How we decide whether to build

Most ventures fail before they launch. Often the product itself was sound. The fault lies in a question that was never properly asked.

The standard approach to entering a new market has not changed much in decades. Find an expert. Commission research. Build a business plan. Raise money. Hire a team. Build the product. Then find out whether anyone wanted it.

This process is expensive, slow, and structurally biased toward false confidence. Experts have opinions. Research firms have methodologies that flatten nuance. Business plans are, by definition, documents written by people who want to believe in what they are describing. By the time reality weighs in, the investment is already made.

We do not work that way.

The question comes first.

Every market we consider begins with a single question. Is there genuine, demonstrable demand for better pricing intelligence here, and will the people who need it tell us so directly?

Not in a focus group. Not in a survey designed to produce the answer the researcher hopes for. In the real world, with real outreach, to real practitioners operating inside the market right now.

If the answer is yes, if the market bites, we build. If it does not, we move on quickly and without sentiment. The cost of a clean exit from a validation exercise is trivial. The cost of building something the market did not ask for is not.

This discipline is the founding principle of Try Ventures. It is also, increasingly, what separates ventures that succeed from those that do not. An analysis of over 400 failed ventures found that poor product-market fit was the leading cause of failure, cited in 43% of cases.

Why 2026 changes the equation.

For most of the history of business building, running a proper market validation exercise required significant resource. You needed researchers, analysts, domain specialists, expensive data subscriptions, and months of time. Small teams could not afford to do it properly. Large teams could, but their size and cost created pressure to confirm rather than challenge the thesis.

That calculus has shifted materially.

The tools available to a focused team in 2026 would have been unrecognisable five years ago. Large language models can now extract, structure, and analyse pricing signals from fragmented, unstructured sources (auction records, dealer listings, permit databases, trade publications) at a speed and cost that makes comprehensive market mapping genuinely accessible. Data extraction and enrichment tools that once required significant engineering resource can now be deployed by a small team in days. Lean campaign infrastructure, the ability to build a credible market presence, reach a targeted audience, and measure real response, has never been cheaper or more precise.

McKinsey estimates that generative AI alone could add between $2.6 and $4.4 trillion annually to the global economy, with the majority of that value coming from exactly the kind of knowledge work and data extraction that underpins market intelligence.

The result is that a rigorous market investigation, one that would once have taken a well-funded team six months, can now be completed in a fraction of the time and at a fraction of the cost. The barrier between curiosity and validated insight has collapsed.

What a proper investigation looks like.

We give every market investigation a defined window. Enough time to build a real signal. Short enough to maintain discipline and prevent the kind of slow accumulation of sunk cost that distorts judgement.

Within that window, we run a structured process.

We map the market from the outside in. Transaction volumes, price distributions, the gap between asking prices and clearing prices, the identity and interests of the dominant incumbents. We do this with data, not with conversations that are vulnerable to the enthusiasm of whoever we happen to speak to first.

We identify the intermediary closest to the transaction. The dealer, the broker, the surveyor, the specialist. The person who feels the friction of poor pricing intelligence most acutely every day. We go to them directly, not to sell anything, but to understand whether the problem we have identified from the outside matches the problem they experience from the inside.

We build a lean market presence and run targeted outreach. Not to describe a product. To test a hypothesis. The response rate, the sentiment, and the specific questions people ask tell us more than any research report.

And then we make a decision. Build or move on.

What the data tells us that experts cannot.

There is a particular kind of confidence that comes from a validated demand signal that no expert opinion can replicate.

When a broker in Rotterdam responds to an outreach campaign within hours asking when the tool will be ready, that is a different quality of information than a consultant telling you the market is attractive. When twenty marine surveyors in three countries independently describe the same frustration with existing valuation tools, that is not anecdote. That is a pattern.

We are not dismissive of expertise. Deep domain knowledge is valuable and we seek it out. But we treat expert opinion as a starting hypothesis, not a conclusion. The data confirms or challenges it. We follow the data.

This is not a novel idea. The lean methodology that underpins modern product development has been well established for over a decade. What is new is the degree to which the tools of 2026 allow it to be applied with genuine rigour to the kind of complex, specialist, high-value markets that were previously the exclusive territory of well-resourced incumbents with years of domain experience.

The playing field has levelled considerably. We intend to use that to our advantage.

The ventures that follow.

Every project in the Try Ventures portfolio begins here. A market that fits the corridor. A hypothesis about the demand for better pricing intelligence. A structured investigation with a defined window and a clear decision point at the end.

Some will validate. Some will not. The ones that do will be built with the confidence that comes from a market that has already spoken.

The ones that do not will be closed quickly, cleanly, and without the kind of emotional investment that comes from having committed too early, too heavily, on the basis of too little evidence.

That is the methodology. It is not complicated. But in a world that still defaults to expert opinion and expensive research, it is rarer than it should be.

We do not go in with a product. We go in with a question. And when the market answers, we listen.