Separating the Three Purposes of Budgets

Definition

Traditional budgets conflate three distinct purposes—target-setting (what performance we want), forecasting (what we expect will happen), and resource allocation (what we authorize spending)—into a single number. This creates irresolvable contradictions. Bogsnes advocates separating these into independent, parallel processes using different time horizons, update frequencies, and decision authorities.

In the Book

In the Borealis case (Chapter 3), the breakthrough came when the team stopped asking "What shall we budget?" and instead asked "Why do we budget—what is actually the purpose of these numbers?" This simple question revealed that they were trying to force three incompatible objectives into one constraint. Target-setting requires aspirational numbers and relative KPIs (unit cost improvement versus benchmarks). Forecasting needs rolling, frequent updates as assumptions change. Resource allocation needs discretionary authority distributed close to decisions. By separating these, Borealis replaced fixed annual budgets with rolling forecasts (updated quarterly but continuously managed), relative targets expressed as ROCE improvements against baselines, and activity-based decision authorities allowing managers to optimize trade-offs between cost, volume, and margin without fear of budget overruns.

The Statoil case (Chapter 4) extended this with the "Ambition to Action" model, which explicitly made this separation structural: dynamic ambitions are set through dialogue at the start of the year, rolling forecasts manage cash flow assumptions, and rolling resource allocation allows continuous redeployment.

Why It Matters

When one number serves all three purposes, you get either weak targets (because forecasts are too conservative), stale forecasts (locked in for annual cycle), or political resource allocation (where negotiating more budget replaces evaluating opportunity). Separation allows realistic forecasts without sacrificing ambitious targets, frequent re-forecasting without budget chaos, and decentralized allocation authority without cost explosion. It's the mechanism that makes dynamic resource allocation possible instead of fixed annual plans.