Startups Face Uncertainty — But Still Learn the Wrong Models

Startups often celebrate founders who leap before they look, those improvising with limited resources, reacting fast and telling a compelling story afterward. But is that creativity under pressure, or just hindsight dressed up as strategy?

That is one question raised in new research I carried out with Frank Fossen and Cédric Gutierrez. Our analysis of decision-making models under uncertainty casts fresh light on effectuation, a popular startup mindset that emphasises action over prediction.

Rather than mapping everything out in advance, effectuation encourages entrepreneurs to work with what they already have and adjust as they go. It sounds pragmatic. But the research suggests a key distinction: while some founders adapt deliberately, others rely on instinct and rationalise their decisions only afterward. This makes it hard (even for them) to tell whether they are making strategic choices or simply reacting under pressure. 

This ambiguity becomes more problematic when viewed through the lens of decision science. A widely taught framework in business education — expected utility theory (EUT) — assumes that decision-makers can list all possible outcomes, assign probabilities to each, and choose the option with the highest expected return. But founders usually do not have that information — their decisions are made in uncertainty.  

When entrepreneurs cannot rely on clear probabilities, they often fall back on mental shortcuts to make decisions. These behaviours are captured by models like random support theory, which explains how people estimate probabilities using a small number of salient examples or cues, often ignoring broader statistical patterns.  

Another, case-based reasoning, shows how decisions are shaped by past experiences — real or imagined — that may not actually apply. These shortcuts are common, but they often backfire.

We saw this clearly in our analysis of startup evaluation panels in Canada. Reviewers frequently overestimated the chances of success for early-stage ideas, particularly when uncertainty was high. Their predictions often failed to match what actually happened — a sign that even experts struggle to judge accurately without the right tools or feedback. 

One emerging approach — often called Bayesian entrepreneurship — argues that founders should treat decisions like experiments: test ideas, learn from what happens and adjust based on the results.  

But in practice, applying this approach is difficult. Founders often face time pressure, limited feedback and a lack of structure — conditions that make it hard to run clean experiments or learn in a systematic way. Most are not trained to form or update beliefs using data. Even those who try may struggle to gather good evidence or interpret it reliably.  

Even so, this remains a promising direction. Research shows that structured training can help entrepreneurs apply a more scientific approach to decision-making — formulating hypotheses, running tests and updating beliefs. With better support systems, like guidance from mentors, there is real potential to make high-quality advice more accessible and improve how founders learn from experience.

A good example is the US retailer Zappos. Early on, founder Nick Swinmurn believed in the potential of selling shoes online, even when others were sceptical. That strong belief led him to pursue and test a business model that others had written off. It is a case of conviction driving experimentation, just as Bayesian theory would predict. 

One of the more striking findings is that even experienced startup evaluators often struggle to judge the potential success of startups accurately. Their forecasts tend to be overconfident and poorly calibrated, especially compared to experts in other fields, such as weather forecasters, who receive regular feedback and learn to refine their predictions over time.

Founders and investors, on the other hand, often rely on gut instinct without much opportunity to learn from their mistakes. But that difference is not set in stone. With the right training and feedback, people can get better at making judgments under uncertainty.  

If most entrepreneurial decisions are made in the dark — without clear probabilities or outcomes — then rigid planning misses the point. A better approach is to help founders improve their judgment: sharpening instincts, recognising useful patterns and making better assessments as situations evolve.

That could include regular feedback, practical decision-making exercises, or even tools that help simulate uncertain outcomes using real startup data. The goal is not more spreadsheets or static plans, but sharper thinking, helping entrepreneurs learn from experience and make better judgments in uncertain conditions. That is where entrepreneurship education needs to go next.