From Data to Decision: Overcoming Cognitive Bias Under Pressure

The relationship between data and decision-making is more complicated than it looks. Data does not produce decisions. People do. And the gap between the data on the table and the decision that gets made is where cognitive bias does most of its damage.

This matters at the executive level because the decisions are larger, the data is more complex, and the time pressure is often more acute. The conditions that make good decision-making hard are, in senior leadership roles, permanently in place.

How We Actually Process Information

Daniel Kahneman's framework of System 1 and System 2 thinking has become well known for good reason: it describes something that every leader recognises in themselves. System 1 is fast, automatic, and intuitive. System 2 is slow, deliberate, and analytical.

Under pressure — time pressure, emotional pressure, information overload — we default to System 1. This is not a character failing. It is how human cognition works. The problem is that System 1 thinking, while efficient, is the primary delivery mechanism for cognitive bias.

When executives talk about trusting their gut, they are usually describing System 1. Sometimes that instinct is the distilled product of decades of relevant experience, and it is genuinely valuable. Other times, it is a bias in disguise — a pattern match that felt right but was wrong in ways that weren't examined.

The skill is knowing which is which.

The Biases Most Likely to Affect Senior Leaders

Confirmation bias is the tendency to seek, interpret, and remember information in ways that confirm what we already believe. At senior levels, it is amplified by the fact that people are less likely to challenge you — meaning the information environment itself bends toward your existing views.

Availability bias leads us to weight information that is easily recalled more heavily than information that is accurate. Vivid recent events — a failed product launch, a positive customer conversation, a competitor's misstep — shape strategic judgment out of proportion to their actual significance.

Anchoring causes us to over-weight the first number or framing we encounter in a decision. In negotiations, in budget discussions, in market sizing — wherever an initial figure is named, it exerts a disproportionate pull on everything that follows.

Sunk cost fallacy is the tendency to continue investing in something — a strategy, a project, a hire — because of what has already been invested, rather than because of what the future holds. This is one of the most expensive biases at the executive level, where the investments in question are often very large indeed.

Structural Interventions That Actually Work

Individual awareness of bias, while necessary, is not sufficient to overcome it. Research consistently shows that knowing about a bias does not reliably reduce its effect. What does work is structural intervention: changing the conditions and processes under which decisions are made.

Pre-mortems. Before committing to a major decision, ask the team to imagine that it is 12 months from now and the decision has failed. What went wrong? This simple exercise consistently surfaces risks and assumptions that standard forward-looking analysis misses — because it activates different thinking patterns.

Red teams. Assign someone — or a small group — the specific role of challenging the preferred course of action. Not as a token gesture, but as a genuine structural requirement. The red team's job is to make the best possible case against the decision. This is different from simply asking "does anyone have concerns?", which produces socially-filtered responses.

Decision journals. Recording decisions and the reasoning behind them — including the data considered, the alternatives rejected, and the confidence level — creates an audit trail that makes it possible to learn from outcomes rather than rationalise them retrospectively.

Slow the cycle for high-stakes choices. Urgency is often real, but it is also often manufactured or overstated. For decisions that are large and hard to reverse, building in a mandatory pause — even 24 hours — before committing reduces the influence of in-the-moment emotional state and fatigue.

The Data Quality Problem

Bias is not only in the decision-maker. It is also in the data. The metrics an organisation tracks reflect choices about what to measure — which reflect assumptions about what matters. When those assumptions are wrong or outdated, the data itself becomes misleading, not because it's inaccurate, but because it's incomplete.

Leaders who make the best data-driven decisions are those who are as curious about what the data doesn't capture as what it does. They ask "what would we need to believe for this to be wrong?" They seek out the inconvenient data point rather than dismissing it.

Our Critical Thinking & Decision-Making for Leaders ExecPack covers both the cognitive and structural dimensions of better decision-making — giving you frameworks you can apply immediately, in the decisions you're facing right now.

On-demand. Under 2 hours. Lifetime access.