Validating an opportunity before you build

Ideation is generous. Almost any market will yield a plausible idea if you look at it long enough — a technology finding a problem, a regulation creating an obligation, an incumbent too slow to notice a shift in behaviour. Evaluation is where that generosity ends, and it is deliberately the most uncomfortable stage of our process.

It is also the cheapest. The work of evaluation is research, conversation and analysis — expensive in attention, negligible in cash. Everything that follows it costs orders of magnitude more, which is why the stage exists and why we refuse to rush it.

Three tests worth failing

• Demand — is there a problem people already spend money, time or workarounds solving? The strongest signal is not what people say they would do, but what they already pay to do badly.
• Feasibility — can a credible version be built and operated with the capabilities available? Not the perfect version; the version that tests the proposition.
• Route to market — is there a way to reach customers that does not rely on outspending established players? Incumbents win bidding wars; new ventures win by going where incumbents can't or won't.

Each test is worth failing honestly. A demand test that fails tells you the problem was milder than it looked. A feasibility test that fails tells you the capability gap was real. A route-to-market test that fails tells you the economics would never have worked at scale. All three are discoveries you want to make in a workshop, not in a P&L.

Enthusiasm is not evidence

The most common failure mode in evaluation is substituting energy for proof. Founders believe, teams get excited, and a compelling narrative forms — narratives are cheap and infectious, and they are exactly what evaluation exists to distrust.

The correction is to prize disconfirming evidence. Talk to the customers least likely to buy. Price the proposition before anyone has seen a brand. Ask the sceptic in the room to make the strongest case against. An opportunity that survives deliberate attack is genuinely stronger than one that has only ever been championed — and one that collapses under it has saved everyone years.

The output is a decision

Evaluation should end in one of three places: proceed to create, revisit the proposition, or stop. Stopping early is a good outcome. It costs weeks rather than years, and it protects the capacity of the team for opportunities that survive the tests.

Where an opportunity does proceed, everything built afterwards inherits the quality of this stage — the proposition, the first hires, the product scope and the commercial model all trace back to what was learned here. That inheritance is why a venture builder treats evaluation as core work rather than due diligence theatre, and why it is the second stage of the seven rather than a preliminary courtesy.

Evaluation also disciplines the capital side of the model. Because opportunities are tested before ventures are created, the staged investment rounds that follow — from pre-seed through Series A — are released against evidence rather than enthusiasm, a discipline we describe in why capital alone doesn't build great companies.

What this looks like in practice

Before Intelligence Fusion became a company, the underlying opportunity — organisations struggling to turn fragmented open-source information into decision-grade intelligence — was tested against exactly these questions: was the problem frequent and expensive, could a product be built with the available technology and team, and was there a route to customers that did not depend on outspending established defence and security suppliers. The answers held, and the venture proceeded through the remaining stages to growth and eventual exit. The full story is in our Intelligence Fusion case study.

Not every opportunity we evaluate proceeds, and that is the point. The companies in our portfolio are the ones that passed the tests — which is precisely what makes the portfolio meaningful.

A working checklist

For founders running their own evaluation, the essentials compress into a short working list. Speak to at least a dozen people who actually have the problem — not people who find the idea interesting. Find out what they currently spend, in money or effort, to live with it. Establish who else is solving it and why their customers stay. Identify the single riskiest assumption and design the cheapest test that could break it. And decide in advance what result would make you walk away, because that decision is much harder to make honestly once you are invested in the answer.

If the checklist feels like it slows things down, that is the point. The weeks spent here are the cheapest risk-reduction available anywhere in company creation — and the discipline transfers directly into how the venture will treat capital later, which is why investors respond so well to teams who have done it properly.

And if you would rather not run it alone, this is precisely the work we do at the front of every venture we build — our own spinouts, founder partnerships and corporate ventures alike. The evaluate stage is where a conversation with us usually becomes concrete.

One final note on timing: evaluation is not a phase you pass once and forget. Ventures that scale well keep re-validating — new segments, new propositions, new pricing — using the same tests at each expansion. The discipline that decides whether a company should exist is the same discipline that decides what it should do next, which is why it sits so early in a process designed to run for years.

Whether you are a founder with a domain insight or an executive with a strategic hunch, the door in is the same: bring the problem, and let the process decide what it deserves to become.