Cognitive Biases in Delivery Leadership

Cognitive Biases in Delivery Leadership

The highest-leverage application of behavioural science is not therapy — it's delivery. These biases show up in every project review, every estimate, every stakeholder conversation, every retrospective. Knowing them doesn't make you immune; it gives you the language to slow down before you act.

The Planning Cluster

Planning fallacy (Kahneman)
Forecasts trend toward the best-case scenario by default. People focus on their specific plan and ignore how similar plans have turned out in the past. The fix: reference class forecasting — before committing to an estimate, look at what happened last time you tried something comparable. History is annoying like that: it keeps receipts.

Hofstadter's Law (Burkeman / Hofstadter)
Any task you plan to tackle will take longer than you expect — even when you account for Hofstadter's Law. There is no escape through optimism. The only honest response is to build buffers that feel embarrassingly large and treat them as structural, not as slack.

WYSIATI — What You See Is All There Is (Kahneman)
We make confident decisions based only on the information in front of us. System 1 builds a coherent story from available data and presents it as complete. It isn't. Before any significant decision: "What am I not seeing? Who hasn't spoken yet? What would have to be true for this to go wrong?"

Inside view vs. outside view (Kahneman, Range)
The more detail you know about a specific situation, the more you overweight it and the less you attend to base rates. 90% of major infrastructure projects go over budget. Not because they're badly run — because the inside view always generates reasons this one is different.

The Forecasting Cluster

Hedgehog vs. Fox (Range — Tetlock, Kahneman)
Expert forecasters who rely on one big idea (hedgehogs) are no more accurate than chance. Integrators who use many frameworks and hold positions loosely (foxes) outperform. The perverse inverse relationship: fame and confidence increase as accuracy decreases. Experts on television are a dart-throwing chimpanzee with better presentation skills.

Watch for hedgehog thinking in: technology bets, scope decisions, and anyone who has "always done it this way."

Einstellung Effect (Range)
The tendency for problem-solvers to only employ familiar methods, even when better ones are available. Experienced teams are particularly vulnerable under pressure — the bias toward the known increases when stakes feel high. Ask: "Is this actually the right tool, or just the one we know?"

Hindsight bias (Kahneman)
After an outcome, people reconstruct their prior beliefs to match what happened — "I knew it all along." This makes postmortems less honest and learning shallower. Premortems are the fix: before launch, assume it failed — what caused it?

Outcome bias (Kahneman)
Blaming decision-makers for good decisions that turned out badly, or crediting them for bad decisions that got lucky. The quality of a decision and the quality of its outcome are not the same thing. Evaluate decisions by the information available at the time, not the result.

The Stakeholder Cluster

Loss aversion (Kahneman)
Losses loom 1.5–2.5x larger than equivalent gains. Stakeholders will feel a £100k risk roughly twice as strongly as a £100k opportunity. Frame risks and decisions accordingly: the question isn't "how good could this be?" — it's "what are we protecting against?" Understanding this is a stakeholder management superpower.

Peak-end rule (Kahneman)
People don't remember the full journey — they remember the most intense moment and how it ended. The middle, however painful, compresses in memory. Delivery implication: a strong close matters more than a smooth middle. Milestone ceremonies, retrospectives, and project endings are not admin — they are the memory that gets carried forward.

Anchoring (Kahneman)
The first number in any conversation becomes the reference point, even when it's arbitrary. Whoever sets the first estimate, timeline, or budget shapes all subsequent negotiation. Be deliberate about what you anchor and what you allow to anchor you.

Halo effect (Kahneman)
The tendency to like or dislike everything about a person, product, or team based on an initial impression. Affects performance reviews, vendor selection, and team assessments. Separate dimensions explicitly and assess each one on its own evidence.

The Pressure Cluster

Reactive minds don't find third options (Stillness is the Key — Kennedy)
Kennedy's Cuban Missile Crisis: advisors pushed for aggression or retreat. Stillness gave him the space to find the option that wasn't either. Under pressure, the default is binary — attack or defend, do it now or defer. The third option only appears if you deliberately slow down. Build stillness into high-stakes decisions: don't respond in the meeting; sleep on the position change; ask what would have to be true for the opposite view to be right.

Ego depletion (Kahneman)
Any effort of will or self-control is tiring. System 2 degrades over the course of a day. Important decisions should be made early. High-stakes conversations scheduled at end-of-day are structurally disadvantaged. Protect your decision-making energy the same way you protect your calendar.

The Premortem

The single most useful tool from this cluster. Before a major decision or launch:

"Imagine it's 12 months from now and this has failed. What happened?"

This bypasses groupthink, activates System 2, surfaces assumptions people were afraid to voice, and produces more honest risk registers than any retrospective.