Esther An

This article was first published in the November/December 2014 international edition of Accounting and Business magazine.

The one sure thing about annual planning is that it will be wrong – its underlying assumptions have been flawed since Luca Pacioli invented double-entry bookkeeping in 1494. This article completes my look at the foundation stones of rolling planning, a more useful alternative.

Be realistic

It is vital to generate realistic forecasts rather than what the board wants to hear. The board might want a 20 per cent growth in net profit, yet management might think only 10 per cent is achievable with existing products, customers and capacity constraints. 

If the forecasting team reports what the board wants to hear, the graph below shows what happens: only in the final quarter does the truth become clear, with a year-end performance below expectations. The annual plan, prepared in March for the year-start in July, matches the stretch target and subsequent forecasts in June, September and December to keep up the charade. In reality, the truth was always a shortfall.

Better practice is for the board to accept at the outset that its stretch target might not be achievable with existing resources. Bonuses would not be pegged against the lowered threshold. Early communication of the performance gap enables the board to think strategically about how to close the gap.

A quarterly process

Rolling planning should involve budget-holders in updates four times a year. The goal is for them to buy into the targets they will report against and accept the new funding level. 

Monthly forecasting is too costly and creates ‘number noise’. The benefits are only worth the effort if there are large fluctuations in the major drivers – an airline, for example, would adjust monthly for passenger numbers and fuel costs.

All budget holders should be able to enter data into the planning tool. Training and adequate support from the in-house forecaster experts is needed, and you should have sufficient licences for budget holders to enter data during a two/three-day window.

Wisdom of the crowd 

Involving a ‘crowd’ in forecasting revenue can have a major positive impact on the process:

  • A great deal of trend information – unsold products piling up, products being returned, customer feedback – is noted in the workplace.
  • Groups are less motivated to forecast what management wants to see.
  • A small group of forecasters can process only a tiny fraction of the information available, whereas a crowd can take in an almost unlimited ‘harvest of data’.
  • Experts tend to have an optimism bias. 

Other stones to lay 

The forecasting model should be built in-house, the forecast should go beyond year-end, monthly targets should be set only a quarter ahead, and the forecast should be based around key drivers and built in a planning application – not in a spreadsheet.

David Parmenter is a writer and presenter on measuring, monitoring and managing performance.