The problem this solves
Campaign review meetings run on the metrics the platform hands over: impressions, clicks, cost per lead. Everyone in the room knows these are proxies, and everyone argues from them anyway, because the metric that would settle the argument - what a campaign's leads were actually worth - lives in other systems. So the meeting negotiates. The loudest advocate's campaign survives; the quiet campaign quietly printing qualified pipeline gets trimmed for looking expensive.
Once conversion events flow back to LinkedIn, the settling metric exists: cost per qualified conversion, per campaign, from your own definition of qualified. But it arrives scattered - spend in Campaign Manager, events in the conversion reporting, qualification logic in your head - and assembling it per campaign per period is exactly the analysis that gets promised after every meeting and delivered before none.
This mission assembles it and makes the call. Spend per campaign joined to conversion events per campaign, the true CPL computed with volume honesty, the ranking charted, and every campaign tagged kill, hold, or scale with the numbers that justify the tag.
How the mission runs
- Join spend to outcomes. Campaign spend for the period is joined to the conversion events attributed per campaign - the events your own loops streamed, carrying your own definition of a qualified outcome.
- Compute true cost per qualified conversion. Spend divided by qualified events, per campaign - with small-sample honesty: a campaign with two events is flagged as thin evidence rather than crowned or executed on noise.
- Chart the ranking. The campaigns render as a ranked bar chart with spend overlaid, so the meeting sees in one figure where money and outcomes diverge - the expensive-but-worth-it and the cheap-but-empty both unmistakable.
- Make the calls. Every campaign gets a tag - kill, hold, scale, or gather-more-evidence - each with its numbers and a sentence of reasoning. Recommendations only: budget changes remain yours to execute.
- Deliver for the meeting. The chart, the table, and the calls land as one artifact ready to open in the review meeting - the negotiation replaced by a ranking everyone can interrogate.
The prompt
This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.
What comes back
The campaign review, settled: true cost per qualified conversion per campaign with volume-honest flags, a ranked chart with spend overlay, and a kill / hold / scale call per campaign with the numbers attached. The meeting starts from evidence and spends its time on the genuinely contested cases.
Make it yours
- Weight events by value where deal sizes differ - cost per dollar of streamed revenue re-ranks harder than cost per event.
- Run it monthly against the same rules so the calls carry trend context: a campaign improving three months straight earns patience.
- Feed the scale winners into the audience playbook - their converters seed the next matched audience.
Frequently asked questions
What does this need in place first?
Conversion events flowing - from the CRM loop, the quality loop, or a backfill. Without events the mission falls back to lead-quality scoring as the outcome measure and says so plainly; with them, the ranking runs on real streamed outcomes.
How does it avoid killing a campaign on noise?
Volume honesty is built into the ranking: campaigns under an evidence threshold are tagged gather-more-evidence rather than kill, and the thin-sample flag rides every number derived from few events. Direction with small numbers, decisions with adequate ones.
Does the mission change budgets itself?
No. It delivers calls with evidence; moving money stays with you. Teams that want the next step wire the recommendation into their own review cadence - the mission's job is to make that review start from settled numbers.