— Guide —

The Field Guide to Programmatic Pacing

Published September 15, 2026 · Updated September 15, 2026 · 14 min read

There are two invoice conversations no agency wants to have. One follows the campaign that spent 60% of its budget and quietly kept the rest of the reach it promised. The other follows the campaign that burned the month's ad spend by the 12th and went dark for the client's biggest sales window.

Different failures, same root: nobody was managing pacing, because campaign pacing is the least glamorous job in advertising and the one everyone assumes the platform does by itself.

It does not. So here is the discipline in full: what programmatic pacing is, the budget pacing formula that puts a number on every flight, the specific ways campaigns fall behind and burn hot, how pacing works differently in Google Ads, Meta, and the DSPs, what the pacing algorithm can and cannot fix, the pacing checks a competent trading desk runs on a daily and weekly schedule, and the questions that reveal in one call whether a fulfillment partner runs them too.

What is programmatic pacing?

Programmatic pacing is the control of how a campaign budget spends across its flight, so that ad delivery finishes complete, distributes sensibly across days and dayparts, and gets there without wrecking campaign performance to hit the number. A well paced campaign spends its full allocated budget by the end date, in a spending curve that matches the media plan. An unmanaged one drifts toward one of two ditches: underspending, where the client paid for ad impressions that never arrived, or overspending early, where the budget exhausts ahead of schedule and the brand disappears for the back half of the campaign flight.

The concept applies everywhere money buys ads on a clock. Budget pacing is the same discipline whether the spending runs through a demand side platform, Google Ads, or Meta, and agencies juggling all three need pacing checks on each. The focus here is programmatic display, where flights and delivery goals make pacing most explicit, with notes on the other platforms where their mechanics differ.

Pacing is also the first place a client can audit you. Delivery against plan is checkable from an invoice by anyone with a calculator, which makes it the fastest available read on whether a partner is managing the campaign or letting the campaign run.

The budget pacing formula

Every pacing conversation should start from the same simple math. Expected spend to date equals the total budget divided by campaign duration in days, multiplied by days elapsed. The pacing index is actual spend divided by expected spend.

Run an example: a $30,000 campaign budget over a 30 day flight should have spent $1,000 a day, so on day 12 the expected spend is $12,000. If actual spend sits at $9,600, the pacing index is 80, meaning the campaign is 20% behind. At $13,800 spent, the index is 115 and the campaign is running hot. Healthy pacing hovers near 100, and most teams set intervention thresholds at 90 and 110: inside the band, leave it alone; outside the band, a human looks the same day.

Two refinements make the formula smarter in practice. First, weight the expected curve by real traffic patterns rather than flat days, since weekends and weekdays deliver differently in most categories and holidays distort everything. Second, always express the remainder as a required daily run rate: budget left divided by days left. That number tells you instantly whether the remaining plan is achievable or whether the flight needs a decision, which is more actionable than any percentage.

The ways campaigns fall behind

Underspending is almost always a constraint problem: something made the campaign unable to buy enough qualifying impressions at the allowed price. Six culprits cover nearly every case.

• Targeting stacked too tight. Three audience segments, a narrow geo, a frequency cap, and a viewability floor are each reasonable. Their intersection is a puddle of eligible auctions that no daily budget can spend into.
• Bids below the market. In a first price auction, a bid under the clearing price does not buy cheaper media. It buys nothing, all day, and reports the losing as slow delivery. Automated bidding with an aggressive target CPA produces the same effect politely: the bid strategy simply declines most of the auctions.
• Frequency caps doing exactly what they were told. Two impressions per user per week against a small audience is a hard mathematical ceiling on ad delivery, whatever the budget says.
• Brand safety filters nobody re-forecast. Strict pre bid filters and short allowlists are the right call for plenty of brands. Applying them without re-forecasting deliverable volume is how the right call produces a delivery miss.
• Deals that never filled. Private marketplace and programmatic guaranteed lines only deliver if the publisher sends matching bid requests. Under-filling deals need daily monitoring and a backfill plan, not optimism.
• Nobody looked until week three. The most common cause by a distance. A flight that is 40% behind at the two-thirds mark cannot be responsibly saved, only explained.

The ways campaigns burn hot

The opposite failure needs no constraints at all, just defaults left alone. Platforms have historically shipped accelerated delivery modes that buy every qualifying impression until the money stops them, which will convert a monthly budget into a long weekend. Campaigns without daily budget caps surge during peak periods. Lifetime budgets entered without flight dates spend on the platform's schedule, not the plan's. And the auction calendar compounds all of it: a campaign paced correctly against October's prices runs hot in early November when CPMs jump with holiday competition, unless a human re-forecasts for seasonality.

Overspending early costs more than calendar coverage. Early exhaustion piles frequency onto whoever happened to be reachable in week one, skews reach away from the plan, and leaves the brand silent in the stretch, usually month-end or the holidays themselves, that the client cared about most. The dashboard still shows the budget fully spent. The client's question is what it was spent on, and when.

How pacing works in Google Ads, Meta, and the DSPs

Each platform paces differently, and agencies managing spending across multiple platforms need to know where the controls actually sit.

Google Ads runs on an average daily budget rather than a flight. The system can spend up to twice the daily budget on high-opportunity days, while holding total monthly spend to the daily budget times the average number of days in a month, so pacing checks in Google Ads happen at the monthly level: is cumulative spend tracking to the monthly plan, and are the strongest days getting the headroom. There is no end date doing the discipline for you, which is exactly why monthly budget pacing reviews exist.

Meta paces at the ad set or campaign level, with a choice between daily budgets and lifetime budgets. Lifetime budgets pace themselves to the schedule and shift spending toward the days the algorithm expects to perform, which works well until a mid-flight edit resets the learning. Campaign budget optimization moves money between ad sets automatically, so pacing checks in Meta include watching whether one ad set is quietly eating the campaign budget while others starve.

Demand side platforms are where pacing is most explicit: flights with start and end dates, budget at the campaign and line item level, even or ahead pacing options, and intraday distribution so the daily budget survives past breakfast. That explicitness is why programmatic pacing is the reference version of the discipline, and why a partner who cannot show you pacing in a DSP context should not be trusted with the platforms where the controls are vaguer.

One thing is constant everywhere: automated bidding and pacing interact. Bid strategies chase performance goals, pacing chases delivery goals, and when the two disagree, the platform resolves the argument silently unless a human is watching both numbers.

What the pacing algorithm can and cannot do

Inside the DSP, pacing works like this: the system takes the flight budget and end date, computes a spending trajectory adjusted for expected auction volume by day and hour, and then throttles the bidder, entering a calculated fraction of qualifying auctions rather than all of them, to hold the curve. Fall behind and the algorithm enters more auctions and, in some setups, bids higher. Run ahead and it throttles down. Intraday pacing spreads each day's budget across hours so spend distributes evenly instead of exhausting by 9 a.m.

Two things follow. First, pacing and performance trade against each other at the margin. A campaign forced to catch up buys less selectively, which is why the cheapest fix for underspending, spending harder at the end, is also the one that damages conversions and CPA most. Second, the algorithm optimizes only inside the constraints it was handed. It cannot widen a too-narrow audience, rescue a structurally losing bid, or call a publisher about an under-filling deal. Those are human jobs, which is the entire argument for treating pacing as an operational discipline instead of a platform feature.

The pacing checks a competent desk actually runs

Mature teams run pacing as a schedule with named checkpoints, not a dashboard someone glances at when things feel quiet. The routine below scales from one campaign to a hundred.

Before launch: forecast the deliverability

Every campaign gets checked against reality before it spends: can this targeting, at these caps and filters, at these bid levels, plausibly deliver this budget over this campaign duration? The DSP's forecasting tools plus historical performance data answer it in minutes, with seasonality and holidays weighted into the expected curve. Campaigns that cannot forecast to full delivery get their constraints renegotiated before launch, when the fix costs nothing, rather than after the miss, when it costs the relationship. This is also where KPIs and campaign objectives get written down, because pacing decisions later always involve trading delivery against performance goals, and that trade needs an agreed reference point.

Hours 24 to 48: verify the launch

Within two days of campaign launch, someone confirms the campaign is winning auctions, spending on trajectory, and serving on the intended inventory. Nearly every catastrophic delivery failure is visible in the first 48 hours, when it is still a cheap fix instead of a make-good conversation.

Every day after: watch the pacing index

Each live campaign carries the number from the formula above, refreshed daily on a pacing dashboard, with the 90 and 110 thresholds wired to alerts a human sees the day a campaign breaches, not at the Friday meeting. Daily monitoring sounds heavier than it is: with the index computed automatically, the daily pacing check for a healthy book of campaigns takes minutes, and the point of the routine is that the unhealthy ones announce themselves.

Weekly: review the curve, not just the number

Once a week the team looks past the index at the distribution: whether daily spending matches the plan, whether weekends are pulling their expected weight, whether creative fatigue or market conditions are moving the click through rates that feed delivery, and whether pacing adjustments made last week actually landed. This is also where budget reallocations across campaigns happen, moving funds from lines that cannot spend to lines that can, while the flight still has days left to benefit.

When a flight runs behind: intervene in order

The fix sequence matters. Raise bids modestly on the best performing inventory first. Then loosen the least valuable constraint: widen a geo, drop a marginal audience layer, add vetted supply. Only then pace harder, accepting the performance cost knowingly. Loosening brand safety comes last and only as a client conversation, never a silent toggle. When a flight runs hot, the same discipline inverts: tighten daily budget caps, check what the early spending actually bought, and let the frequency data say whether it was reach or repetition.

The final week: land it deliberately

The last week of a campaign flight gets its own check: remaining budget confirmed, required daily run rate computed, and any projected shortfall handled by decision rather than discovered on the reconciliation call.

Questions that expose a partner's pacing discipline

Pacing is where fulfillment vendors differ most in practice and least in their sales decks, so make the evaluation questions concrete:

1. Where does delivery versus plan appear in your standard reporting, and can I see the pacing dashboard mid-flight without asking?
2. Who gets alerted when a campaign breaches pacing thresholds, how fast, and what is the written intervention playbook?
3. Tell me about the last campaign that under-delivered and exactly what happened next. A partner with a real answer has been tested. A partner who says it does not happen has not been watching.
4. What do the make good terms say? Unspent budget belongs to the client, and shortfalls get made good on comparable inventory or credited, in the contract, not in the apology email.
5. How does your pacing interact with bid strategies and performance targets? The straight answer admits the trade off and describes buying selectively early so the flight never needs a desperate final week.

These five double as a proxy for the whole operation. A vendor that runs pacing checks with thresholds, alerts, and a playbook almost certainly runs QA, brand safety, and reporting the same way, because it is one habit applied to different numbers.

How Rambunctious Rhino handles pacing

Pacing at Rambunctious Rhino is deliberately boring. Delivery forecasts run before campaign launch, launch verification happens inside 48 hours, every flight carries a pacing index with human alerting on threshold breaches, and interventions follow the written order above, performance and brand safety protected while the flight catches up. And because the pricing takes no spread on media, nothing in the model gets paid for spending fast on cheap inventory or letting a flight quietly come up short. Campaigns finish complete, and the paperwork proves it without anyone having to write the apology email.