ASAP Pacing and Auto-Bidding Burn a Monthly Budget by Day 14
An advertiser sets a $90,000 monthly budget with a $55 CPA goal for a lead-gen campaign. The flight starts on the 1st, pacing is set to ASAP, and the ad group uses a performance bid strategy with a $22 max bid and a frequency cap of 10 impressions per user per day. The audience is a 2.2 million in-market segment.
By day 14, spend is $54,300 versus the $42,000 an even pace implies, and CPA has climbed from $49 in week one to $81 in week two; the 14-day CPA is $68. Win rate is 74%, average CPM has risen from $9.20 to $14.60, and reach has flattened at about 1.1 million unique users. The client's finance team is nervous about the burn rate and asks whether to cut the budget or pause the campaign until month-end.
1. Diagnose why spend is ahead of pace and CPA is rising. What is the platform doing, and what would you check to confirm each hypothesis?
Read the worked answer
Three mechanisms are interacting, and the diagnosis should separate them.
- ASAP pacing front-loads spend by design. It is telling the platform to win as much eligible inventory as possible, as early as possible. Spending 60% of the monthly budget in 14 days is the setting working as configured, not a malfunction.
- The performance bid strategy raises bids to capture that volume. With a $22 max bid, the system has room to bid up. CPM rising from $9.20 to $14.60 (+59%) is the signature of buying progressively more expensive, marginal supply. Check the bid landscape and the share of impressions won near the max bid.
- The frequency cap of 10/day against a 2.2M audience is effectively no cap. Reach plateaued at 1.1M - 50% of the audience - meaning the remaining spend is largely repeat impressions on the same users. Check the frequency distribution: if a large share of users are above 5–6 impressions, the incremental impressions are low-value.
Also check before concluding: conversion lag. Week-two CPA is likely overstated because recent clicks have not converted yet; compare cohort-based CPA rather than same-day reported CPA. And confirm whether the conversion event includes lower-intent actions (e.g., newsletter sign-ups) that inflate volume at a lower quality.
2. Recommend specific fixes with numbers. What settings, caps, and budget reallocation would you change for the remainder of the month?
Read the worked answer
Step 1: stop the front-loading. Switch pacing from ASAP to Even. Remaining budget is $90,000 − $54,300 = $35,700 over 16 days, which is $2,231/day versus the $3,879/day currently being spent. Even pacing enforces that trajectory without a manual pause, which preserves learning and avoids the restart loss a pause would cause.
Step 2: cap the bidding. Lower the max bid from $22 to roughly $14–15, in line with the CPM band that produced the week-one $49 CPA. If a performance strategy is retained, set a target CPA close to $55–60 rather than letting the algorithm chase volume at any price.
Step 3: tighten frequency. Move from 10/day to 4–5/day. On a 2.2M audience, the current cap allows more than 20M daily impressions - far beyond available supply - so it is not constraining anything.
Step 4: expand supply rather than repeat it. Add a lookalike of converters or a second in-market segment to move beyond the 1.1M plateau, and exclude users who have already converted or submitted a lead in the last 30 days.
Step 5: reallocate. Use dayparting toward the hours with the lowest CPA, and shift budget to the best-performing ad group or creative rather than spreading it evenly.
Numbers to state: at $35,700 remaining and a $55 goal, the campaign needs roughly 649 leads from the remaining 16 days. At the current $68 CPA it would produce about 525 - the gap is bridgeable with the bid and frequency changes, but not if spend continues at 1.7× pace.
3. How would you decide between cutting the budget and fixing the settings? Frame the decision for the finance team.
Read the worked answer
Cutting budget and fixing settings solve different problems, and conflating them is the error.
- Cutting the budget from $90,000 to, say, $60,000 changes how much is spent. It does not change ASAP pacing, the $22 bid ceiling, or the loose frequency cap. The campaign would still front-load, still buy marginal supply, and still miss the CPA goal - just with less money.
- Fixing the settings changes how efficiently each dollar is spent. That is the prerequisite for any conversation about the size of the budget.
Sequence to recommend:
- Apply pacing, bid, and frequency fixes now, and hold the total budget for 7–10 days.
- Judge results on cohort-adjusted CPA (allowing for conversion lag) and on marginal CPA by frequency bucket, not on the blended 14-day figure.
- If marginal CPA is still above $55 after the fixes, reduce the remaining monthly budget or shift it to a better-performing audience - at that point the constraint is genuine demand, not settings.
- Consider lead quality: a $55 CPL that produces poor downstream conversion is worse than a $70 CPL with strong conversion. If downstream data is available, make the budget decision on cost per qualified lead.
Message to finance: the current burn rate is a configuration outcome, not evidence that the channel is too expensive. Give the fixes one measurement window, with a pre-agreed CPA threshold that triggers the budget reduction.