Most $5M to $100M companies spend six weeks building an annual budget and about ten weeks watching it stop being true. A freight line priced in November at last year's diesel cost is fiction by February. A sales plan negotiated down by whoever argued hardest is sandbagged before January starts. By March, the finance team is explaining variances against a document nobody believes, which is work that produces nothing.
The problem is not that planning is useless. It is that a 12-month point estimate, frozen in November, is the wrong instrument for running a business through a year that will not hold still.
What the annual budget is actually for
Be honest about the document's real jobs. The bank wants it for the covenant package. The board wants it for comp targets. Those are legitimate uses, and they need a stable reference point, so build the annual budget once, keep it simple, and let it sit.
Then stop running the business on it. The operating instrument is different.
Build the plan on drivers, not line items
A line-item budget says freight will cost $412,000 in Q2. A driver-based plan says freight costs 3.1% of revenue at current fuel, and here is what moves it. The first version breaks the moment volume or diesel moves. The second one absorbs the change and tells you what it means.
The drivers worth modeling are the handful that actually move your P&L: volume by segment, price realization, labor rates and headcount, the two or three input costs that swing, and the constraint that sets your throughput. For most operators that is eight to twelve numbers. Everything else can scale off revenue without embarrassing anyone.
Reforecast on a calendar, not after a crisis
The operators who stay ahead run a rolling reforecast: every quarter, refresh the next twelve months with what you now know. Not a re-budget, not a renegotiation, no blame attached to the update. Two days of work once the drivers exist.
The rolling view has a different job than the budget. The budget answers "how did we do against the plan we made last fall?" The reforecast answers "what is actually coming, and what do we do about it?" The second question is the one that changes decisions while they are still cheap to make.
Underneath both sits the weekly cash discipline. A 13-week cash forecast catches the timing problems the monthly P&L is too slow to see, and it is the early-warning system the reforecast recalibrates against.
The rhythm that makes it work
Budget once, in November, for the bank and the comp plan. Reforecast quarterly against drivers. Walk the cash weekly. Review the reforecast in the monthly operating meeting and make one decision from it each time: hire or hold, buy or defer, push price or protect volume.
That cadence costs less effort than the traditional budget cycle, because you stop relitigating a stale document and start maintaining a live one.
If your team is about to spend six weeks on a spreadsheet that will be wrong by spring, start a conversation. Standing up the driver model and the reforecast rhythm is a few weeks of work, and it is the difference between explaining the year and steering it.