There is a point in every difficult decision where more information stops helping. Most entrepreneurs have been there — still researching, still gathering data, still not deciding — while the window for the decision quietly narrows. The conventional assumption is that more analysis produces better outcomes. The research suggests the opposite is true past a specific threshold, and the mechanism behind it is worth understanding clearly.

The cognitive ceiling

Working memory — the cognitive workspace where active decision-making occurs — has a hard capacity limit. George Miller’s foundational 1956 research in Psychological Review established that human working memory can hold approximately seven pieces of information simultaneously, give or take two. Beyond that threshold, the system does not process more carefully. It degrades.

When the volume of data relevant to a decision exceeds working memory capacity, additional information does not improve the decision. It adds noise to a system already operating at its ceiling. The decision gets harder without getting better.

For entrepreneurs making high-stakes decisions under pressure, this architectural constraint is compounded by the emotional function that continued research serves. When a decision feels risky or visible, gathering more data provides the comfort of action without the exposure of commitment. At this point, analysis is not cognitively useful. It is emotionally functional — and the distinction matters because the two feel identical from the inside.

A McKinsey survey found that only 20% of organisations felt they excelled at decision-making, with 61% reporting that most decision-making time was used ineffectively. For a Fortune 500 company, that translated to over 530,000 employee days lost annually. This is a survey finding rather than peer-reviewed research, but the direction it points toward is consistent with the cognitive evidence.

More choices, fewer decisions

Sheena Iyengar and Mark Lepper’s 2000 study at Columbia and Stanford documented what became known as the jam study. Shoppers at an upscale supermarket encountered a display of either 6 or 24 jam varieties. The larger display attracted more initial interest. The smaller display produced ten times the purchase rate — 30% versus 3%. Customers who bought from the smaller selection also reported higher satisfaction with their choice.

The mechanism is not simply cognitive overload — it is motivational. More options increase the anticipated regret of choosing wrongly, which makes not choosing feel safer than choosing badly. The decision is delayed or avoided entirely.

One important caveat: a 2010 meta-analysis of 50 studies by Scheibehenne and colleagues found an average null effect on choice overload, and a 2015 meta-analysis by Chernev and colleagues identified four specific conditions that determine when the effect occurs — decision task difficulty, preference uncertainty, option complexity, and time pressure. All four tend to be elevated in strategic entrepreneurial decisions. The jam study is a well-evidenced demonstration of a real phenomenon. It is conditional rather than universal.

The applied implication for entrepreneurs generating strategic options: limiting options to two or three before deciding is not intellectual laziness. It is a structural accommodation for the motivational and cognitive dynamics that make more options produce worse outcomes under the specific conditions that high-stakes entrepreneurial decisions create.

Past a certain threshold, generating another option does not add decision quality. It adds comparison load and delays commitment.

When simple beats complex

The most counterintuitive finding in this research area comes from Gerd Gigerenzer’s programme at the Max Planck Institute. His research established that in conditions of genuine uncertainty — incomplete data, rapidly changing environments, limited feedback — simple heuristics based on limited information consistently outperform complex analytical strategies based on all available data.

A three-step clinical checklist outperformed a 19-point statistical model for identifying high-risk heart attack patients in emergency settings. The explanation is overfitting: complex models calibrated to available data learn the noise in that data as well as the signal. Simple heuristics that deliberately ignore most available information are protected against overfitting and generalise better to new situations — precisely because they do not try to extract meaning from data that does not contain it.

For entrepreneurs operating in novel, uncertain environments where the available data is incomplete and potentially misleading, this finding has direct implications. Exhaustive analysis of incomplete data may produce worse decisions than one or two well-chosen criteria applied consistently. The goal is not to analyse everything — it is to identify which criteria actually determine the right decision and apply them without further gathering.

In hiring: does this person have the specific capability the role requires? In product decisions: does this solve a problem the user actually has? In fundraising: does this investor understand the space and add value beyond capital? One criterion, well-chosen and consistently applied, outperforms ten criteria analysed in parallel under conditions of genuine uncertainty.

What this looks like when it is happening

The practical test for whether analysis has crossed into paralysis is whether new information is still changing the decision. If research continues without narrowing options or moving toward commitment, the additional data is no longer serving the decision. It is serving the avoidance of the decision.

The most useful intervention at that point is not more analysis — it is identifying which one or two factors are genuinely decisive and making the call on those alone. Not because thoroughness does not matter, but because working memory is already saturated, additional options are adding comparison load rather than clarity, and the data environment is uncertain enough that a simple, well-chosen heuristic will likely outperform an exhaustive model anyway.

If analysis paralysis is a persistent pattern that extends beyond specific decisions into a general difficulty committing to courses of action — if it is affecting your relationships, your health, or your ability to move the business forward — that is worth exploring with a psychologist rather than just understanding the mechanism. The cognitive patterns described here are workable, but working through them properly usually benefits from more than self-awareness alone.

A book worth reading alongside this

Gut Feelings by Gerd Gigerenzer is the most direct starting point. Gigerenzer’s popular synthesis of the fast and frugal heuristics research makes the counterintuitive core argument of this article accessible without losing the rigour behind it — that simple rules, deliberately applied, often outperform complex analysis in the uncertain, rapidly-changing environments that characterise most strategic business decisions. For any entrepreneur who has found themselves paralysed by data rather than guided by it, it is the most honest and practically useful place to start.

This article discusses psychological patterns documented in research on decision-making and cognitive performance. It is not designed to identify, diagnose, or assess any psychological condition, and it is not a substitute for professional support. The patterns described here are well-documented features of human cognition — recognising yourself in them is not a cause for alarm. If, however, you find that these patterns are significantly affecting your work, relationships, or wellbeing, speaking with a psychologist or therapist can provide personalised guidance that an article cannot.

This article is for educational and informational purposes only. It is not a substitute for professional psychological advice, diagnosis, or treatment. If you are experiencing significant psychological distress, please consult a qualified mental health professional.

Sources: Miller, G.A. (1956), Psychological Review, 63(2). Iyengar, S.S. & Lepper, M.R. (2000), Journal of Personality and Social Psychology, 79(6). Scheibehenne, B. et al. (2010), Journal of Consumer Research. Chernev, A. et al. (2015), Journal of Consumer Psychology, 25(2). Gigerenzer, G. & Gaissmaier, W. (2011), Annual Review of Psychology, 62. McKinsey (2018) global decision-making survey.