where a simulated test fits in

Around stage 3, a rough read on how different kinds of people might react can save you a few wasted conversations. SellTo.us does this with AI personas: you describe the idea and who it’s for, and each simulated participant lands on Interested, Maybe or Not interested with a one-line reason. The report splits those reactions by audience group and pulls out the objections that came up most.

Use it to sharpen stage 3 and to decide who to call first. It can’t tick stage 4. Nobody in a simulated crowd is a real person or spends real money, and the result is neither demand nor a forecast, so the real-world boxes still need real-world proof.

Your next step

Sit down with stage 1 tonight and be strict. For each check you can’t back with something you saw or heard, write down one conversation or small test that would settle it. Book the first of those for this week, and leave the 2027 launch date blank until you have.

When you can’t tell which belief is shakiest, the article on finding your riskiest assumptions helps you pick one, and giving the idea an honest roast often shows which stage needs you first.

Written by the SellTo.us Team. Examples in this article are hypothetical. SellTo.us results come from AI-simulated participants, not real people, and don’t predict sales or guarantee success.