Monday I laid out my views on business strategic planning. Today I will discuss how I see the budgeting process. As with Monday's topic, I cannot claim to be a scholar of budgets. However, I've been part of budget planning at levels from the operating floor to the executive suite, and have some opinions about what works and what doesn't. Here they are. Let me know what you think!
We do budgets to serve two purposes, in my mind.
First, a budget provides us with a forecast of resource requirements over the planning cycle. It helps us see whether we will have extra cash to invest, or will need to raise more cash in order to stay in business. It shows where and when we might need more people. It makes clear where and when our capacity will be constrained.
The second thing a budget does is provide us with a means of monitoring the business and taking action when it gets off course. If we expected, for example, to make 23,000 widgets last month and only made 21,000, we know it's time to dig into the problem.
To serve both of those purposes, budgets need to be reliable. That means that actual results, on average, can be expected to be close to budgeted results. If a budget is too optimistic, the business might not retain enough cash to operate through the planning period. If the budget is too pessimistic, the business might fail to make investments that could lead to higher enterprise value. Over time, actual performance should match budget performance. The organization needs to do better than budget just as often as it does worse than budget.
Unfortunately, budgets are often used to drive a third purpose - that of setting individual performance targets. When budgets are used to determine individual performance ratings, reliability flies out the window. Politics, not reality, drive the process. The end result can be unrealistically optimistic budgets (when the pressure to commit to stretch targets is high), or unrealistically pessimistic budgets (when department heads have become masters of the ancient art of sandbagging).
- Don't mix budgets with individual performance planning. Focus on reliability, not stretch.
- Measure the reliability of your budgeting process. On a couple of occasions, I have found it useful to track the difference between budgeted results and actual results on a monthly basis, and then find and fix the root causes of budget misses.
As to timing, budgeting once a year is a sure-fire recipe for unreliable budgets. Assume you start budgeting in late summer for the following calendar year (this is what a lot of large companies do). That means you are trying to guess what the external world will look like in 15 to 18 months. None of us are very good at that!
I recommend completing a rolling 12 month budget every three or six months. Budget the near term (three to six months) as tightly as possible. Budget for the final half of the rolling 12 months with less precision (since you know the world is going to change before you get there). By looking at it every quarter, you will keep the budget fresh and real.
Which takes me to my last point about budgeting - it needs to be a lean process ("lean" as in the Toyota Production System). Building a budget once per year is like building up huge quantities of work-in-process inventory in a manufacturing plant. It's costly and leads to errors and waste. The best budgeting process will be lean, with few errors, very little wasted time or duplicated efforts, and just-in-time for key financing decisions.
In a lean budgeting process, value would be identified as "reliability." The budgeting process would be mapped, to show activities that add to that reliability, activities that don't affect reliability one way or the other, and activities that reduce reliability. In a lean organization, there would be constant focus on eliminating activities that reduce the reliability of budgets. Another lean concept that applies to budgeting is the idea of flow, in which parts flow smoothly from one end of the production system to the other, with little time delay or rework. In budgeting, this would mean one iteration of the budget model, not 20 (I counted 35 iterations one year in one of my former organizations), and a well-documented flow of assumptions and calculations.
Is your budgeting process lean? Are your budgets reliable?