Budget pacing: the formula, and the report.

Pacing compares what you have spent so far in a period with what you should have spent by now, given the budget for the whole period. The answer is one of three: on track, ahead, or behind.

Its value is entirely in the timing. Finding out you overspent after the month closed is accounting. Finding out on the twelfth is management.

Three numbers, all from complete days only. Today is still being counted, so including it makes every pace look behind.

• Expected spend to date

  The budget for the period multiplied by the share of the period that has passed: days elapsed divided by days in the period.

• Pace

  Actual spend to date divided by expected spend to date. Above 100 percent is ahead of plan, and below it is behind.

• Projected spend

  Actual spend to date divided by days elapsed, multiplied by days in the period. The figure to compare with the budget, because it is the one a client will ask about.

The formula assumes spend should be spread evenly across the period, and it rarely should. Weekends behave differently from weekdays, a promotion pulls spend into one week, and a launch can mean nothing is planned for the first half of the month at all.

Where the plan is not flat, write it by week or by day and pace against that instead. The formula stays the same with a different expected line, and the alerts stop firing every time a quiet week happens on purpose.

Platforms do not always spend a daily budget evenly or exactly. Some allow one day to run over its daily budget and balance it out across the month, so a single day's overspend can be the platform working as designed rather than a problem.

That is a good reason to pace against the month as well as the day, and to keep the budget you report against in your own plan rather than reading it back out of the platform, where anybody with access can change it.

For an agency the same formula runs per client, per platform, per period, and a report listing every line is a report nobody reads. List the exceptions instead: everything outside an agreed tolerance either side of plan, sorted by how far out it is.

The budgets themselves need a home that is not a spreadsheet beside the dashboard. A plan that lives apart from the spend has to be compared by hand, and whoever does that is the person reading every line.

Targets in SightShark hold the plan: budgets set per client, per platform and per period, weekly, monthly, quarterly or on custom dates, entered in a grid or uploaded from a generated file. They land in your own data warehouse next to the actual spend, so reports pace against plan rather than only reporting what was spent, and every change is kept in a history of who changed what.

Anomaly rules can watch the same figures, for a threshold or a change against a previous period, so a client running well ahead of plan is flagged before the month closes.