Save this part
The short version
- Describe one specific customer and painful situation before describing the product.
- Classify the idea for that customer: the same product can be a painkiller for one segment and a vitamin for another.
- Look for recent behaviour, existing workarounds, spending, and consequences-not polite encouragement.
- Validate distribution by identifying where the first 20 customers can actually be reached.
- Use a score to expose weak assumptions, then test the riskiest one before building a full product.
A startup idea is not validated because people call it interesting. It becomes less risky when real behaviour shows that a specific, reachable customer experiences the problem often, dislikes the current alternative, and will make a meaningful commitment to solve it.
Prefer the 60-second version?Watch the original painkiller-versus-vitamin startup reel↗Step 00
Write the idea in one specific sentence
If the idea needs a long explanation, the customer and problem are probably still vague. Remove technical language, feature lists, and claims about changing an industry.
Use this structure: “[Specific customer] needs a better way to [job] because [current pain or cost].” It forces the sentence to name who, what, and why before proposing how.
- Weak: “An AI platform for the future of productivity.”
- Stronger: “Small support teams need a faster way to draft repetitive email replies because backlogs delay paying customers every day.”
A clear sentence is not validation. It is the first test of whether the idea is precise enough to validate.
Step 01
Is the idea a painkiller or a vitamin for this customer?
A painkiller addresses a problem customers already notice and actively try to solve. Delay has a meaningful consequence, so urgency and willingness to pay can exist. A vitamin improves convenience, enjoyment, identity, or quality of life, but the customer can postpone it with limited immediate damage.
This is not a permanent label attached to a product. Canva may be optional for someone making one invitation and operationally important for a creator or marketing team producing assets every day. The same product can be a painkiller for one segment and a vitamin for another.
- What happened the last time the customer did nothing?
- Does the problem occur yearly, monthly, weekly, daily, or multiple times per day?
- Does it cost time, revenue, customers, opportunity, reputation, safety, or compliance?
- Does the customer search for a solution without being prompted by your pitch?
- Have they created a workaround or paid for an imperfect alternative?
Painkillers are not guaranteed businesses, and vitamins are not automatically bad. Vitamins often need exceptional habit, brand, entertainment, community, or distribution. Pain only helps when the affected customer is reachable and able to pay.
Examples with context
The product name alone does not determine the category
Use these examples to practise the classification. The same product can move between categories when the customer, frequency, deadline, or consequence changes.
| Product | Customer and situation | Classification | Reason |
|---|---|---|---|
| Stripe | Online business that cannot otherwise accept customer payments | Painkiller | Without a payment path, the business cannot collect revenue. |
| Calendly | Recruiter coordinating dozens of interviews each week | Painkiller | Repeated scheduling creates operational delays and avoidable back-and-forth. |
| Resume builder | Active job seeker applying before near-term deadlines | Painkiller | The task is urgent and tied to employment opportunities. |
| Canva | Someone creating a one-off birthday invitation | Vitamin | It improves convenience, but delay has limited consequences. |
| Canva | Marketing team producing campaign assets every working day | Painkiller | It supports a frequent workflow tied to deadlines and revenue. |
| Loom | Someone sending an occasional personal explanation | Vitamin | Text, a call, or another recorder can usually replace it. |
| Loom | Distributed team replacing recurring status meetings with async video | Painkiller | It can remove repeated coordination cost across the team. |
| Password manager | Casual user who currently reuses memorable passwords | Vitamin at first | Security value is real, but the user may not feel immediate urgency. |
| Password manager | IT team managing employee credentials and access risk | Painkiller | Credential failures can create security, recovery, and compliance costs. |
| Wallpaper app | Phone user who wants a different visual style | Vitamin | It improves aesthetics without resolving a costly recurring problem. |
Do not copy these labels into your pitch. Write your own row with a precise customer, triggering situation, current alternative, frequency, and consequence.
Step 02
Name the customer without saying “everyone”
Broad labels such as students, founders, businesses, or creators hide major differences in urgency and buying behaviour. Describe a role, situation, trigger, current alternative, and purchasing context.
- Role: who personally experiences the problem?
- Context: what are they trying to complete?
- Trigger: what makes them look for help now?
- Frequency: how often does the situation occur?
- Buyer: who approves and pays for a solution?
- Constraint: what prevents them from using the obvious alternative?
“Job seekers applying to ten or more technical roles this month” is more testable than “people who need a résumé.”
Step 03
Score frequency and consequence separately
A problem can be frequent but harmless, or rare but catastrophic. Do not combine those into one vague feeling. Ask about the last event, how often it occurs, and what the customer lost because of it.
- Frequency 1: once a year or less.
- Frequency 2: every few months.
- Frequency 3: every month or week.
- Frequency 4: most working days.
- Frequency 5: several times per day.
- Consequence 1: mild annoyance with no meaningful cost.
- Consequence 5: material loss of money, customers, time, opportunity, safety, or employment.
Do not award a five because the problem sounds emotionally important. Record the event, frequency, duration, and consequence that justify the score.
Step 04
Find competitors, substitutes, and evidence of spending
Competition is useful evidence that customers recognise the problem and allocate resources to it. Search for direct products and indirect substitutes: agencies, employees, spreadsheets, templates, consultants, internal tools, and manual routines.
No direct competitor can mean an early opportunity, but it can also mean low demand, an unreachable customer, poor economics, or a problem people tolerate. Treat absence as a question to investigate-not proof of originality.
- Record at least three direct competitors or substitutes.
- Capture their target customer, promise, price, and primary acquisition channel.
- Note whether buyers pay with money, employee time, manual effort, or accepted risk.
- Search Product Hunt and Uneed for launches, G2 and Capterra for B2B tools, AlternativeTo for substitutes, and relevant marketplaces for consumer products.
- Verify pricing and positioning on each company’s own website before relying on a directory entry.
Step 05
Mine negative reviews without copying feature requests
One- and two-star reviews can reveal switching triggers, broken workflows, missing integrations, pricing objections, and customer language. Search G2, Capterra, app stores, browser-extension stores, Reddit, support forums, and issue trackers relevant to the category.
- Copy the exact problem and context, not only the requested feature.
- Group complaints by customer segment and job-to-be-done.
- Record how recently and repeatedly each problem appears.
- Separate product defects from structural gaps the market still fails to solve.
- Check positive reviews too: a proposed improvement must not remove what existing buyers value.
“I wish it had feature X” is not automatically a roadmap. Ask what the reviewer was trying to accomplish and why the current path failed.
Step 06
Prove you can reach the first customers
A painful problem is not enough if you cannot find the affected people at a reasonable cost. Write down the first channel and twenty real prospects before building.
- Developers: relevant GitHub projects, technical communities, Hacker News, conferences, or focused subreddits.
- Designers: portfolios, professional communities, events, agencies, or teams already using adjacent tools.
- Parents: trusted local groups, schools, activity communities, newsletters, or professional referrals depending on the problem.
- Students: course communities, campus groups, Discord servers, subreddits, or search queries tied to a deadline.
- Founders: accelerator networks, Indie Hackers, LinkedIn, X, founder communities, or direct outreach based on a visible trigger.
A list of platforms is not a distribution plan. Name the community, search query, partnership, audience, or outbound filter that reaches this exact customer.
Step 07
Interview recent behaviour, not future intention
Recruit people who recently experienced the situation. Ask for a walkthrough of the last real event and listen for frequency, urgency, workarounds, spending, decision-makers, and consequences.
- “When was the last time this happened?”
- “Walk me through what you did next.”
- “What was the hardest or slowest part?”
- “What have you already tried?”
- “What do you pay for today?”
- “What happens if you leave it unsolved?”
Avoid “Would you use this?” Past behaviour is evidence; future compliments are predictions.
Step 08
Ask for a commitment stronger than a compliment
Validation becomes more useful when a potential customer gives up something scarce: money, time, access, reputation, internal coordination, or data. Match the commitment to the product’s maturity and risk.
- A follow-up meeting with the actual buyer or decision-maker.
- Access to anonymised examples or observation of the current workflow.
- A pilot with a clear owner, start date, and success condition.
- A refundable deposit, pre-order, paid trial, or signed commercial commitment when appropriate.
A waitlist click is an interest signal. It is not equivalent to payment or repeated usage.
Step 09
Test whether you can deliver the outcome quickly
Do not estimate the time required to build the final company. Estimate the smallest version that can produce the promised outcome for one well-matched customer. That may be a prototype, spreadsheet, manual concierge service, integration, or narrow application.
- Can you test the riskiest assumption in days rather than months?
- Which parts can be performed manually before they are automated?
- Do security, regulation, data access, hardware, or platform review make a fast test unrealistic?
- Can you deliver enough value to ask for a meaningful commitment?
- Will the test teach you something even if the result is negative?
Momentum matters, but a universal three-day MVP rule is misleading. The right scope is the fastest ethical test that produces credible evidence for this market.
Decision point
Score the startup idea out of 25
Score each dimension from one to five and write the evidence beside it. The total is a comparison tool, not a scientific forecast. A score of 16 can be a useful personal review threshold, but weak pain or zero distribution should not be hidden by high enthusiasm in other categories.
- Pain severity and frequency: how costly and recurring is the problem?
- Reachable customer and distribution: can you identify and contact buyers efficiently?
- Spending or commitment evidence: what do customers already invest to solve it?
- Speed and feasibility: how quickly can you run a credible first test?
- Founder advantage and motivation: do you understand, reach, or serve this market unusually well?
STARTUP IDEA
Describe the idea in one sentence:
[SPECIFIC CUSTOMER] needs a better way to [JOB] because [CURRENT PAIN OR COST].
CUSTOMER
Who experiences this problem most intensely?
What event makes the problem urgent?
Where can you reach the first 20 people?
PAINKILLER OR VITAMIN
What happens if they do nothing?
How frequently does the problem occur?
What do they lose: time, money, customers, status, safety, or opportunity?
CURRENT BEHAVIOUR
How did they solve it the last time?
What tools, services, employees, or workarounds do they use?
What do they already pay?
MARKET EVIDENCE
Three competitors or substitutes:
1.
2.
3.
Recurring complaints from negative reviews:
1.
2.
3.
FIRST TEST
Riskiest assumption:
Smallest test:
Success threshold:
Decision date:
SCORE — 1 TO 5
Pain severity and frequency: __ / 5
Reachable customer and distribution: __ / 5
Evidence of spending or commitment: __ / 5
Speed and feasibility of the first test: __ / 5
Founder advantage and motivation: __ / 5
TOTAL: __ / 25
DECISION
Pursue / revise / pause / reject:
Evidence behind the decision:Never use a strong total to ignore a fatal assumption. Write what would have to become true before a low-scoring dimension can improve.
Final review
Check the strongest red and green flags
Use these signals to decide whether the next step is to pursue, revise, pause, or reject the idea.
- Red: the idea cannot be explained without vague language or technical novelty.
- Red: no specific customer, trigger, reachable channel, current workaround, or spending evidence.
- Red: the business needs enormous scale before one customer can be served profitably.
- Red: founder excitement is about the technology while customer evidence remains weak.
- Green: the same costly problem appears independently in recent customer stories.
- Green: customers already spend money, time, headcount, or reputation on alternatives.
- Green: you know where the first twenty prospects are and can contact them now.
- Green: the riskiest assumption can be tested with a small, defined experiment.
Common questions
Startup idea validation FAQ
Should a startup be a painkiller or a vitamin?
Painkillers are often easier to sell because customers already feel urgency, but the label depends on the customer and situation. Vitamins can still become strong businesses through habit, identity, entertainment, brand, or exceptional distribution. Test frequency, consequence, spending, and reachability instead of relying on the label alone.
How many customer interviews are enough to validate a startup idea?
There is no universal number. Start with five to ten well-matched people in one segment and look for repeated patterns in recent behaviour. Continue until conversations stop materially changing your understanding, then test demand with a real commitment.
Does finding competitors mean my startup idea is bad?
No. Competitors and substitutes can show that customers recognise the problem and spend resources on it. Study who buys, why they switch, what they dislike, and which segment remains poorly served. No competitors should trigger more research, not automatic excitement.
Should I build an MVP before talking to customers?
Usually, no. First learn whether the problem is frequent, consequential, and already produces action. Then build the smallest ethical test needed to answer the next uncertain question; that may be a prototype, manual service, or landing page rather than production software.
Does a waitlist prove demand?
A waitlist is an early interest signal, not proof of purchase intent. It becomes more useful when sign-ups match the target audience and take a higher-effort step such as booking a call, sharing a workflow, joining a pilot, or paying a deposit.
What is a good startup validation score?
A score such as 16 out of 25 can be a personal comparison threshold, not a universal rule. Record evidence for every point and inspect each dimension separately. Severe weakness in pain, customer access, willingness to pay, or feasibility can invalidate a high total.
Found something worth remembering?
Save the useful idea in Offload with the context that made it matter. Ask for it naturally when you’re ready to use it.
Try Offload free