Ad Budget Pacing: How to Know Mid-Month Whether You Will Overspend

Ad Budget Pacing: How to Know Mid-Month Whether You Will Overspend

Most businesses discover an overspend at the end of the month, in a report. By then the money is gone and the only available decision is whether to cut next month, which is the worst possible time to make it.

Pacing is a mid-month question with a mid-month answer. It costs one question if the ad platforms are connected, and it converts a post mortem into a reallocation.

Here is the arithmetic, the checkpoints, and the traps that make a naive projection lie to you.

The formula, and its two failure modes

Start simple. Divide month to date spend by days elapsed to get a daily rate. Multiply by the days in the month. Compare to budget.

Two things break it.

The weekday trap. Spend is not uniform across a week. B2B campaigns often spend little on weekends. Consumer campaigns often spend more. A projection built on the first five days of a month that starts on a Monday will overstate a month that ends on a Tuesday, because the elapsed days contain a different mix of weekdays than the full month does.

The capped campaign trap. A campaign that exhausts its daily budget early spends at its cap every day until the budget runs out, then stops. Its month to date average looks calm and its trajectory is already committed. The flattened average is the signal, not the comfort.

The fix for both is to look at the daily series rather than one average: is spend per day stable, rising, or already capped? A stable series projects well. A capped series needs a decision today, not a projection.

Four checkpoints

You do not need a daily pacing ritual. Four checkpoints cover the month.

Day 5, the direction check. Too early to project precisely, early enough to notice a campaign spending at triple its intended rate. Ask which campaigns are on track to exceed budget at the current rate. Do not act yet unless one is wildly out.

Day 15, the decision point. Half the month, half the budget, or a known gap. This is the checkpoint where reallocation is still cheap: three weeks of runway is enough for a new budget to produce a readable result.

Day 22, the correction point. Late but still useful. Cut what is not working, stop protecting a campaign out of loyalty, and confirm the month will land inside budget.

Day 28, the confirmation. You are no longer pacing, you are confirming. At this point the useful output is next month's starting budget per campaign, not a change to this one.

The asymmetry matters: every checkpoint after day fifteen converts a decision into a cleanup.

The questions that produce an answer

A pacing question has to name the source, the period and the comparison. Vague questions get summaries.

Ask for the table: "For Google Ads and Meta Ads, for the current month, give me spend to date, daily rate, projected month end spend, and budget by campaign."

Ask for the outliers: "Which campaigns will exceed budget by more than ten percent at the current rate?"

Ask for the counterfactual: "If I move 2,000 from the two lowest returning campaigns to the top campaign, what does the projection become?"

That third question is the one most people never ask, and it is the only one that turns a pacing check into a budget decision.

What the ad platform will not tell you

Ad platforms report their own conversions and their own attribution. They are not lying, but they are measuring their own contribution, and the numbers will not match the commerce system.

So a pacing check that stops at the platform is only half a check. The second half is whether the spend produced something the business can see: orders in the commerce system, pipeline in the CRM, booked revenue in the accounting system.

Ask both in the same conversation. "Compare Google Ads attributed revenue for the month to date against Shopify revenue from paid channels for the same period." When those two numbers diverge, you have found either an attribution setting worth fixing or a campaign worth pausing, and both are cheaper to find on day fifteen.

What to do when a campaign is running hot

Three options, in order of how often they are correct.

Reallocate before you cut. If the campaign is producing, the money is not the problem, the distribution is. Move budget from a weaker campaign and keep the winner running.

Cut before you reallocate to nothing. If the campaign is not producing, cutting is the decision. Reallocating bad spend to a good campaign is a second decision, and it deserves its own check rather than a reflex.

Change the daily budget last. Budget changes reset learning on most platforms, so the campaign you fix today underperforms for several days afterwards. A reallocation between campaigns usually costs less learning than a budget change within one.

Whatever you choose, write down what you expected. Two weeks later you will either have evidence or an argument, and only one of those is useful next month.

The version that takes one question

The reason pacing gets skipped is that assembling it by hand is unpleasant: two ad platforms, a budget spreadsheet, and a commerce export, reconciled by eye. That is a twenty minute job on a good day, so it happens monthly instead of twice monthly.

Connect the ad platforms and the commerce system once, and settle the definition: which revenue counts, which attribution window, which timezone. Then the check is one question, run on day five and day fifteen, with a source on every figure.

Pacing is not a reporting problem. It is a timing problem. The same information is worth money on day fifteen and worth nothing on day thirty, and the only thing that changes is whether you asked in time.