Every validation playbook we have, organised by what you are actually trying to decide. Start with the framework, or jump straight to the path that matches your situation.
Idea validation is the practice of gathering evidence that a problem is real and already costing people money, before you build a solution. This page is the index to every validation playbook we have, arranged by the decision you are actually trying to make rather than by publication date.
There is a reason this library exists rather than a single article. Validating a B2B SaaS idea and validating a consumer app are genuinely different jobs with different evidence sources, and a solo founder with no audience faces a different constraint than someone with a mailing list. One generic guide serves none of them well.
Validation rests on three independent proofs. Miss any one and you are guessing with extra steps.
1. The problem is documented at scale. Not that you believe it exists, but that you can find at least twenty unconnected strangers describing the same frustration in their own words, in places they had no reason to expect you were reading.
2. A budget already exists. Someone is already paying to solve this, whether through a competing tool, a freelancer, or an employee’s hours. Absence of spend is a red flag, not an untapped market.
3. A stranger acts. Cold traffic converts, or better, someone pre-pays. Warm traffic and friendly opinions do not count, because the people giving them are being kind.
Friends saying they like it. A large total addressable market. An AI confidence score. Survey respondents saying they “would definitely use this.” Every one of these produces a number that feels like progress while telling you nothing about whether a specific person will pay a specific price.
The distinction between market and product validation matters here too. Market validation proves the problem is real and funded. Product validation proves your particular solution wins. Market validation comes first and is far cheaper. Running them in the wrong order is the most expensive mistake in this whole library.
If you read one thing, read the framework. Everything else in this library is a deeper cut on one of its steps.
| Your situation | Start with | Why |
|---|---|---|
| I have a specific idea | The validation framework | Eight steps in order, with kill thresholds at each |
| I want a fast verdict | The idea evaluator | Scores your idea and shows the complaints behind the score |
| I have no idea yet | Start from complaints | A complaint already contains demand evidence a brainstorm lacks |
That last row is the counter-intuitive one and worth dwelling on. Most founders brainstorm first and validate second, which means they spend their validation effort trying to rescue an idea they are already attached to. Starting from a documented complaint inverts the order: the evidence comes first and the idea is derived from it. For the mechanics see how to find startup ideas in 2026 and how to brainstorm business ideas if you prefer the traditional direction.
Pick the path matching your situation. Each is three to four playbooks in a deliberate order.
Your binding constraint is reach, not ideas. This path front-loads the evidence work so you do not spend your one shot on the wrong problem.
You can build, so your risk is building the wrong thing quickly and well. This path is about finding the gap incumbents leave.
Your constraint is time and a single pair of hands. This path optimises for cheap, fast, reversible tests.
No single source validates an idea. Convergence does. A problem that shows up in software reviews and community threads and paid freelance work is real in a way no landing-page metric can match. These are the seven independent signals we triangulate across, each with what it proves and what it cannot.
| Signal | Proves | Cannot prove |
|---|---|---|
| Software review complaints (270,000+ Capterra, 150,000+ G2) | The problem is documented and severe | That anyone will switch |
| Community threads (160+ subreddits) | The problem in the customer’s own words | Willingness to pay |
| App store reviews (130,000+) | Where mobile products fail users | Anything outside that app’s niche |
| Freelance job posts (5,300+) | People already pay to solve it manually | That software beats a human |
| Verified revenue (8,600+ startups) | What comparable businesses earn | Your outcome, survivorship applies |
| Category saturation (30,000+ companies) | How crowded the space already is | Whether the space is contestable |
| Acquisition listings (650+) | What these businesses sell for | Closed prices, these are asks |
The practical rule: two independent sources plus one behavioral test. Two sources establish the problem is real and funded. The behavioral test, a cold-traffic pre-sale or a paid hand-delivered pilot, establishes that someone will actually act. Full method in the multi-signal validation framework.
The same idea needs different evidence at different points. These are the deeper cuts, ordered by where you are.
Before validation there is selection. Finding SaaS ideas from real pain points covers the complaint-first method, and the pain-point tools roundup compares the instruments. If you want ready-made candidates, SaaS ideas backed by pain points lists ones that already cleared the evidence bar, and niche SaaS ideas narrows by category.
The heart of validation. The core framework covers steps three through five, and the eight-stage framework is the longer, more structured version. Validating niche viability handles the case where your market is narrow by design.
A real problem in a saturated category is still a bad bet. SaaS market saturation in 2026 covers how to read crowding, and the Stripe Index database shows who already operates in your niche. Existing competitors validate that budget exists; what you need to judge is whether the space is contestable.
Validate the business, not just the idea. Revenue benchmarks show what comparable businesses actually earn, MRR versus ARR versus TTM revenue clears up the metrics, and the state of SaaS acquisitions covers what these businesses eventually sell for.
Constraints differ more than methods do. A technical solo founder and a non-technical founder with an audience should validate the same idea in different orders.
Every playbook here is built on the same corpus rather than on received wisdom. It spans 1M+ documented complaints across eleven independent sources, continuously expanded through automated pipelines. Two figures from it are worth internalising before you start.
Most ideas do not survive contact with evidence. Across 22,000+ recorded judgements on researched ideas, 80.5% were rejected outright. Notably, rejections took longer on average than approvals, about 8.6 minutes of deliberation against 6.7. Saying no is the considered decision, not the lazy one.
Validated businesses are usually small. Across 3,700+ startups with verified recurring revenue, the median is roughly $145 per month, with the 90th percentile near $5,107. Validation improves your odds of building something wanted; it does not promise scale. That asymmetry is exactly why spending two weeks on evidence beats spending six months on code.
Explore the underlying data directly in the pain points database, revenue intelligence, freelance demand analysis, and the Stripe Index. The complaint analysis platform guide explains how the sources fit together, and the pain-points database guide covers day-to-day use.
Every figure on this page was re-queried on August 5, 2026 and rounded down to a stable floor, because the corpus grows continuously. The limitations are worth stating plainly, because a validation library that hides them is not much better than a confidence score.
Review and community data is directional, not transactional. A complaint proves frustration, not willingness to pay. Only the freelance-demand and revenue layers speak to money, and each has its own skew: freelance posts capture budget but not whether software beats a human, and revenue data suffers survivorship bias because every business in it shipped and charged. Saturation counts come from Stripe’s public directory, so categories that do not run on Stripe are under-counted. Acquisition figures are asking prices rather than closed transactions. Our idea-reaction data measures what founders say they would pursue at first read, which is appetite rather than an outcome, and our audience is founder-heavy rather than representative of end customers.
This is why the standard here is convergence rather than any single number. Full per-source disclosure sits in the core framework’s methodology table.
There is a growing category of tools that will validate your idea in under two minutes and return a confidence score. They are useful for triage and are not evidence, because no new information about the world enters the system when a language model summarises what it already believes.
The question worth asking any of them: can you show me the underlying complaint, from a named source, with a date? If yes, you have something you can verify. If it only returns a number, you have bought confidence.
Comparisons, written to be fair rather than flattering: the best idea validation tools for 2026, AI idea validators scored on evidence, IdeaProof alternatives, BigIdeasDB versus IdeaProof, and BigIdeasDB versus ValidatorAI. Free instruments: idea evaluator, market size calculator, niche finder, and competitor finder. Product reference lives in the SaaS idea validation tool guide.
Four failure modes account for most wasted validation effort, and all four feel productive at the time.
Validating the solution instead of the problem. Showing people your idea and asking if they like it tests your pitch, not the market.
Only sampling friendly traffic. Your network is being kind. Validation requires strangers with no reason to spare your feelings.
Treating signups as proof. An email address is free to give. Until money changes hands you have measured curiosity, not intent.
Never finishing. Validation running past three months is avoidance dressed as rigor. Run the pre-sale this week.
Longer treatments in common validation pitfalls, lessons from failed business ideas, and how to stop delusional thinking taking over.
BigIdeasDB indexes 1M+ documented complaints across eleven sources, plus verified revenue on 8,600+ startups and saturation data on 30,000+ companies. Every score traces back to a complaint you can read.
Idea validation is the practice of gathering evidence that a problem is real and already costing people money, before building a solution. It rests on three proofs: the problem appears repeatedly in documented complaints from strangers, a budget already exists because people pay for a workaround, and at least one stranger acts, ideally by paying before the product exists. Opinions from friends, market-size estimates, and AI confidence scores are not validation.
Start with the core framework, which walks the eight steps in order from naming the technology shift through delivering the outcome by hand. If you already have a specific idea and want a fast verdict, run it through the free business idea evaluator instead. If you have no idea yet, start from documented complaints rather than brainstorming, because a complaint already contains the demand evidence a brainstormed idea lacks.
Convergence across independent sources matters more than any single score. A problem documented in software reviews, discussed in community threads, and paid for in freelance job posts is real in a way no landing page metric can match. The practical bar is at least two independent sources plus one behavioral test, meaning a cold-traffic pre-sale or a paid hand-delivered pilot.
Market validation proves the problem exists, is painful, and has a budget attached. Product validation proves your specific solution is the one people will pay for. Market validation comes first and is much cheaper because it can be done from documented complaints without writing code. Running them in the wrong order is the most common and most expensive validation mistake.
Yes, and the free path is genuinely good, just slower. Search community threads with the customer’s own words, read one-star and two-star software reviews, and search freelance marketplaces for jobs where people pay to solve the problem by hand. That gives you problem and budget evidence at zero cost. The paid shortcut is having those sources already indexed and scored so the research takes an afternoon instead of three weeks. Start with the pain points database or the discovery tool.
BigIdeasDB Research. (2026). Idea Validation: The Complete Playbook Library. BigIdeasDB. Retrieved from https://bigideasdb.com/idea-validation