A Shared Campaign Budget Eats the Prospecting Ad Group
A retail client hands you a $240,000 budget for an eight-week always-on campaign and wants both prospecting and site retargeting live at launch. You build one campaign with two ad groups: Prospecting, targeting a broad in-market segment, and Retargeting, targeting a 1P pixel audience of roughly 340,000 users. Campaign budget is $240,000, pacing is set to Even, flight is fixed to the eight weeks, and no ad group budgets are set. Both ad groups carry the same base bid of $9.50 CPM and a frequency cap of 5 impressions per user per day.
At the end of week one, total spend is $11,400. Retargeting has consumed $8,700 of it at a $13.80 CPM; prospecting has spent $2,700 at $6.10 CPM and is delivering under 20% of its forecast impressions. Retargeting CPA is $38; prospecting CPA is $112 against a $70 blended goal. The client wants to know why the "cheap" prospecting ad group is the problem and whether to pause retargeting.
1. Diagnose the delivery and efficiency pattern. What is actually happening inside this campaign structure, and what would you verify in the platform before changing any settings?
Read the worked answer
The pattern is budget cannibalization, not a bad prospecting audience.
- No ad group budgets were set, so both ad groups draw from one $240,000 pot. The retargeting audience is small (340,000) and highly qualified; it wins a large share of the auctions it is eligible for and takes spend first. The campaign-level budget never restricts it.
- Prospecting's realized $6.10 CPM against a $9.50 base bid is the tell. It is only winning the cheapest, least contested impressions. It is not performing "cheaply"; it is being outbid on the quality supply that would convert. Low CPM here is a symptom of losing the auctions that matter.
- Retargeting's 5/day cap is probably binding. That limits its volume, which caps how much of the budget it can absorb and pushes the remainder into weak prospecting supply. The high retargeting CPM is expected and fine given its $38 CPA.
Verify before editing: ad group win rate and available impressions; unique reach and frequency distribution per ad group; conversions by ad group under the intended attribution window; whether the pixel fires once per order (not on multiple pages); and whether both ad groups are optimizing to the same conversion event. If retargeting is genuinely capped and efficient, pausing it would be the wrong move - it would just strand more budget in underperforming prospecting supply.
2. Propose a revised structure and budget allocation for the remaining seven weeks, with numbers, and explain the trade-offs of your choice.
Read the worked answer
Separate the budgets and the jobs. Remaining spend is $228,600 over seven weeks.
Option A (recommended): one campaign, explicit ad group budgets.
- Prospecting: $137,000 (60%), base bid raised to $11–12 CPM so it can compete for quality inventory.
- Retargeting: $68,500 (30%), bid reduced to $10–11 CPM, frequency cap tightened to 3/day.
- Test/expansion ad group: $22,900 (10%) for a lookalike or second in-market segment.
Option B: split into two campaigns. Cleaner reporting and independent pacing, but you lose shared budget flexibility and add management overhead; it is worth it only if the client needs separate KPIs per funnel stage.
Bidding: keep base bids until prospecting has ~50 conversions per week, then test a performance/CPA bid strategy with a goal near $85 (prospecting CPA must be allowed to run above the blended $70 because retargeting subsidizes it). Do not apply an auto-bid strategy to a starved ad group - it will overbid to chase volume.
Trade-off to state explicitly: hard ad group budgets protect prospecting from cannibalization but remove the campaign's ability to shift spend to whichever ad group is efficient on the day. That flexibility is exactly what caused the week-one outcome, so the cap is the point.
3. What QA checklist would you run before relaunching, and which single check is most likely to invalidate the new plan?
Read the worked answer
Pre-relaunch QA:
- Budget math: ad group budgets sum to no more than the campaign budget, and the remaining flight days match the remaining spend. A mismatch here silently re-creates the original problem.
- Flight and time zone: campaign start/end dates and the advertiser time zone are correct; ad group dates are not defaulting past the campaign end.
- Audience integrity: retargeting pixel audience is populated and growing, not stalled; the prospecting segment is available in the target geography; no accidental overlap.
- Frequency and exclusions: caps set per ad group; recent converters and existing customers excluded from prospecting if the goal is new-customer acquisition.
- Creative: all sizes/format approved, no disapprovals, landing pages resolving, click trackers not double-redirecting.
- Conversion tracking: pixel fires once per order, correct event is mapped to the optimization goal, attribution window documented.
Most likely plan-invalidating check: audience size versus budget. If the retargeting pool is only 340,000 users and you allocate $68,500 over seven weeks at $13 CPM, that is roughly 5.3 million impressions - an average frequency near 15 unless the pool grows. If the pool is not growing fast enough, cap retargeting lower and move the difference to prospecting before launch, not after week two.