VettedGaps
All research
Idea validation11 min read

Idea validation: what it is, a framework and a template

Idea validation explained: what it can and cannot show, a five-level evidence framework, a copyable template and real examples from VettedGaps data.

Idea validation is the work of collecting outside evidence, before you build, that real people have the problem your idea solves and already spend money, time or search effort on it. The output is a graded record of what was observed for each part of the idea, with links, plus the parts where nothing was found. Treat it as evidence for a decision: it can’t tell you the product will sell.

The framework we use grades every function of an idea on five levels: E0 not observed, E1 people said it, E2 people search for it, E3 people pay someone to do it, E4 people pay for a product. A function with only E1 needs interviews. A function at E3 or E4 has money behind it. This page covers the definition, the levels, a template and examples; the step-by-step method is in our guide on how to validate a business idea.

This page is for indie developers and micro SaaS builders who have an idea, or three, and want to know what counts as evidence before they write code. Every example is about software, because that is what our data covers: we read public complaints in app marketplace reviews, vendor forums, Reddit and one-star Chrome extension reviews, drop praise and support noise, and group posts about the same frustration into pain cards. As of November 2026 the index holds 7,775 pain cards across 78 marketplaces.

Key takeaways

  • Idea validation records what was observed about demand. It can’t show that your product will sell.
  • Grade each function of the idea separately, from E0 (not observed) to E4 (people pay for a product).
  • In our public sample report, one function out of four had direct evidence, and the verdict was Needs proof.
  • A complaint is a lead. Money already spent on the job is stronger evidence than anyone saying they would pay.
  • Write the result that ends the idea before you run any test.

What is idea validation

In plain terms, it’s a check of an idea against things people already did, rather than things they say they might do. Unprompted complaints, searches, freelance orders and paid products all count. A friend’s opinion doesn’t, and neither does a survey answer about a product that doesn’t exist.

The idea validation meaning that matters for a builder is narrow: you test claims about the idea one at a time. “Shopify merchants with several warehouses lose orders when stock sync fails” is a claim you can check. “Merchants want inventory peace of mind” can’t be checked, since nobody searches for it, complains about it in those words or pays a freelancer for it.

That is why we split every idea into functions. A function is one sentence with a verb that describes one job of the product. To validate an idea is to find evidence for each function on its own, because a pitch can read well while only one of its parts has any proof.

What idea validation in entrepreneurship can and cannot show

Idea validation in entrepreneurship answers one question: is there observable demand for the jobs this product would do? People also search for “validation of your ideas ensuring success in business and entrepreneurship”, and the honest answer is that no check does that. Validation lowers the odds of building something nobody wanted. Whether buyers accept your price and keep paying shows up only after launch. Each source of evidence has a limit:

Evidence It can show It cannot show
Complaints The problem exists and who has it That they would pay you
Search volume People look for a fix often enough to measure That they want your fix
Paid freelance work The job already has a price and buyers That they would switch to software
Paid products People pay for a product that does the job That there is room for another one
Your commitment test People pay or pre-order your version That the result holds at scale

Our method page puts it in one line: “No single signal proves a business. Agreement between independent signals is what makes an opportunity worth investigating.” (source)

The idea validation framework: five evidence levels

An idea validation framework is a fixed set of rules for what counts as evidence and how strong each kind is, so two people grading the same idea reach the same result. Ours, the VettedGaps evidence levels, uses the function as its unit and a five-step scale that records what was observed, never how convincing the idea sounds:

Level What was observed Typical source
E0 Not observed Nothing found, or nobody looked
E1 People said it Complaints in reviews, forums, Reddit
E2 People search for it Monthly searches for task phrases
E3 People pay someone to do it Freelance gigs with buyer reviews
E4 People pay for a product Paid apps with real customers

Our sample validation report states the rule for E0 plainly: it means we have not looked, and it counts neither for nor against the idea.

Two more rules sit on top of the scale. First, each piece of evidence is judged for relevance: the validator marks every retrieved pain card as directly supporting the idea, merely adjacent, or unrelated. An adjacent complaint describes a neighbouring problem and doesn’t raise the level. Second, count distinct people. One person posting in five threads is one data point.

Read the levels as a ladder for what to do next. Park an E0 function until a customer raises it unprompted. Functions at E1 and E2 need problem interviews. Once a job reaches E3 or E4, money has already moved for it, and the open question becomes price and competition.

The idea validation process at a glance

The idea validation process has four stages, and each one produces something you can write down:

  1. Describe the user, the problem and what they do about it today, in one paragraph.
  2. Split the description into functions, each one job with a verb.
  3. Grade each function from E0 to E4, with a link for every piece of evidence.
  4. Decide per function: build, test, park or drop. Before any experiment runs, write its kill criterion: the result that ends the idea.

That is the whole business idea validation process in outline, and it works the same way whether you validate ideas one at a time or compare three. The step-by-step guide linked at the top covers each stage in detail. If you don’t have an idea yet, generating business ideas starts one step earlier.

An idea validation template you can copy

An idea validation template is a table with one row per function and columns for the evidence, its level and the test that could kill it. Above the table, write four lines: who the product is for, the problem in their words, what they do about it today, and where they would find your product (a marketplace, a browser store, a search). The evidence in the two rows below, and the first row’s test, come from the sample report.

Function Evidence found (with links) Distinct people Level Direct or adjacent Next test Kill criterion
Compares stock levels after a failed sync 1 pain, 2 quotes, 190 searches/mo 2 E2 Near match: the pain is overselling Interview 10 merchants Fewer than 3 of 10 express strong interest
Emails the bookkeeper a daily discrepancy list No pain passed the similarity floor 0 E0 None Park Raise only if a merchant asks for it

Fill the “Distinct people” column even when the number is small. Two loud posts from one person look like a market until you count. The kill criterion goes in before the test runs, so a weak result can’t be argued into a pass afterwards.

Idea validation examples from real data

Three idea validation examples from VettedGaps pages show different evidence shapes. We re-read all three on 2026-11-20, and the counts move: card 4136 had 304 posts from 256 people when our step-by-step guide quoted it a day earlier. Only the first level came from the validator; the other two are our reading of each page’s numbers on the scale above.

Example What was observed Strongest level What it still lacks
Multi-warehouse stock sync, Shopify App Store 4 functions checked against 254 pain clusters (groups of complaints about one problem); 1 function with direct evidence (2 people, 190 searches/mo) E2 Proof of payment; 3 functions at E0
Squarespace site export 10 people complained, about 2,220 Google searches a month, 37+ paid Fiverr orders with a median order of $92 E3 Payment for an exit-specific tool
Meta ad spend for dropshippers, Shopify App Store 307 posts from 258 people, demand score .94, explicit willingness to pay E1 A product idea and its own checks

The first verdict was Needs proof, with medium confidence and partial evidence. It also warned that the one supported function rests on a multichannel overselling pain, so the multi-warehouse angle may be narrower than the evidence suggests.

The Squarespace example is the strongest of the three, because three independent signals point at the same job. One complaint reads: “300 a year for a site you barely touch is wild when free options exist.” (source) Even so, the idea page says no paid orders for an exit-specific tool were observed, so a price test is still needed.

The third row is a pain card, not an idea. Its demand score is a number from 0 to 1 that weighs reach, recency, engagement and willingness to pay. It shows strong E1 evidence before anyone has proposed a product: one dropshipper wrote, “At these CPM numbers I can’t see this ever being profitable.” (source)

FAQ

How to check idea validity without spending money?

Validating an idea for free starts with reviews, forum threads and Reddit: save each link and count distinct people. Check freelance marketplaces for the job done by hand. Then interview people from those threads. Low cost idea validation costs time, and only a paid pilot needs money.

Can a tool validate my idea for me?

Our Validate idea tool checks each function for complaints, paid work, search volume and competitors, and says so when nothing supports the idea. It can’t run your interviews or your commitment test. What AI validators can and cannot check is covered in AI tools for validating product ideas.

How much evidence is enough to start building?

It depends on what you are about to commit. A function with E3 or E4 evidence plus a passed commitment test is a reasonable bar for a first build; quitting a job or raising money needs your own buyers paying. The stage-by-stage rule is in startup idea validation.

Does this framework work for ideas that are not software?

Yes, the levels apply to any idea, and only the sources change: product reviews, local groups and customer conversations. Our complaint corpus covers software only, so for a bakery or a skincare line you fill the template by hand.

Start your idea validation with one function

Idea validation turns a pitch into a list of functions with evidence next to each one. Pick your idea, write its functions into the template above and grade them tonight. To validate your idea against real marketplace complaints, paid demand and search volume, run it through Validate idea, our own tool and part of the paid VettedGaps Pro plan: $15 a month with 5 validations a day (plans).

Other ways people search for this

Some people type how to check validity in idea, or ask for an idea validity check. Both are the same question as the FAQ above: grade each function by what you can observe and link every piece of evidence.

Others search for idea validation customer development methodology. That search is usually about talking to customers before building: problem interviews and commitment tests. Those methods are covered in how to validate a product idea. For software and micro SaaS ideas checked before any code, read how to validate SaaS ideas.