Most ideas die of politeness. You describe the concept to friends, they say "that's great, I'd use that," and you spend six months building something nobody pays for. Validation is the discipline of replacing those polite opinions with evidence — and you can get most of the way there in 30 days, without writing a line of code or spending money on product development.
Week 1: Write down what would kill it
Before you collect any evidence, decide what you're testing. Every idea rests on a handful of assumptions, and they are not equally dangerous. List them, then rank them by one question: if this turned out to be false, would the business be dead?
For most ideas the killer assumptions are the same three:
- The problem is painful enough that people already spend time or money on it.
- The specific person you have in mind actually has this problem (not "businesses" — a named kind of person).
- They would pay your kind of price for your kind of solution.
Notice what's not on the list: "can I build it?" For most first businesses, buildability is the safest assumption you have. Test the market, not your skills.
Week 2: Talk to ten real people — badly is fine
Find ten people who match your target customer and ask about the problem, not your solution. The difference matters enormously. "Would you use an app that chases late invoices?" invites politeness. "What happened the last time a client paid you late?" invites the truth.
Three rules keep these conversations honest:
- Ask about the past, not the future. What people did is data. What they would do is theatre.
- Follow the money and the workarounds. If they've cobbled together a spreadsheet, a reminder app, and a recurring calendar event, you've found real pain. If they shrug, you haven't.
- Don't pitch. The moment you pitch, every answer afterwards is contaminated.
Ten conversations sounds small. It's enough to hear the same phrase three times — and that repeated phrase is usually your eventual marketing copy.
Week 3: Build the door, not the house
Now create the cheapest possible thing someone can say yes or no to. Not a product — a commitment test. A one-page site with a price on it. A pre-order link. A "book a pilot" calendar. A concierge version where you do the work manually for the first three customers.
The form matters less than the rule: the test must cost the customer something — money, a deposit, a signed pilot agreement, or meaningful time. Email signups are encouragement, not evidence. A €50 deposit from a stranger outweighs a hundred "I'd definitely use this" comments.
Week 4: Run the numbers against your kill criteria
Before the results come in, write down what counts as a pass — and do it in advance, because afterwards you will be tempted to grade on a curve. Something like: "I contact 50 people, get 10 conversations, and 3 commit to a paid pilot." Concrete, dated, and slightly uncomfortable.
Then honour the result. Three outcomes are possible:
- Pass: you have your first customers and a reason to build. Build the smallest thing that serves them.
- Fail on willingness to pay: the problem is real but not worth money. Reposition (different customer, different price, different wedge) or stop.
- Fail on the problem itself: nobody recognises the pain. Stop. This is a success of the process — you just saved six months.
The point of the deadline
Thirty days is arbitrary but the constraint is not. Validation expands to fill whatever time you allow it, and an open-ended "research phase" is just procrastination with a respectable name. A month forces the only question that matters into the open: did anyone commit?
If the answer is yes, your business plan stops being speculation — every section sharpens, because it's now built on what real people told you and paid you. If the answer is no, the next idea costs you 30 days instead of a year. Either way, you win.