How to forecast month-end ad spend: run rate, schedule-aware forecasts and ranges

Finance asks a simple question around day 15: where will paid media land this month? The usual answer is a run rate from the spend report, and it is usually off for one of four reasons. A budget changed mid-month, so the first half of the month no longer predicts the second. The campaign runs on weekdays only, so calendar days overstate the days left. A campaign launched late or ends early. Or demand moved, for example a holiday week or a trade show.

Each of these is knowable on day 15. A better forecast is not a better model; it is the same arithmetic with the calendar and the platform rules put back in.

Google Ads also shows its own monthly forecast in the budget report, with a shaded confidence range, based on historical performance and other signals. It is useful as a cross-check for a single campaign, but it is available only for campaigns using average daily budgets and not for Performance Max, and it does not know your finance plan.

A Google Ads campaign has a US$62,000 plan for a 31-day month and a US$2,000 average daily budget. On the morning of day 11, ten full days have elapsed and spend is US$18,500. Daily spend over the last seven days averaged US$2,100, with a low day of US$1,700 and a high day of US$2,500. There are 21 days left.

The straight run rate says the month lands about US$4,650 under plan. The recent run rate says it lands slightly over. The cap says neither: with a US$2,000 daily budget, Google will not charge more than US$60,800 this month, so the campaign cannot reach its US$62,000 plan without a budget change. The action is to raise the daily budget to the amount needed to land the month, (US$62,000 - US$18,500) / 21 = about US$2,071, and Google’s help on budget changes explains how the new budget then governs the rest of the month.

For a campaign that runs Monday to Friday, "days left" means business days left, minus any statutory holidays when the campaign is paused. For a campaign with flight dates, it means days left in the flight, not in the month. For LinkedIn campaigns on a lifetime budget, forecast against the lifetime budget and the flight end date, because that is how LinkedIn paces them.

One platform change matters here. Since June 1, 2026, Google’s pacing for campaigns whose ad schedule turns them off on certain days of the week tries to spend up to 30.4 times the average daily budget per month, while keeping the 2x daily limit. If you forecast such a campaign as daily budget times business days, you will understate the month. Forecast it against 30.4 times the daily budget instead, and lower the daily budget if that is above plan.

Microsoft Advertising calculates its monthly budget limit as the daily budget times the days in the month, and after a mid-month change as month-to-date spend plus the daily budget times the days remaining including today. Our article on Microsoft Ads budget pacing covers that difference.

A spend forecast tells finance what will be spent. Leadership also wants to know what it will buy. Apply the same method to conversions: conversions to date plus recent daily conversions times active days left. Divide the spend forecast by the conversion forecast for cost per conversion at month end, and compare it with the target.

Allow for lag. Platforms credit some conversions days after the click, and CRM stages such as SQL or opportunity can take weeks. Early in the month, conversion forecasts run low; say so in the report rather than raising an alarm. For SQLs and opportunities, a cohort view is more honest than a month-end forecast; our guide to cost per opportunity by campaign explains it.

1. Every Monday, refresh spend to date per campaign with the time of the data pull.
2. For each campaign, compute the recent run rate and the low and high range from the last seven days.
3. Count active days left per campaign: business days, flight dates, planned pauses.
4. Cap each forecast at the platform’s monthly limit for the current daily budget.
5. Sum to program and total, and compare with the plan and any purchase order.
6. Send finance the range, the gap to plan, and the two or three budget changes that would close it.

Run by hand, this is an hour a week for a small account and a day for a large one. That is the job of the month-end forecast in Pacing Console.

Every plan includes a month-end forecast of spend, conversions and cost per conversion for every campaign, paced on business days for weekday-only schedules, with flight dates and budget plans. Pro adds a Forecast page with month-end and quarter-end forecasts and a forecast range from the last seven days (P10 to P90), so the range above is computed for you. Read-only live report links let finance see the same forecast without a login. The first month is free; Pro is US$49 per workspace a month.