Works with the AI Agents you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+22 moreClaude CodeCursorCodex CLIMuseDots
OpenClaw
Manus
Grok
Grok Bot+18 agentsCampaign Performance Doctor
Diagnose campaign results across strategy, audience, creative, delivery, conversion, customer quality, economics, and measurement—then identify what the evidence supports and what to test next.
$9.99
Campaign Performance Doctor
Example session with this skill installed
Using only the fictional data below, diagnose the campaign without assuming causality. Campaign A ran for 14 days. The ad platform reports 1,200 clicks, 60 purchases, and $6,000 attributed revenue on $2,000 spend. Web analytics reports 900 sessions and 24 purchases. CRM records 18 orders, but refunds, gross margin, new-versus-existing customer mix, and repeat purchase are unknown. A landing-page redesign shipped on day 8; tracking changed on day 9; no holdout or experiment was used. Summarize what is observed, reconcile the measurement gap, identify plausible explanations, calculate safe descriptive ratios, state what cannot be concluded, and propose the next evidence and tests needed before reallocating budget.
- Read your context and instructions
- Compiled the campaign performance doctor
Campaign A: evidence-led diagnosis
Executive read
The reports disagree materially, and two mid-flight changes complicate comparison. The platform reports 60 purchases and $6,000 attributed revenue; analytics reports 24 purchases; CRM lists 18 orders. This is a measurement reconciliation problem before it is a performance verdict. With no holdout or controlled experiment, the data do not establish incremental purchases or that the campaign caused the orders.
Observed facts and descriptive ratios
- Spend: $2,000 over 14 days
- Platform: 1,200 clicks; 60 attributed purchases; $6,000 attributed revenue
- Analytics: 900 sessions; 24 purchases
- CRM: 18 orders
- Platform-reported click-to-purchase rate: 60 / 1,200 = 5.0%
- Platform-reported attributed revenue / spend: $6,000 / $2,000 = 3.0x (attributed ROAS, not profit or incrementality)
- Analytics purchase/session rate: 24 / 900 = 2.7%
- Platform clicks exceed analytics sessions by 300; platform purchases exceed analytics purchases by 36; CRM orders are 6 below analytics purchases. These counts may use different definitions, windows, identity rules, and deduplication.
Confounders and checks
A landing-page redesign shipped on day 8 and tracking changed on day 9. Compare daily data with change markers; verify event firing, consent effects, attribution windows, timezone, click/session definitions, order IDs, duplicate events, cancellations, and refund treatment. Review device, channel, campaign, audience, geography, new/existing customer mix, and page performance only where reliable.
What cannot be concluded
We cannot claim 60 incremental purchases, a true 3x return, profitability, redesign lift, or media effectiveness. Gross margin, refunds, repeat behavior, customer quality, baseline, seasonality, and counterfactual are missing.
Next steps
- Reconcile order-level counts across platform, analytics, and CRM using consistent dates and definitions.
- Validate tracking before and after day 9; annotate the redesign and other changes.
- Add margin, refunds, new-customer status, and downstream quality to evaluate contribution.
- Compare cohorts cautiously; avoid attributing the redesign effect to media.
- Run a properly powered holdout or controlled test before making causal claims or major budget shifts. Preserve uncertainty in the retrospective and record decisions for follow-up.
Connects securely to your tools. The creator never sees your data.
About this skill
Campaign Performance Doctor
Turn campaign reports into a disciplined diagnosis of what happened, where value changed, what may have contributed, and what the available evidence cannot establish. Trace the system from objective, audience, proposition and creative through delivery, destination, conversion, customer quality, economics and measurement.
Reconcile platform, web analytics, CRM, sales, finance, inventory, and cohort signals when available. Separate observed facts from hypotheses and causal claims. Examine attribution windows, tracking changes, sample size, seasonality, promotions, audience mix, auction and delivery shifts, landing-page changes, sales handoff, returns, retention, and profit contribution before recommending action.
Produce a prioritized diagnostic, confidence ratings, evidence gaps, root-cause candidates, preserved practices, targeted tests, and a learning plan. Do not treat correlated movement as proof of cause, platform-reported conversions as incremental outcomes, or attributed revenue as profit. Never fabricate data, benchmark, experiment results, customer intent, or actions taken. Use authorized data only and keep customer-level information minimized. Human owners approve budget and campaign changes.
How to use
Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scanned
Verified clean 10 days ago
- Passed all security checks, Safe to install