guides
Idea Validation hubIdea Validation Kit
Validate the problem before you build the app.
This kit is for beginner solo founders using AI tools to build profitable apps. The goal is not to prove your idea is genius. The goal is to find one painful, specific problem that real people already try to solve.
If the idea passes, your next step is to build the smallest Simple, Lovable, Complete version and try to earn your first dollar. If it does not pass, you saved yourself from building the wrong thing.
TL;DR
- Pick one narrow customer.
- Write the problem in one sentence.
- Score it with the Problem Scorecard.
- As a lightweight first pass, collect enough real examples to hear repeated language.
- Start with about 5 potential-customer conversations.
- Scan a handful of competitors or workarounds.
- Decide: kill, pivot, or continue.
- Only build the next smallest offer test if the evidence says people already care.
What Validation Actually Means
Validation does not mean asking people, "Would you use this?"
Most people are polite. They will say yes because it costs them nothing. Better validation looks for evidence that the problem already exists.
Use an evidence ladder, not a vibe check:
| Evidence Level | Examples |
|---|---|
| Weak | Likes, compliments, waitlist joins, "I would use this," hypothetical willingness to pay |
| Better | Specific recent pain, current workaround, budget/current spend, repeated manual effort, call booked |
| Stronger | Paid pilot, deposit, preorder, first payment, repeated usage, referral, buyer asks for a payment link |
The question is not: "Do people like my idea?"
The question is: "Is this problem painful enough that a specific person is already trying to solve it?"
Five interviews will not prove demand. They can be enough to surface early confusion, weak pain, repeated language, and whether this is worth a small build-and-sell test.
The One-Sentence Idea
Before research, write the idea like this:
I help [specific customer] solve [specific painful problem] so they can [valuable outcome].Examples:
I help solo newsletter writers turn messy sponsor notes into polished ad reads so they can publish faster and avoid rewriting brand copy.I help small gym owners reply to missed-call leads so they can book more trial sessions without hiring a receptionist.I help Etsy sellers rewrite weak product titles using buyer search language so they can improve listing quality without learning SEO.If the customer is "businesses," "creators," "students," or "everyone," narrow it before continuing.
Problem Scorecard
Score each category from 0 to 3.
This scorecard is a decision aid, not a scientific model. Use it to expose weak assumptions, then verify with conversations and paid tests.
| Category | 0 Points | 1 Point | 2 Points | 3 Points |
|---|---|---|---|---|
| Specific customer | Everyone or vague group | Broad segment | Clear niche | Clear niche with reachable communities |
| Pain frequency | Rare | Monthly | Weekly | Daily or several times per week |
| Pain intensity | Mild annoyance | Noticeable frustration | Costs time, money, or opportunities | Urgent, expensive, embarrassing, or revenue-linked |
| Current workaround | No workaround | They complain but do nothing | Manual spreadsheet, assistant, template, or hack | They already pay for a partial solution |
| Search behavior | No evidence | A few vague posts | Repeated questions in communities | Many specific posts with strong language |
| Payment evidence | No budget | Might pay someday | Pays for adjacent tools | Already pays, has budget, or problem is tied to revenue |
| Founder access | You know nobody like this | Can reach them cold | Can reach communities | Can talk to 5+ this week |
| Build scope | Requires complex integrations or marketplace | Large app with many workflows | 2-3 screens | One workflow, buildable in a weekend |
| Differentiation | Same as existing tools | Slightly cheaper or faster | Clear niche or workflow angle | Clear wedge competitors ignore |
| First-dollar path | No obvious buyer action | Free user first, maybe paid later | Simple paid plan possible | Can ask for pre-order, paid setup, or first month now |
Scoring
These cutoffs are deliberately rough heuristics for forcing a decision, not statistically validated thresholds.
| Total Score | Decision | What To Do |
|---|---|---|
| 0-12 | Kill | The problem is too vague, weak, or hard to reach. Pick a narrower customer or a more painful problem. |
| 13-20 | Pivot | There may be something here, but the current version is not sharp enough. Narrow the customer, workflow, or evidence. |
| 21-26 | Continue cautiously | Good enough to run interviews and a small offer test. Do not build the full product yet. |
| 27-30 | Run the SLC offer test | Strong signal. Confirm with interviews, then build the smallest complete version and ask for money quickly. |
A high total score does not save an idea with zero customer access or zero pain intensity. If either of those is 0, treat the idea as a pivot.
FAQ
How many interviews do I need before building?
You do not need a perfect sample size before a tiny offer test. Start with about 5 relevant conversations to find repeated language, weak assumptions, and obvious disqualifiers. Keep the first build small enough that being wrong is cheap.
Is a waitlist enough validation?
No. A waitlist is weak evidence. It becomes more useful when the people joining match the target buyer, explain their pain, ask for access, or take a stronger action such as booking a call or paying for a pilot.
Should I research Reddit and public communities?
Yes, as long as you participate respectfully and follow platform rules. Public discussions are useful for language and pain discovery, but they do not replace direct customer conversations.
Sources checked
- Y Combinator, Startup School
- Indie Hackers, Home
- Reddit, Rules
Members get the full resource
Unlock the complete walkthrough
- Full implementation notes, examples, checklists, and templates.
- Private Discord access for feedback, accountability, and support.
- Monthly Startup Club resource drops for solo founders building with AI.