The problem this solves
LinkedIn's cost per lead climbs quietly. A campaign that opened at a defensible number drifts upward as the audience saturates, and because the drift is gradual, the response is usually a budget shrug rather than a diagnosis. Meanwhile the platform keeps serving, the invoice keeps growing, and nobody can say whether the problem is who is being targeted, what they are being shown, or how the auction is being entered.
The three failure modes demand three different fixes, and guessing wrong burns another month. Fixing creative when the audience is exhausted, or rebuilding audiences when the bid strategy is the leak, is how ad budgets teach expensive lessons twice.
How the mission runs
- Read the accounts against benchmarks. The mission pulls campaign-level performance from your connected LinkedIn Ads account and lines CPL, CTR, and frequency up against B2B benchmarks, flagging the campaigns that have drifted past defensible.
- Diagnose the leak. For each flagged campaign, the evidence points at a culprit: saturated audience (frequency up, response down), tired creative (CTR decay on stable reach), or bid strategy (costs up on stable engagement). The verdict comes with its numbers.
- Rebuild the audience where that is the diagnosis. When the audience is the leak, a replacement is built from live buying-intent signals - companies in motion on your category right now - and staged as a matched audience for your approval.
- Prescribe the rest. Creative and bid diagnoses arrive as concrete prescriptions: which ads to retire, what angle the refresh should take, what bid change to test - each traceable to the data behind it.
- Report to the channel. The verdict lands where the team works: a summary of what is leaking, why, and what was staged or prescribed, posted to your performance channel.
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
A per-campaign diagnosis of your LinkedIn spend - audience, creative, or bid, each verdict with its evidence - plus a replacement audience built from live intent and staged for approval wherever saturation was the disease.
Make it yours
- Standing watch: run it weekly and only hear about campaigns that crossed the benchmark line since last run.
- Deeper rebuild: pair the new audience with a drafted refresh campaign so the fix ships as one reviewable unit.
- Portfolio view: run across every active campaign and get the reallocation picture, not just per-campaign fixes.
- Benchmark your own history: after a quarter of runs, the diagnosis compares campaigns against your account's own seasonal baselines rather than industry averages alone, which catches drift earlier and with fewer false alarms.
Frequently asked questions
Will it change my campaigns on its own?
No. Diagnosis is a read; the rebuilt audience stages for approval; prescriptions are recommendations. Anything that would alter a live campaign remains a human action.
Whose benchmarks are used?
Published B2B LinkedIn benchmarks, applied as context rather than gospel - the diagnosis leans primarily on your own campaigns' trends, where drift is measurable without industry averages.
What if the diagnosis is wrong?
Every verdict shows its evidence, so you can overrule it before anything is staged. Treat the diagnosis as a prepared case, and judge it the way you would an analyst's.
My CPL is fine - is there still a reason to run this?
A healthy number is the best time to learn what is carrying it. The same diagnosis run on a performing account shows which audiences and creatives do the lifting, which are passengers, and where frequency is quietly climbing toward saturation - the early warning that lets you refresh before the drift starts instead of after the invoice notices.