agentleFS
narayan-metaflownarayan-metaflow/metaflow-marketing-skills/skills/paid-media-reporter/SKILL.md

paid-media-reporter

Skill14 starsChanged 7 months agoMITView on GitHub

Build performance reports, dashboards, and stakeholder communications that turn ad data into business narratives — client reports, QBRs, audit deliverables, and executive summaries.

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 10 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈10

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every report should answer four questions:

    line 43
  2. Distinctive:

    Rule: If the "is this working?" answer takes more than 5 seconds to find, the dashboard is broken.

    line 177
  3. Distinctive:

    When onboarding a new client or performing crisis diagnosis, deliver an audit report structured as:

    line 207
  4. Distinctive:

    Cross-reference: Use the google-ads-audit skill for Google deep audits and meta-ads-audit skill for Meta deep audits.

    line 242
  5. Distinctive:

    Every report section should deliver 3 specific things:

    line 254
  6. Distinctive:

    When asked to produce a report, deliver:

    line 287
  7. Distinctive:

    September was a recovery month: After August's mid-month fatigue dip on Meta ASC (frequency hit 5.4 on top creatives), we executed a …

    line 349
  8. Distinctive:

    [ICE: 480]: Run Meta prospecting incrementality test (geo split) — 14 days

    line 375
  9. Distinctive:

    Don't report in isolation: Always include competitive/market context (CPM trends, platform changes, seasonality).

    line 395
  10. Distinctive:

    Don't over-report in crisis: When something breaks, send a pulse + action plan, not a 30-page analysis.

    line 399

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Process

  1. Identify the audience — CMO, channel manager, finance, or founder? Different metrics for different eyes
  2. Select cadence — Daily pulse, weekly tactical, monthly strategic, quarterly business review
  3. Choose metrics by level — Don't overwhelm; match metric depth to audience depth
  4. Build the narrative — Data alone fails. Story: what happened → why → what's next
  5. Design for scanability — Executives skim; surface the 3 things that matter
  6. Tie to business outcomes — Not impressions, but revenue, pipeline, CAC, LTV
  7. Include competitive/market context — "Meta CPMs rose 9% this year" explains some decline
  8. Deliver an action plan — Every report ends with "what we're doing next"

Report Types & Audiences

Report Cadence Framework

CadenceAudienceDepthTime Invested
Daily pulseChannel specialist / teamAlert-driven, raw metrics<5 min/day (automated)
Weekly tacticalAccount manager + client point-of-contactTrends, actions taken, immediate issues30-60 min
Monthly strategicMarketing director / VPFull performance, wins, recommendations2-4 hours
Quarterly business review (QBR)CMO, executives, foundersStrategic narrative, forecasts, budget decisions4-8 hours
Ad hoc auditNew client onboarding, crisis momentsFull diagnostic with ICE-scored recommendations4-8 hours

Audience-Tailored Metric Selection

AudienceCares AboutBores ThemLead With
CMO / C-suiteRevenue impact, ROAS trends, market share, competitive positionCTR, QS, ad-group structureMER, revenue, 30/60/90-day trend
Finance / CFOCAC, LTV, blended CAC, payback period, MERCreative fatigue, bid strategiesCAC trend, LTV:CAC ratio, cash efficiency
Channel managerCTR, CPC, CVR, QS, creative performance, impression shareHigh-level narrativePlatform-specific deep metrics
Founder (SMB)Revenue, new customers, is this workingAcronyms, technical detailPlain English: "spent $X, got $Y, here's what's next"
Agency account managerPacing, client satisfaction levers, scaling opportunitiesPure technical deep-divesStatus summary + priorities

Critical rule: Lead with the metric the audience cares about. CMO reports that lead with "CTR improved 0.3%" fail. CMO reports that lead with "Revenue grew 22% at constant MER" succeed.

The Narrative Framework

Every report should answer four questions:

QuestionPurposeExample
What happened?Factual summary"Revenue grew 18% MoM, MER held at 3.4x, CAC down 8%"
Why did it happen?Causal explanation"Creative refresh in Week 2 reversed fatigue; new UGC variants lifted ROAS 22%"
What's next?Action plan"Scale winning creative concepts, test 3 new hooks, refresh retargeting pool"
What should you decide?Stakeholder decision points"Approve $5K budget increase for scaling Meta ASC + $3K for YouTube test"

Don't: Report metrics without narrative. Do: Tell the story of the month with metrics as evidence.

Monthly Report Template

# [Client Name] — Performance Report
## [Month Year]

## Executive Summary
- **Headline:** [one sentence on the month]
- **Revenue:** $[X] ([+/-X%] MoM)
- **Spend:** $[X] ([+/-X%] MoM)  
- **MER:** [X.X]x ([+/-X%] MoM)
- **Key win:** [most important positive]
- **Key concern:** [most important issue]
- **What we're doing next:** [top 3 actions]

## Performance at a Glance
| Metric | This Month | Last Month | MoM | vs Target |
|--------|-----------|-----------|-----|----------|
| Revenue | $[X] | $[X] | [+/-%] | [status] |
| Ad Spend | $[X] | $[X] | [+/-%] | [status] |
| MER | [X]x | [X]x | [+/-] | [status] |
| New Customers | [X] | [X] | [+/-%] | [status] |
| Blended CAC | $[X] | $[X] | [+/-%] | [status] |

## Channel Breakdown
| Channel | Spend | Revenue | ROAS | Trend | Role |
|---------|-------|---------|------|-------|------|

## What Happened — The Story
[2-3 paragraphs explaining performance, what drove changes, key insights]

## What's Working
- [3 specific wins with data]

## What's Not Working
- [2-3 specific issues with diagnosis]

## Actions Taken This Month
- [Key optimizations, tests launched, changes made]

## Next Month's Plan
- [ICE-prioritized actions]
- [Tests planned]
- [Budget recommendations]

## Decisions Needed From You
- [specific asks for approval or input]

QBR Template (Quarterly Business Review)

# [Client Name] — Q[X] Business Review

## Agenda
1. Q[X] at a Glance (5 min)
2. Performance Deep Dive (15 min)
3. Strategic Wins & Learnings (10 min)
4. Market Context & Competitive Position (5 min)
5. Q[X+1] Strategy & Forecast (15 min)
6. Budget & Decisions (10 min)

## Q[X] At a Glance
- **Total Revenue:** $[X] ([+/-X%] QoQ, [+/-X%] YoY)
- **Total Spend:** $[X]
- **Blended MER:** [X]x (target: [X]x)
- **Blended CAC:** $[X] (target: $[X])
- **LTV:CAC:** [X]:1
- **Market position:** [competitive context]

## Performance Narrative
[What happened this quarter, key turning points, strategic decisions and outcomes]

## Channel Scorecard
| Channel | Q[X-1] ROAS | Q[X] ROAS | Trend | Action |
|---------|------------|-----------|-------|--------|

## Strategic Wins
- [3-5 biggest wins and what drove them]

## Key Learnings
- [what didn't work and what we learned]

## Market Context
- [CPM trends — Meta +9% in 2025]
- [Platform changes — AI Max, attribution windows, etc.]
- [Competitive shifts]

## Q[X+1] Strategy
### Strategic Priorities
1. [Priority 1 with rationale]
2. [Priority 2 with rationale]
3. [Priority 3 with rationale]

### Budget Allocation
[Proposed allocation with scenarios: conservative / base / aggressive]

### Key Tests Planned
- [Creative tests]
- [Channel tests]
- [Incrementality tests]

### Forecast
| Scenario | Spend | Expected Revenue | MER |
|----------|-------|-----------------|-----|

## Decisions Requested
1. [Budget approval]
2. [Strategic direction]
3. [Resource allocation]

Dashboard Design Principles

The 5-30-5 Rule

LayerTime to ComprehendContent
5 secondsTop 3 numbers that define "is this working?"MER, Revenue, CAC vs targets with traffic lights
30 secondsTrend charts and channel split30/60/90-day trendlines, channel contribution
5 minutesDrill-down detailCampaign-level tables, creative performance, anomaly callouts

Rule: If the "is this working?" answer takes more than 5 seconds to find, the dashboard is broken.

Dashboard Stack (Common Tools)

ToolBest ForCostData Connectors
Looker StudioClient reports, free tierFreeNative Google, limited Meta (needs connector)
SupermetricsConnector layer for Sheets/Looker/BI$99+/moAll major ad platforms
ImprovadoEnterprise unified reportingEnterprise pricingAll major + CRM + warehouse
Google Sheets + Apps ScriptCustom, flexibleFreeVia API or Supermetrics
Tableau / Power BIEnterprise BIEnterpriseVia warehouse
OptmyzrPaid media-specific reportingPaidGoogle + Meta + more

2025 consideration: Meta Insights API removed 7d_view and 28d_view attribution windows (October 2025). Any dashboard pulling those windows needs updating.

Core Dashboard Widgets

WidgetWhat It ShowsWhy
Spend pacingMTD spend vs targetDaily budget health
MER trend (90 days)Blended efficiency over timeStrategic trajectory
Channel contributionRevenue split by channelAllocation visibility
Creative leaderboardTop 10 creatives by ROASCreative performance
Anomaly flagsMetrics >20% off baselineAlert-driven attention
Impression share (Google)Coverage vs opportunityUnder/over-investment signals
Frequency (Meta)Fatigue leading indicatorCreative refresh timing
Forecast vs actualsPredicted vs realForecasting credibility

Audit Report Framework

When onboarding a new client or performing crisis diagnosis, deliver an audit report structured as:

# Paid Media Audit — [Client Name]

## Executive Summary
- **Account health score:** [X]/100
- **Estimated monthly waste:** $[X]
- **Top 3 opportunities:** [list]
- **Estimated impact of fixes:** [+X% revenue / -Y% CAC]

## Findings (by severity)
### Critical Issues (fix this week)
| Issue | Impact | Evidence | Fix |
|-------|--------|----------|-----|

### High-Priority (fix this month)
| Issue | Impact | Evidence | Fix |
|-------|--------|----------|-----|

### Medium-Priority (fix this quarter)
| Issue | Impact | Evidence | Fix |
|-------|--------|----------|-----|

## ICE-Scored Action Plan
| Priority | Action | ICE Score (I×C×E) | Owner | Timeline |
|----------|--------|------------------|-------|----------|
| 1 | [action] | 10×9×8 = 720 | [who] | [when] |

## Monitoring Plan
[What to check and when]

## 30/60/90-Day Expected Outcomes

Cross-reference: Use the google-ads-audit skill for Google deep audits and meta-ads-audit skill for Meta deep audits. This skill structures the deliverable — those skills drive the analysis.

Performance Narrative Rules

Good Narrative Structure

Bad: "Revenue was $150K, ROAS was 3.2x, CAC was $45."

Good: "Revenue grew 18% to $150K as the new UGC creative rotation reversed last month's fatigue decline. MER held at 3.2x despite rising Meta CPMs (+9% YoY). Blended CAC dropped 8% to $45 as retargeting efficiency improved after audience refresh. Action: scale winning UGC concept with 5 new variants in Week 1 of next month."

The Rule of 3s

Every report section should deliver 3 specific things:

  • 3 wins
  • 3 issues
  • 3 actions

More than 3 = scattered. Fewer = unsubstantive. The discipline of 3 forces prioritization.

Connect to Business Outcomes

MetricVanity VersionBusiness Version
Impressions"5M impressions""Reached 40% of target market"
Clicks"50K clicks""8,000 new site visitors at 1.6% of ad spend"
CTR"2.3% CTR""CTR improved 15% after creative refresh, driving 7% revenue lift"
CPA"$35 CPA""CAC $35 → $32 improved blended CAC to $48, payback period now 4.2 months"

Always connect the metric to revenue, customers, or cash outcomes.

Common Reporting Mistakes

MistakeWhy It's BadFix
Reporting every metricReader can't prioritizeLead with top 3, put rest in appendix
No context"CTR is 2.3%" — is that good?Include target, benchmark, or trend
Metrics without storyData dumpExplain what happened and why
No action plan"Here's the data, goodbye"Always end with next steps
Same report for all audiencesCMO doesn't care about QSTailor metrics to audience
Hiding bad newsKills trust when discoveredBe transparent, always pair with fix
Over-reporting in crisisAnalyzes instead of actingPulse report + action; full analysis later
No competitive contextDecline blamed on team when it's macroInclude market context (CPM trends, competitor moves)

Output Format

When asked to produce a report, deliver:

## [Report Type] — [Client/Account]
## [Period]

### Executive Summary
[3-5 bullets with the story and key numbers]

### Key Metrics
[Tabular comparison vs prior period / target]

### What Happened
[Narrative: wins, issues, causes]

### Actions Taken
[What you did this period]

### What's Next
[ICE-prioritized action plan for next period]

### Decisions Needed
[Specific asks of stakeholder]

Example

Input: "I'm an agency account manager. I need to write a monthly report for a DTC skincare client. September 2026 numbers: spend $48K ($18K Google, $22K Meta, $5K TikTok, $3K YouTube), revenue $165K, 1,200 new customers, MER 3.44x. Previous month: spend $45K, revenue $153K, MER 3.40x, 1,100 new customers. Client is concerned because last week Meta CPMs were high and some campaigns dipped. What should the report look like?"

Output:

Monthly Performance Report — [Skincare Brand]

September 2026

Executive Summary

  • Revenue grew 7.8% MoM to $165K on a 6.7% spend increase, with MER improving from 3.40x → 3.44x
  • New customers up 9.1% to 1,200, driven by Meta ASC creative refresh and Google Search impression share recovery
  • Blended CAC dropped 1.5% to $40, improving payback period toward 4.0 months
  • Key concern addressed: Meta CPMs did spike in Week 3 (+12% WoW) due to seasonal competitive pressure, but our frequency-based creative rotation kept ROAS stable
  • Next month focus: Scale Meta ASC winners, launch Google PMax new customer acquisition goal, test YouTube expansion

Performance at a Glance

MetricSep 2026Aug 2026MoMvs Target
Revenue$165,000$153,000+7.8%✅ Above target
Ad Spend$48,000$45,000+6.7%✅ On budget
MER3.44x3.40x+1.2%✅ Above 3.2x target
New Customers1,2001,100+9.1%✅ Above target
Blended CAC$40.00$40.91-2.2%✅ Below $45 target

Channel Breakdown

ChannelSpendRevenueROASTrendRole
Google Search$10K$60K6.0x⬆️Primary driver
Google PMax$8K$26K3.25x→Scaling
Meta ASC$15K$48K3.2x⬆️Primary driver
Meta Prospecting$7K$14K2.0x⬇️Scaling — watch
TikTok$5K$8K1.6x→Experimental — under review
YouTube$3K$5K1.67x→Support — underfunded

What Happened — The Story

September was a recovery month. After August's mid-month fatigue dip on Meta ASC (frequency hit 5.4 on top creatives), we executed a 6-variant creative refresh in Week 1. By Week 3, frequency normalized to 3.2 and ROAS recovered from 2.7x → 3.4x on Meta ASC.

The CPM spike you noticed in Week 3 was real — Meta's auction pricing rose 12% WoW due to Q4 prep from competing advertisers. Despite that headwind, our MER improved because (a) the creative refresh absorbed impressions at better conversion rates, and (b) Google Search captured more demand as PMax feeder creative drove bottom-funnel intent.

Google Search impression share improved to 72% (from 64% in August) after we expanded non-brand keyword coverage based on search term mining. This was the biggest contribution to revenue growth.

What's Working

  • Meta ASC creative refresh playbook — 6-variant weekly refresh is keeping frequency below 4.0 and preventing fatigue-driven decline
  • Google Search non-brand scaling — impression share gains are producing efficient incremental revenue at 3.5x marginal ROAS
  • Blended CAC discipline — despite CPM pressure, CAC dropped 2.2% because conversion rates improved

What's Not Working

  • Meta prospecting ad set underperforming — 2.0x ROAS below target. Likely feeding ASC but marginal incrementality is unclear. Recommend incrementality test before reallocating
  • TikTok at $5K is undersized — below signal threshold for meaningful learning. Either commit more or sunset
  • YouTube at $3K is underfunded — same issue, can't drive real signal at this spend

Actions Taken This Month

  • Launched 6-variant Meta ASC creative refresh (Week 1)
  • Expanded Google non-brand keyword coverage (Week 2)
  • Refreshed Meta retargeting audiences (Week 3)
  • Paused 4 fatigued creatives at frequency >5 (ongoing)
  • Added automated frequency-pause rule at 5.0 threshold

Next Month's Plan (October 2026)

  1. [ICE: 720] Scale Meta ASC with 8 new creative variants targeting winning UGC concept
  2. [ICE: 640] Launch Google PMax new customer acquisition goal (predicts high-LTV users)
  3. [ICE: 480] Run Meta prospecting incrementality test (geo split) — 14 days
  4. [ICE: 420] Decision: commit $10K to YouTube OR sunset — current $3K is wasted
  5. [ICE: 360] Creative pipeline: brief 10 UGC variants for November buffer

Decisions Needed From You

  1. YouTube investment decision: Scale to $10K/mo or sunset? Current $3K is below signal threshold
  2. Meta prospecting: Approve geo-split incrementality test (no additional budget, just a 14-day measurement)
  3. October budget approval: Holding at $48K with reallocation, or approving $55K for Q4 scaling?

Market Context

  • Meta CPMs continue to trend +9% YoY in 2025 — rising costs are a macro headwind, not a team failure
  • Google AI Max (global beta September 2025) is driving avg 34% conversion lift for early adopters — evaluating for October rollout
  • Competitive CPM pressure from Q4-prepping advertisers expected to continue through mid-November

Guidelines

  • Don't dump every metric in the report. Lead with 3-5 headline numbers. Put detail in appendix or drill-down.
  • Don't report without a narrative. Data without story = noise. Every report explains what happened, why, and what's next.
  • Don't forget the audience. CMO reports differ from channel-manager reports. Tailor metric depth to decision depth.
  • Don't bury bad news. Transparency builds trust. Pair every issue with a diagnosis and fix.
  • Don't report in isolation. Always include competitive/market context (CPM trends, platform changes, seasonality).
  • Don't end without decisions needed. Reports that ask for nothing leave stakeholders passive. Force the decision moments.
  • Don't ignore the 7d_view/28d_view removal (October 2025). Any report still pulling those attribution windows is broken. Update.
  • Don't report attribution as fact. Platform attribution overcounts. Use MER as your north star. Reference the ad-spend-allocator skill for cross-channel allocation decisions.
  • Don't over-report in crisis. When something breaks, send a pulse + action plan, not a 30-page analysis. Analysis comes later.
  • Don't build dashboards for yourself. Build for the reader. The "5-30-5 rule" — headline in 5 seconds, context in 30, detail in 5 minutes.
  • Cross-references: For audit analysis, use the google-ads-audit and meta-ads-audit skills. For cross-channel allocation decisions, use the ad-spend-allocator skill. For deep campaign analysis that feeds into reports, use the campaign-analyzer skill.
  • Confidence: Report templates are starting points — calibrate to your specific client expectations and contract scope. Tool recommendations reflect 2025 landscape; new tools emerge regularly.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.