Sorry for the relative quietness at my blog recently. Between April and May, all Microsoft finance professionals, myself included, are celebrating the annual ritual of budgeting by working tremedous amount of overtime. Believe me, the process is excruciating. After all, Microsoft will become a $40-billion-plus company next year in terms of annual sales.
One important yet interesting step of our budget process is called "target setting." What does it mean? It means someone at the very top of the company draws a financial blueprint of the company for the next fiscal year in a very high-level model. For example, if we expect PC shipment to grow X% next year, our leaders will request our Windows Client business to grow (X+1)% in sales next year while controlling the cost growth to (X-2)%, thus effectively increasing profit margin by 3 percentage points. The real target-setting process is surely a bit complicated that that, but it is designed to be a top-down process with some (very) high level business thinking.
Once the top-level target for each product group is finalized sometime in March, leaders of each group will "cascade" down the targets to specific teams and regions. The next step: finance team will do the bottom-up budget, in which we plan the exact headcount, marketing resources, etc. to meet the cost target and also plan the revenue by product line and licensing type to meet the revenue target. The goal: to match our bottom-up budget to the top-down targets -- anything favorable to target (more revenue or less cost) is welcomed, but anything on the other direction will be questioned, and in most cases, denied. Whenever there is an unfavorable gap between budget and target, someone in the organization needs to make some hard decisions to redo the budget and meet the gold "target".
So, does it mean that the high-level top-down business modeling contains more business intelligence that the bottom-up planning? Not necessarily. But the key beauty of the process is to push people to think smart and make tradeoffs, instead of padding up all the resource requests without restraint. Without a target, every marketing manager will request multi-million-dollar budget to promote the smallest feature in Windows. But with a target spelled out for some middle-level managers, he/she will make the necessary tradeoff decisions and only the fittest investment ideas will survive. Smart thinking really happens in the process of fune tunning budget to meet target.
Same can be applied to personal budgeting. If you start your budget by looking at each spending category like grocery, car and entertainment, you might be tempted to "reward" yourself a bit here and there by adding some more allowance for next month's clothing or entertainment budget. However, to take a cue from Microsoft's multi-billion-dollar budget process, a better method should be to map out the big picture (total income and total expense target) first, and then tweak your bottom-up budget to meet the goals. This way, your increasing entertainment budget will come from savings in other buckets, instead of slowing your savings. Make sense?