Building something nobody wants is the most common reason startups fail. CB Insights found that 42% of failed startups created a product with no market need1.
Not because the code was bad. Not because the founder didn't work hard. Because nobody asked for it.
The fix isn't working harder. It's validating smarter.
This framework gives you seven sequential phases to test your idea at increasing levels of commitment. Each phase costs more—but proves more. You stop when the evidence stops you.
Core principle: Prove demand before you build. Validate through behavior, not opinions.
The 7 Validation Phases
| Phase | Name | What You Validate | Cost | Time |
|---|---|---|---|---|
| 1 | Problem Discovery | Real pain, documented | $0 | 1–2 weeks |
| 2 | Market Sizing | Worth pursuing? | $0 | 1 week |
| 3 | Solution Fit | Your approach resonates | $0 | 1–2 weeks |
| 4 | Budget Validation | Will people pay? | $0–$500 | 2–4 weeks |
| 5 | Landing Page Test | Will strangers click? | $50–$100 | 1–2 weeks |
| 6 | Pre-Sell / Waitlist | Will they commit? | $0 | 2–4 weeks |
| 7 | First 10 Customers | Real feedback, real revenue | $500–$5,000 | 4–8 weeks |
TL;DR: Start cheap. Stop early if signals are red. Only spend more when earlier phases prove demand.
TL;DR: Start Here
| Your Situation | Start Here | Why |
|---|---|---|
| On a budget? | Phase 4: Budget Validation | Test willingness to pay without spending money |
| Need users fast? | Phase 7: First 10 Customers | Prove demand with real humans |
| Not sure it's real? | Phase 1: Problem Discovery | Confirm pain before solving |
| Have a prototype? | Phase 5: Landing Page Test | Test if strangers care |
Phase 1: Problem Discovery
Duration: 1–2 weeks
Cost: $0
Input: Your idea, target customer in mind
Output: 20+ complaint threads + 10 interview notes with repeating language
The lean startup methodology starts with one question2: What problem are you solving for whom?
How to validate:
- Search communities: Reddit, Hacker News, X, niche forums. Look for recurring complaints—not single mentions.
- Run 10–15 interviews: Ask "Walk me through your last experience with [current solution]"3. Not "What's your biggest problem?"
- Measure intensity: Rule-of-thumb—you need ~30% showing high problem intensity—people actively frustrated and actively looking3.
Stop / iterate gate:
- ✅ Proceed if: 20+ complaint threads found, 10+ interviews with repeating pain language
- ⚠️ Iterate if: Few threads, weak interview responses → refine target customer
- ❌ Stop if: No community discussion, "nice to have" responses only → pivot or kill
Red flags:
- No community discussion = no problem worth solving
- "Nice to have" responses = not painful enough
- "Everyone would want this" = nobody specifically needs it
What good looks like:
"I've talked to 12 founders in the last 2 weeks. Eight of them explicitly complained about X. Three have already tried to build workarounds. One called it 'the thing that keeps me up at night.'"
[→ Deep dive: Problem Discovery Guide] (coming soon)
Phase 2: Market Sizing
Duration: 1 week
Cost: $0
Input: Problem validated from Phase 1
Output: Competitor list + search volume estimate + TAM range
Goal: Is this big enough to pursue?
How to validate:
- Search volume: Google Keyword Planner, Ubersuggest (free tier)
- Competitor presence: 2+ competitors with traction = validated market
- TAM estimate: Top-down from industry reports
Stop / iterate gate:
- ✅ Proceed if: 2+ competitors with traction, reasonable search volume
- ⚠️ Iterate if: <100 search volume → explore B2B/communities, not kill
- ❌ Stop if: Zero competitors + zero search = potential opportunity OR no market (requires deeper digging)
Important caveat for B2B:
Search volume <100/month can still work in B2B if distribution is outbound/communities. Treat low search volume as a caution flag, not a kill switch4.
Quick benchmark:
| Search Volume | Consumer Verdict | B2B Verdict |
|---|---|---|
| <100/month | Too small | Caution flag—check communities |
| 100–1,000/month | Niche but viable | Viable if outbound works |
| 1,000–10,000/month | Solid market | Good |
| >10,000/month | Competitive but large | Very competitive |
[→ Deep dive: Market Sizing Guide] (coming soon)
Phase 3: Solution Fit
Duration: 1–2 weeks
Cost: $0
Input: Problem + market confirmed
Output: Demo/mockup + 5+ prospects interested in paying
Goal: Your solution actually solves the problem.
Eric Ries defined the MVP as "that version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort"5.
How to validate:
- Create a demo: Mockups, wireframes, or a 2-minute video. Not code.
- Test with prospects: Show to the same people from Phase 1
- Measure engagement: Do they say "Where can I get this?"
The key question:
Don't ask "Would you use this?" Ask "Would you pay $X for this?"
Stop / iterate gate:
- ✅ Proceed if: 5+ prospects show strong interest, ask about pricing
- ⚠️ Iterate if: Mild interest, feature requests diverge → refine solution
- ❌ Stop if: "Sounds cool" without commitment, no pricing interest
Red flags:
- "Sounds cool" without commitment
- Feature requests that diverge from core problem
- Interest drops when you explain the trade-offs
[→ Deep dive: Solution Fit Guide] (coming soon)
Phase 4: Budget Validation
Duration: 2–4 weeks
Cost: $0–$500
Input: Solution validated from Phase 3
Output: Survey results: % willing to pay + preferred price point
Goal: Will people pay—and how much?
This is the critical gate. Most indie hackers should not write code past this phase unless they can show clear willingness to pay6.
Important distinction: This is a willingness-to-pay survey (pre-launch). It's different from the PMF survey (post-launch) developed by Sean Ellis, which measures retention among existing users7.
How to validate:
- Survey with price points: "Would you pay $X? What about $Y?"
- Offer pre-orders at 2–3 price tiers (or a refundable deposit)
- Track commitment: There's a huge difference between "I would buy this" and "Take my money now"3
The 30% rule (rule-of-thumb):
If <30% say they'd pay anything, the idea needs rethinking.
Stop / iterate gate:
- ✅ Proceed if: 30%+ willing to pay at your price point
- ⚠️ Iterate if: <30% → try different positioning/pricing or pivot
- ❌ Stop if: <10% willing to pay → kill or major pivot
Price research:
Use ValidSpark to see what competitors charge. Test 2–3 price points.
Illustrative example: One founder (Steve) had a VR platform idea. Initially targeted developers at $50/month. Through validation, discovered architects would pay $5,000/month for the same tech—100x revenue from repositioning3.
→ Read: How to Validate on Any Budget
Phase 5: Landing Page Test
Duration: 1–2 weeks
Cost: $50–$100 (ad spend)
Input: Pricing validated from Phase 4
Output: Live page with conversion data
Goal: Will strangers click and sign up?
How to validate:
- Build a page: Carrd, Leadpages, or custom. One page, one CTA
- Run ads: $50–$100 on Meta or Google
- Measure conversion: Signups ÷ visitors
Typical heuristic benchmarks:
| Conversion Rate | Verdict |
|---|---|
| >5% | Strong—messaging resonates |
| 3–5% | Good—valid interest |
| 1–3% | Average—test variations |
| <1% | Problem—messaging unclear |
Stop / iterate gate:
- ✅ Proceed if: >3% conversion, clear messaging resonates
- ⚠️ Iterate if: 1–3% → test headlines, CTA,hero image
- ❌ Stop if: <1% after 3 variants → major messaging rethink
Red flags:
- High traffic, low conversions = messaging problem
- Low traffic = targeting problem
- High bounces = page doesn't match ad
[→ Read: Landing Page Benchmarks] (coming soon)
Phase 6: Pre-Sell / Waitlist
Duration: 2–4 weeks
Cost: $0
Input: Landing page validated from Phase 5
Output: Pre-orders or committed waitlist signups
Goal: Real commitment, not just interest.
How to validate:
- Launch waitlist: Use Carrd, ConvertKit, or Typeform
- Offer early access: In exchange for feedback
- Track conversion: Email signup ≠ commitment. Pre-order = commitment
The "I'll pay later" test:
When people say "I'll pay later," that's not validation. That's politeness6.
Stop / iterate gate:
- ✅ Proceed if: Pre-orders, paid deposits, signed LOIs
- ⚠️ Iterate if: Only email signups → strengthen offer
- ❌ Stop if: "I'll pay later" responses only → no real commitment
What validates:
- Pre-orders with credit card
- Paid waitlist deposits
- Signed LOIs or contracts
[→ Read: The Credit Card Pre-Sell Method] (coming soon)
Phase 7: First 10 Customers
Duration: 4–8 weeks
Cost: $500–$5,000
Input: Pre-sell validated from Phase 6
Output: 10 paying customers + retention data
Goal: Real feedback, real retention, real revenue.
This is your final proof. PMF is measured, not declared.
Important: The Sean Ellis 40% "very disappointed" benchmark is a post-launch PMF survey—it's for measuring existing users, not pre-launch willingness to pay7.
How to validate:
- Launch to small cohort: 10–20 users, not 1,000
- Interview each personally: Every single one
- Measure retention: After 30/60/90 days
The PMF benchmark (post-launch):
| % "Very Disappointed" | Market Verdict |
|---|---|
| 40%+ | Strong PMF |
| 25–40% | Moderate |
| <25% | Weak—don't scale |
Stop / iterate gate:
- ✅ Proceed if: 10+ paying, >40% "very disappointed," retention growing
- ⚠️ Iterate if: <10 customers, moderate PMF → get more feedback
- ❌ Stop if: Can't acquire 10 customers → reconsider before building more
The goal:
If you can't get 10 paying customers who actively use the product, reconsider before building more.
→ Read: How to Get First 10 Customers Before Code
When to Stop
Each phase is a gate. If signals are red, don't proceed to the next—go back and iterate.
| Signal | What It Means | Action |
|---|---|---|
| <30% willingness to pay (survey) | Weak demand | Pivot or kill |
| No community discussion | No visible problem | Go back to Phase 1 |
| Landing page <1% conversion | Messaging off | Rework value prop |
| "I'll pay later" responses | No real commitment | Keep validating |
| <40% "very disappointed" (post-launch PMF) | Weak PMF | Don't scale yet |
How ValidSpark Helps
ValidSpark accelerates every phase:
- Phase 1–2: Instantly find problem discussions, size your market
- Phase 3: See what features competitors already offer
- Phase 4: Research pricing strategies that work
- Phase 5–6: Generate landing page copy from proven angles
Paste your idea → ValidSpark gives you exactly which keywords to search and which pain-point threads are already ranking.
FAQ
How long should validation take?
12–20 weeks total if you move fast. Each phase is 1–4 weeks. The key is not to rush—each gate proves demand before spending more.
Do I need to do customer interviews?
Yes. Community search helps, but direct interviews reveal pain intensity that Google can't. Budget 10–15 interviews in Phase 1.
Can I validate B2B with low search volume?
Yes. Many viable B2B niches have low Google search volume but strong community/discord presence. Treat low search as a caution flag, not a kill switch—check LinkedIn groups, industry forums, and outbound channels instead.
What's the difference between willingness-to-pay and PMF?
- Willingness-to-pay (pre-launch): "Would you pay $X for this?" — survey prospects
- PMF (post-launch): "How would you feel if you could no longer use this?" — survey existing users
They're measuring different things at different stages.
The Bottom Line
This framework exists because building first is the expensive way to learn.
42% of startups fail because they build something nobody wants1. The difference between the ones that succeed isn't luck—it's systematic validation before writing code.
Start at Phase 1. Work forward. Each phase costs more than the last—but only spend when earlier phases prove demand.
Ready to validate your idea?
- Want the cheapest test? → Phase 4: Budget Validation
- Want the fastest proof? → Phase 7: First 10 Customers
- Not sure the problem is real? → Start at Phase 1
References
Footnotes
-
CB Insights: The Top 12 Reasons Startups Fail — "42% of startups fail because there is no market need for their product or service." ↩ ↩2
-
Lean Foundry: How to Validate Any Startup Idea in 90 Days — Validates through customer behavior, not opinions. 30% high problem intensity benchmark. ↩ ↩2 ↩3 ↩4
-
Viima: What is the Lean Startup Methodology? — B2B often relies less on search volume and more on community/outbound channels. ↩
-
Eric Ries: The Lean Startup — MVP definition ↩
-
Startup Grind: The Ultimate Step-by-Step Guide to Validating Your Startup Idea — "I'll pay later" is not validation. ↩ ↩2
-
Sean Ellis: PMF Survey — "Very disappointed" benchmark is for post-launch measurement among existing users. ↩ ↩2