· 6 min read
Validate an idea before you build: the four-question method
Before the website, the logo and social media: who has the problem, how they solve it today, who else is working on it, and which first test will provide the proof.
It is tempting to start with what shows: a name, a logo, a website, social media accounts. It is concrete, it is pleasant, and it feels like progress. But none of these answers the only question that matters at the start: does anyone want what you offer badly enough to pay for it?
Validating an idea does not mean proving it will work. It means reducing, as quickly and cheaply as possible, the risks that would kill it. Here are the four questions we ask of every project, the ones Elpis’s Research agent documents in its validation report.
1. Who has the problem, exactly?
“Small businesses” or “parents” are not a target: they are millions of people with different needs. A good answer fits in one precise sentence: owners of independent hair salons who lose appointments because they cannot answer the phone during a haircut.
The more precise the target, the easier it is to find, to talk to, and to check whether the problem really exists for them.
2. How do they solve it today?
A problem with no current solution is often a problem nobody is trying to solve. Conversely, people who cobble something together — a spreadsheet, a WhatsApp group, an intern, a tool that is too expensive — show real demand.
Look for concrete signals: questions asked on forums, negative reviews of existing solutions, job offers that describe this work, budgets already being spent.
3. Who else is working on it?
Having competitors is rather good news: it proves a market exists. The real question is what sets you apart for your target: a price, a specialisation, a channel, a quality of service.
The Elpis validation report lists the competitors found, their positioning and their prices when they are public, with a link to each source. When a piece of information is missing, it says so instead of making it up.
4. Which first test will provide the proof?
Desk research, even well sourced, is still desk research. It does not replace a conversation with a real customer. That is why the 30-day plan ends with a measurable first demand test, for example:
- twenty sign-ups on the waiting list of your home page;
- five conversations with potential customers, at least two of whom ask for a price;
- a first pre-order.
If the test fails, it is not a failure: it is information, obtained before building everything.
An honest score, even when it disappoints
The report ends with a score out of 100 and a verdict: Go for it, Adjust or Risky. When an idea looks fragile, we say so, and we suggest ways to adjust it. It is not pleasant to read, but it costs far less than finding out six months later.