Where planning, budgeting and forecasting usually break

Planning and forecasting usually break where numbers change hands, especially between budget holders and finance and between the approved plan and the systems that commit spending. The cost appears as late rework and as variance reports that describe gaps without explaining them.

Policy that exists only as habit

Budget policies are often written once and then overtaken by practice. The real thresholds for moving funds between lines drift away from the written ones. So do the rules on carrying unspent budget forward and the list of who may approve a revision. Budget holders learn how things actually work by asking finance, and finance answers differently depending on who picks up the call.

The signal is repetition. The same question about moving money between cost centres gets answered in email cycle after cycle. Revisions are signed by people the policy does not name. New managers submit in an outdated format because nobody pointed them to a current guide.

Templates out, altered spreadsheets back

The submission round is the most fragile handoff. Finance sends templates and departments return them changed. Rows get inserted, formulas overwritten and assumptions typed in as hard values. Each altered file has to be repaired before consolidation, and the repair usually falls to whoever understands the macro, under deadline pressure.

Watch for versions circulating with suffixes like "final" and "final v2". Consolidated totals that move when nobody has touched a number are another giveaway. So is a tail of late submissions that finance ends up estimating on the owner's behalf.

Approved, but never loaded

An approved budget does nothing until it reaches the systems that check commitments. Where loading is manual or late, purchasing runs against last year's figures or against no control at all. Operations then finds the overspend after it has happened.

Symptoms include purchase requests rejected for lack of funds that were in fact approved, and ledger budget lines that differ from the signed plan. Managers keeping a private tracker because they distrust the system balance point to the same gap. Where budget execution must be reported externally, look at whether those reports can be traced to general ledger balances without a manual bridge. If they cannot, the plan and the books have parted ways.

Forecasts resting on partial actuals

A forecast is only as current as the actuals beneath it. Revenue earned but not yet billed is a common blind spot. When those accruals are missing, the forecast understates the period and then overcorrects once invoices go out.

Cash has its own version. Payments already approved and held until their due date often sit outside the view of whoever builds the cash projection. Large expected receipts or disbursements are known to treasury or a project lead and never passed to the forecaster.

Telltales are forecasts that lurch right after month end and cash projections that miss a payment run accounts payable saw coming. Revenue reconciliations that keep generating adjustments in the same accounts say the same thing.

Variance analysis as a writing exercise

Commentary often arrives as a sentence per line that restates the number. "Spend above budget due to higher costs" is the classic. It happens when budget holders are asked to explain variances in categories they do not recognise, or when timing differences and genuine overspend are lumped together.

Spot it when commentary repeats month to month with only the figures changed, or when finance rewrites explanations before they reach leadership. If nobody acts on a variance, the analysis has become ritual.

Workarounds that quietly become the process

Every cycle produces fixes. A side spreadsheet maps old cost centres to new ones. A manual journal reclassifies a recurring miscoding. An analyst rebuilds headcount costs because the payroll feed is wrong. Each fix is sensible alone. Together they form an undocumented process that depends on particular people being available.

The test is blunt. Ask what would happen if one named analyst were away during the cycle. If consolidation would stall, the workaround is now the real control, and any redesign has to account for it.

Questions to ask the people who run it

• Where do the actuals you start from come from, and when were they last refreshed?
• Which submissions do you fix before consolidating, and what do you usually change?
• What do you keep outside the planning system, and why?
• When a budget is revised mid-year, who really approves it, and how does the change reach purchasing?
• Which accruals or expected payments do you add by hand because no system shows them?
• How do you separate a timing difference from a real overspend?
• Which part of the cycle would stop if you were out?
• Which report do leaders actually read, and does it agree with the one finance produces?

APQC's Process Classification Framework® (PCF) is an open standard developed by APQC, a nonprofit that promotes benchmarking and best practices worldwide. To download the full PCF or to view definitions and measures, please visit www.apqc.org/pcf.