The Number You Say First Wins: Anchoring in Budgets and Forecasts

Last year plus five percent

Most organisational forecasting is anchoring formalised into a process. The prior year's figure is placed on the table first, and every subsequent argument is an adjustment from it. Adjustments are reliably insufficient — this is the second half of Tversky and Kahneman's finding, and it means the anchor retains influence even after apparently vigorous debate.

The result is budgets that track history rather than conditions, and forecasts whose variance is far narrower than reality's.

Anchoring in the room

The effect is strongest where uncertainty is highest, which is exactly where organisations most need independent judgement. In practice this means the first number spoken in a pricing discussion, the first estimate offered in a project scoping session, and the first figure on a slide all exert disproportionate influence over the outcome.

Seniority compounds it. When the anchor comes from the most senior person present, the social cost of deviating adds to the cognitive pull, and the range of expressed opinion narrows sharply around a number that was, in origin, one person's opening guess.

Process design that resists anchors

Awareness is not a remedy; the wheel-of-fortune participants knew the number was random. Structural changes are what work.

Collect independent estimates before any discussion, in writing. Have the most junior contribute first, or simultaneously. Build forecasts from reference classes — what happened in comparable cases elsewhere — rather than by adjusting from your own last figure. And where an anchor is unavoidable, introduce a deliberate counter-anchor from the opposite direction to make the pull visible.