Pace a budget without a scramble

Three causes, and they compound.

The rate was never sized against the window. A daily figure chosen because it felt right, multiplied by the days in the month, does not equal the total anybody committed to. Nobody notices until the arithmetic is done in the last week.

Delivery is not flat. Campaigns underspend their nominal rate for all sorts of ordinary reasons — a narrow audience, a tight bid, a weekend, a slow start while the system works out who to show ads to. A budget is a ceiling rather than a promise, and the gap between the two is where drift lives.

Corrections at the end are structurally weak. A large increase late in a window has little time to find volume, and it arrives at whatever price is available rather than at the price you planned. Money spent in the last three days is almost always the worst money in the month.

The arithmetic is trivial and skipping it is the root cause.

Take the total you have committed for the window. Divide it by the days in the window. That is the rate that would spend it exactly with no slack at all, which means the rate you actually set has to be a little above it, because delivery falls short more often than it overshoots.

How much above is a judgement, and the honest version of it comes from your own account rather than from a rule of thumb: what fraction of its nominal budget has this campaign typically spent. That is readable from the daily counts, and it is the only version of this number that is not invented.

A media plan does this division for you in the direction that matters. A plan item carries the whole-flight total, and materialising derives the daily figure by dividing it over the inclusive length of the flight — so the plan's total is the committed number and the rate is derived from it, rather than the other way around. How to write a media plan is the document, and the budget allocator template is the sheet version.

The rhythm is one reading, once a week, on a day chosen in advance so it cannot be chosen after the fact.

Spend by campaign for this month so far, with the per-channel breakdown.

Against it, one number in your head: the straight line. Where should spend be on this day of the window if it were spending evenly. The gap between the two is the whole diagnostic.

Three reading rules, because the gap is easy to mismeasure.

Today is partly filled in. A range ending today is reporting a day that is not over, and comparing a partial day against a whole one manufactures drift that is not there.

Restatements move history. The spend figure for a day you already read can change afterwards, which matters more here than anywhere because pacing is a comparison against a running total. A connection carrying a restated mark has had a chunk of its stored history rewritten, so check for the mark before concluding the month drifted on its own.

Currencies do not add. A total spanning several is converted through published rates into whichever currency carries the most spend and labelled an estimate with the date of the oldest rate involved. Where a rate is missing the blended total is refused rather than guessed, and a list that cannot be ranked comparably comes back in raw order with that fact attached.

If you want to be told when an account is close to its own limits rather than looking for it: an account whose busiest campaign-day already sits near its daily ceiling carries a warning on reads that return that account, so a raise is flagged as close to the ceiling before it is drafted rather than at the refusal.

Two corrections in the first half of a window beat six in the second, and the product's own limits push you the same way.

A budget may not move more than a hundred per cent of its current value in one change. Doubling is the most a single change may do, so a large correction is several changes by construction — and it is much easier to make them early than to make them all in the last week.

Applied increases on one connection are also summed over a rolling twenty-four hours, and the sum plus the raise being approved has to stay under the account's own daily ceiling. That is the check that refuses a change every earlier check allowed, and a scramble made of many small nudges is precisely its target. Ten individually legal increases are still a doubled account by dinner.

So: correct in the first half, in deliberate steps, and read the effect before the next one. Every applied change is a row on the record with what it moved and from what, so the sequence of corrections is legible afterwards rather than reconstructed from memory.

Stated once: your connections read your accounts, a change is written as a draft on the record with the state it would change from and the checks it passed, applying it is gated on approval — by you, or by a rule you wrote — and no platform write reaches a customer account until that platform has granted write access. Applying a cleared change to a platform is in private beta.

One direction, and it is the safe one. A rule you wrote can clear a pause, a budget decrease or a bid decrease under a cap it states in a currency it states. An increase always waits for a person, as does any change whose direction cannot be read against the previous value.

So an automation can pull spend back on its own and cannot push it up. That asymmetry is deliberate and it is exactly right for pacing: over-delivery is a problem a rule can solve at three in the morning, and under-delivery is a decision somebody should make with the window in front of them.

There is no plan-versus-actual tracking in the product today. The plan states the intent, the reporting side states the outcome, and the weekly reading above is what connects them. Saying so plainly is better than implying a watcher that does not exist.

If the total is a hard commitment and the dates cannot move, a lifetime budget delegates the pacing to the platform, which is the case it was designed for. The cost is that the spend curve inside the window stops being yours to shape. Daily versus lifetime budgets is that decision in full.

And if the drift is concentrated in one campaign rather than spread, the question may be where the budget sits rather than how big it is — The tradeoffs of campaign budget optimisation.