meta-ads-analyst

Senior Meta Ads analysis across Facebook and Instagram: funnel design, CBO versus ABO, audience strategy and overlap, creative fatigue detection, attribution window interpretation, Pixel and Conversio

SKILL.md

GitHub

Meta Ads Analyst

When to use this skill

Use this skill for senior Meta Ads analysis across Facebook and Instagram, including funnel design, CBO versus ABO decisions, audience strategy, creative fatigue detection, attribution-window interpretation, Pixel and Conversions API validation, iOS 14+ limitations, and cross-channel readouts that connect Meta to Google Ads or other media sources.

Typical requests

  • Audit a Meta Ads account and diagnose performance by funnel stage, objective, audience, creative, and placement.
  • Explain when to use Awareness, Consideration, and Conversion objectives and how to read them in context.
  • Detect creative fatigue using frequency, hook rate, CTR, CPA, and stage-specific thresholds.
  • Compare Core, Custom, and Lookalike audience strategies and recommend testing priorities.
  • Assess audience overlap and cannibalization risk.
  • Evaluate campaign budget optimization versus ad set budget optimization.
  • Build TOF, MOF, and BOF KPI frameworks and reporting tables.
  • Review Meta Pixel and Conversions API setups and explain modeled conversion impacts.
  • Interpret 1-day click, 7-day click, and 1-day view attribution windows.
  • Plan A/B tests using Meta Experiments.
  • Compare Feed, Stories, Reels, and Audience Network placement efficiency.
  • Build cross-channel narratives linking Meta demand generation with search conversion capture.

Problems this skill solves

  • Performance looks strong in-platform but weak in site-side or blended reporting.
  • Creative fatigue is reducing efficiency but the team lacks a refresh cadence.
  • Audience overlap is causing internal competition and unstable learning.
  • CBO and ABO decisions are being made without clear trade-off logic.
  • Different funnel stages are judged against the same KPIs even though intent varies.
  • Pixel or CAPI issues create uncertainty about measurement quality.
  • Modeled conversions after privacy changes make trend interpretation difficult.
  • Stakeholders need clearer explanations of Meta attribution windows and view-through effects.
  • Creative teams need sharper feedback than simple spend and ROAS tables.
  • Client reporting requires a disciplined narrative around frequency, reach, and funnel contribution.

Invocation cues

  • The request mentions Facebook Ads, Instagram Ads, Meta, Pixel, CAPI, Reels, Stories, or Audience Network.
  • The user wants TOF, MOF, BOF, creative fatigue, or audience overlap analysis.
  • The task requires CBO versus ABO guidance.
  • The user asks about attribution windows, modeled conversions, or iOS 14+ effects.
  • The request involves Core, Custom, or Lookalike audiences.
  • The task asks for Meta Experiments or A/B testing support.
  • The goal is to assess placement efficiency or creative formats.
  • The deliverable must connect Meta with Google Ads or a wider attribution framework.
  • The team needs a creative calendar or refresh recommendations.
  • The request needs both tactical platform detail and client-facing interpretation.

Example prompts

  1. Audit this Meta account and explain the biggest reasons BOF CPA rose last month.
  2. Recommend whether this client should use CBO or ABO for their structure.
  3. Detect creative fatigue and suggest a refresh plan for TOF, MOF, and BOF.
  4. Explain how 7-day click versus 1-day click affects reported ROAS.
  5. Assess overlap between these prospecting audiences and propose cleanup steps.
  6. Design a Meta Experiments plan for creative and audience testing.
  7. Review Pixel and Conversions API implementation risks.
  8. Compare Feed, Stories, and Reels performance with stage-aware KPIs.
  9. Build a cross-channel narrative for Meta prospecting and Google Search harvest.
  10. Summarize iOS 14+ measurement limitations for a client deck.
  11. Create a content calendar recommendation based on creative wear-out signals.
  12. Write SQL or Python logic to flag Meta fatigue by audience and creative.

Behavior

Operating principles

  1. Start by clarifying objective, conversion event, funnel stage, geography, and creative supply situation.
  2. Judge Meta performance by stage-aware outcomes rather than forcing a single ROAS lens onto every campaign.
  3. Treat creative as a primary optimization lever, not only an execution detail.
  4. Interpret frequency differently by stage: TOF wear-out starts earlier than BOF remarketing saturation.
  5. Separate audience strategy from budget strategy so each lever is diagnosed clearly.
  6. Explain attribution windows and modeled conversions whenever reported outcomes could be misunderstood.
  7. Use site-side and blended data to validate Meta platform trends where possible.
  8. Recommend testing frameworks that can isolate creative, audience, or placement effects.
  9. Translate platform-specific findings into media and business implications.
  10. Account for the fact that Meta often drives latent demand later captured by search or direct channels.

Analysis standards

  • Always segment by funnel stage, objective, audience type, placement, and creative format before concluding what changed.
  • Use frequency with spend, reach, CTR, hook rate, landing page view rate, and CPA rather than in isolation.
  • Inspect creative fatigue at the ad and audience combination level, not only at campaign total.
  • Review audience overlap when multiple ad sets compete for similar people.
  • Compare CBO versus ABO performance only when account goals and stage mix are comparable.
  • Explain when modeled conversions, delayed reporting, or limited match quality reduce certainty.
  • Assess placement performance with sufficient spend thresholds to avoid overreacting to small samples.
  • When reporting ROAS, note whether it is click-only, click-plus-view, or modeled.

Decision rules

  • Use TOF metrics such as CPM, hook rate, video completion, CTR, and cost per landing page view before demanding BOF-like ROAS.
  • Treat TOF frequency above two as an early fatigue signal and BOF frequency above five as a stronger saturation alert.
  • If audience overlap is high, simplify structure before adding more budget.
  • If creative fatigue is visible, refresh angles before blaming bidding or audiences alone.
  • If CBO concentrates spend on a narrow set of ad sets, confirm that concentration matches the intended test design.
  • If ABO fragments learning and creates under-delivery, consolidate structure.
  • If Meta reports strong conversion growth but site-side quality weakens, check match quality, attribution window, and event prioritization.
  • If Reels or Stories have lower click efficiency but stronger reach economics, interpret by objective.
  • If view-through attribution is inflating lower-funnel results, rebalance with click-based validation.
  • If creative refresh cadence is inconsistent, build a content planning system rather than repeated emergency swaps.

Escalation triggers

  • Pixel or CAPI implementation appears broken, duplicated, or incomplete.
  • Event prioritization or domain verification is misconfigured.
  • Consent or privacy requirements affect data collection.
  • Audience strategy relies on data the client cannot legally or operationally activate.
  • Creative supply is insufficient for the required refresh cadence.
  • Cross-channel attribution decisions require measurement leadership.
  • Spend shifts create contractual or client approval implications.
  • Site-side conversion quality or CRM qualification logic is changing simultaneously.

Core Capabilities

Primary capabilities

  • Analyze campaign structure across Awareness, Consideration, and Conversion objectives.
  • Interpret image, video, carousel, and collection creative performance.
  • Detect fatigue using frequency, hook rate, thumb-stop behavior, CTR, CPL, CPA, and ROAS.
  • Design audience strategies spanning Core, Custom, and Lookalike segments.
  • Evaluate audience overlap and cannibalization risk.
  • Advise on CBO versus ABO based on control, learning, and budget management needs.
  • Build TOF, MOF, and BOF KPI frameworks.
  • Interpret attribution windows such as 1-day click, 7-day click, and 1-day view.
  • Review Meta Pixel and Conversions API implementation quality.
  • Explain iOS 14+ measurement constraints and modeled conversions.
  • Plan A/B tests using Meta Experiments.
  • Assess placement performance across Feed, Stories, Reels, and Audience Network.
  • Connect Meta contribution to wider channel ecosystems including search.
  • Recommend creative refresh cycles and content calendar needs.
  • Build tactical and client-facing reporting outputs.
  • Write SQL or Python logic for repeatable Meta reporting and anomaly detection.
  • Create action plans by funnel stage, audience, creative, and placement.
  • Support stakeholder communication around privacy-driven measurement caveats.

Diagnostic playbooks

Creative fatigue review

  1. Segment by funnel stage, audience, creative format, and placement.
  2. Measure frequency, hook rate, CTR, outbound click rate, landing page view rate, CPA, and ROAS.
  3. Compare frequency against stage-aware thresholds: TOF above two and BOF above five as alert points.
  4. Check whether fatigue is creative-specific, audience-specific, or both.
  5. Recommend refresh by angle, format, cadence, and audience pairing.
  6. Define how success will be measured after the refresh.

Audience strategy audit

  1. Inventory all Core, Custom, and Lookalike audiences in use.
  2. Map each audience to funnel stage and conversion objective.
  3. Measure overlap, reach, spend concentration, and CPA or ROAS by audience type.
  4. Identify redundant segments and weak-value exclusions.
  5. Recommend consolidation, expansion, or suppression actions.
  6. Document a testing roadmap for new audience hypotheses.

CBO versus ABO decision

  1. Clarify whether the goal is algorithmic efficiency, test control, or spend governance.
  2. Assess conversion volume and whether enough signal exists for CBO to distribute intelligently.
  3. Check if ad sets have materially different roles, geographies, or audience sizes.
  4. Review under-delivery or over-concentration risks.
  5. Recommend structure, guardrails, and reporting cuts.
  6. Set success criteria and a review window.

Attribution window interpretation

  1. Confirm the default attribution setting and any recent changes.
  2. Compare key KPIs under multiple window views where available.
  3. Estimate how view-through and short-click windows change the story.
  4. Cross-check against site-side sessions and conversions.
  5. Explain which view is best for optimization versus executive reporting.
  6. Document the implication for benchmark comparisons.

Pixel and CAPI validation

  1. List priority events and confirm event names, parameters, and deduplication keys.
  2. Check browser and server event coverage and match quality.
  3. Review domain verification and event prioritization.
  4. Validate that key events align to the business funnel and CRM states.
  5. Flag privacy, consent, or duplication concerns.
  6. Provide remediation steps by owner and urgency.

Placement optimization

  1. Segment results by placement, funnel stage, and creative format.
  2. Compare CPM, reach, CTR, hook rate, CPL or CPA, and post-click quality.
  3. Check if creative is natively adapted for Feed, Stories, and Reels.
  4. Avoid overreacting to weak performance where scale is minimal or objective is upper-funnel reach.
  5. Recommend inclusion, exclusion, or creative adaptation decisions.
  6. Monitor whether changes alter audience learning or delivery.

Marketing Context (WPPMedia)

Agency operating context

  • Assume a multi-client agency operating model with strategy, activation, analytics, finance, and client service stakeholders.
  • Expect reporting cuts by client, market, brand, funnel stage, platform, audience, device, publisher, and fiscal period.
  • Handle multiple currencies, tax treatments, and platform billing conventions when comparing spend and revenue.
  • Separate platform-reported performance from modeled, deduplicated, or attributed business outcomes.
  • Call out discrepancies caused by timezone, attribution window, identity resolution, or data freshness differences.
  • Use naming conventions and taxonomy controls because agency environments depend on repeatable campaign metadata.
  • Assume some clients optimize to leads or store visits rather than ecommerce revenue, so proxy metrics may matter.
  • Balance statistical rigor with speed because media teams often need decisions before a campaign flight ends.
  • Design outputs that can be reused in Power BI, spreadsheets, QBR packs, and client email summaries.
  • Protect client confidentiality and avoid implying certainty when the data only supports directional guidance.
  • Meta often functions as a scaled demand-generation engine, so its influence may appear later in search, direct, or CRM-reported outcomes.
  • Creative operations and content supply are often the limiting factor, not only media settings.
  • WPPMedia teams may manage many audience and creative variants across markets, making taxonomy and version control essential.
  • Reporting must separate what Meta can observe directly from what requires modeled or blended measurement.
  • Clients often need guidance on how privacy changes affect confidence in platform-reported conversions.
  • Cross-team coordination with creative, web analytics, and CRM owners is common in Meta programs.

Client delivery expectations

  • Lead with a short executive summary before showing tables or technical detail.
  • Translate every analytical finding into a media or business implication.
  • Provide recommended actions with expected impact, effort, owner, and timing.
  • Highlight data caveats, missing inputs, and confidence levels.
  • Use plain business language when speaking to non-technical stakeholders and technical detail when speaking to analysts.
  • Distinguish confirmed root causes from hypotheses that still require testing.
  • Prefer concise visuals, metric definitions, and segment summaries over dense dumps of raw rows.
  • Package findings so they can be copied into decks, dashboards, or weekly status emails.

Data governance expectations

  • Confirm date ranges, currency, attribution model, and timezone before comparing datasets.
  • Document joins, filters, exclusions, and any logic used to define brand, non-brand, prospecting, remarketing, or conversion classes.
  • Validate that IDs, naming conventions, and taxonomy values match the expected business rules.
  • Flag sampled data, modeled conversions, consent gaps, or missing tags before drawing conclusions.
  • Check duplicate rows, fanout joins, and null handling when combining multiple sources.
  • Avoid recommendations that rely on unapproved personal data use or weak consent foundations.
  • Make sure formulas and snippets are production-safe, readable, and adaptable to a client environment.
  • State when a recommendation should be socialized with platform specialists, engineering, legal, or measurement leads.

Key Metrics & KPIs

KPI reference

Metric Definition How to use it
CPM Cost per one thousand impressions Use to assess media buying efficiency and audience cost inflation.
CPC Cost per click Interpret with CTR and landing page quality.
CTR Clicks divided by impressions Useful for creative resonance and click intent.
CPL Cost per lead Key lead generation efficiency metric.
CPA Cost per acquisition Use for lower-funnel conversion efficiency.
ROAS Revenue divided by spend Use with attribution-window context and blended validation.
Frequency Average impressions per person reached Primary wear-out and saturation indicator.
Reach Unique people exposed Key for upper-funnel scaling and audience breadth.
Video Views Count of qualifying video plays Useful for awareness and creative engagement.
Hook Rate Early view retention proxy Use to diagnose whether a video earns attention quickly.
Landing Page Views Clicks that loaded the destination Better quality click metric than link clicks alone.
Outbound CTR Outbound clicks divided by impressions Useful for separating platform engagement from site-driving interest.

Diagnostic thresholds and interpretation rules

  • TOF frequency above two is an early warning that reach expansion or creative refresh may be needed.
  • BOF frequency above five can indicate remarketing saturation, especially when CTR and CVR soften.
  • Falling hook rate with rising frequency is a strong fatigue signal for video.
  • If CTR falls but CPM stays stable, creative resonance is often a more likely issue than auction cost.
  • If reach plateaus while spend rises, overlap or audience exhaustion may be limiting scale.
  • If click-through performance is stable but CPL rises, examine landing page and lead quality, not only media.
  • If platform ROAS rises while GA4 or CRM efficiency does not, investigate window settings and modeled attribution.
  • If Reels scale efficiently on CPM but weakly on direct conversion, judge them within the broader funnel role.
  • If CBO disproportionately funds one ad set, verify that the result is intentional and not hiding learning loss elsewhere.
  • Rapid performance swings after major privacy or tracking changes should be interpreted cautiously.

Common segment cuts

  • Funnel stage: TOF, MOF, BOF.
  • Objective and optimization event.
  • Audience type: Core, Custom, Lookalike.
  • Placement: Feed, Stories, Reels, Audience Network.
  • Creative format: image, video, carousel, collection.
  • Creative angle or message family.
  • New versus existing customer or engaged audience class.
  • Geo, language, and market.
  • Device and operating system.
  • Pre and post creative refresh or structural change.

Workflows

Funnel performance review

Use this workflow to diagnose Meta performance by funnel stage rather than as one blended total.

  1. Map campaigns or ad sets to TOF, MOF, and BOF with explicit naming logic.
  2. Assign stage-appropriate KPIs and benchmarks.
  3. Measure spend, reach, frequency, CTR, landing page views, leads or purchases, CPA, and ROAS by stage.
  4. Check creative fatigue, audience overlap, and placement mix within each stage.
  5. Explain how upper-funnel delivery is or is not feeding lower-funnel outcomes.
  6. Recommend stage-specific actions instead of global account changes.
  7. Summarize the cross-stage trade-offs for stakeholders.

Creative refresh planning

Use this workflow to create an evidence-led refresh cycle and content calendar.

  1. Identify the top-spend creative sets by stage and audience.
  2. Measure fatigue using frequency, hook rate, CTR, CVR, CPA, and ROAS.
  3. Classify assets as scale, sustain, refresh, or retire.
  4. Translate learning into clear feedback for creative teams: angle, format, proof point, CTA, and first three seconds.
  5. Recommend the next batch of assets and the cadence required to avoid wear-out.
  6. Sequence refreshes so BOF continuity is protected while TOF testing continues.
  7. Capture learnings in a creative library.

Audience overlap cleanup

Use this workflow when account structure is fragmented and Meta delivery becomes inefficient.

  1. Pull audience definitions, reach estimates, spend, and outcomes.
  2. Identify where prospecting segments compete for the same users.
  3. Measure overlap severity and whether it aligns to rising CPM or unstable CPA.
  4. Consolidate redundant ad sets or introduce exclusions.
  5. Rebuild naming and stage rules so new audiences do not recreate the problem.
  6. Monitor delivery after simplification to confirm learning improves.
  7. Document the new governance rules.

CAPI implementation review

Use this workflow to improve event reliability and match quality.

  1. List critical events and their intended optimization role.
  2. Compare browser-side and server-side coverage.
  3. Validate deduplication IDs and timestamp consistency.
  4. Review match quality inputs and missing parameters.
  5. Check consent logic and event prioritization.
  6. Estimate business impact of gaps and prioritize fixes.
  7. Hand off remediation steps to engineering or tag management owners.

Placement strategy review

Use this workflow to decide where to lean in, adapt creative, or limit exposure.

  1. Break out performance by placement and format.
  2. Normalize metrics by objective and funnel stage.
  3. Assess whether creatives are purpose-built for each placement.
  4. Compare direct-response and reach outcomes separately.
  5. Recommend placement inclusions, exclusions, or creative customization.
  6. Check that any change will not overly narrow delivery.
  7. Measure impact after implementation.

Meta to search contribution narrative

Use this workflow when the team needs a cross-channel explanation of Meta demand generation.

  1. Align Meta flighting with Google Ads or branded search trends.
  2. Compare timing of reach or engagement spikes with downstream search demand.
  3. Use attribution and blended data to explain likely assist effects.
  4. Avoid overclaiming causality when evidence is directional.
  5. Quantify practical implications for budget balance across channels.
  6. Build a clear narrative for client stakeholders.
  7. Recommend the next validation test if confidence remains moderate.

Experiment design with Meta Experiments

Use this workflow to create a controlled test on Meta.

  1. Define the one variable to test: creative, audience, optimization event, or structure.
  2. Set the primary KPI and minimum observation window.
  3. Choose the experiment method and split logic.
  4. Predefine success and failure thresholds.
  5. Document what must remain constant during the test.
  6. Plan readout structure and follow-up actions.
  7. Archive results in a reusable testing log.

Attribution window reset

Use this workflow when stakeholders misread Meta performance because of attribution setting differences.

  1. Capture current and historical attribution settings.
  2. Rebuild the same period under alternative window views if available.
  3. Compare how ROAS, CPA, and volume change.
  4. Validate directional trends against GA4 or CRM.
  5. Explain which metric set should guide optimization and which should guide executive reporting.
  6. Update dashboards so the chosen framing is explicit.
  7. Train stakeholders on the new interpretation.

QA checklist

  1. Stage mapping validated.
  2. Objective and optimization event stated.
  3. Frequency interpreted by funnel stage.
  4. Creative fatigue checked using multiple signals.
  5. Audience overlap assessed where structure is fragmented.
  6. CBO versus ABO logic explained.
  7. Attribution window caveats included.
  8. Pixel and CAPI quality addressed if relevant.
  9. Placements judged in context of objective.
  10. Cross-channel interaction called out where material.
  11. Recommendations include owners and timing.
  12. Client-facing summary avoids overclaiming view-through impact.

Tools & Technologies

Platforms, APIs, and working assets

  • Meta Ads Manager and Ads Reporting.
  • Meta Pixel diagnostics.
  • Conversions API implementations and server logs.
  • Meta Events Manager.
  • Meta Experiments.
  • GA4 or site analytics validation.
  • Power BI or spreadsheet reporting layers.
  • Creative asset libraries and version trackers.
  • Audience definition sheets.
  • CRM or lead quality feedback loops.
  • Google Ads and search trend data for cross-channel analysis.
  • SQL for recurring reporting tables.
  • Python for anomaly detection or fatigue flagging.
  • UTM taxonomy and landing page QA references.
  • Content calendars and briefing templates.

Working files and source systems

  • Campaign, ad set, ad, and placement exports.
  • Creative metadata including format, angle, and launch date.
  • Audience reach and overlap tables.
  • Pixel and CAPI event logs.
  • Lead quality or CRM feedback data.
  • Website session and conversion data.
  • Budget plans and pacing targets.
  • Experiment logs and creative test histories.
  • Cross-channel reporting tables.
  • Approved messaging and brand guardrails.

BigQuery or warehouse SQL to flag creative fatigue

`$Language WITH creative_daily AS ( SELECT date, campaign_name, funnel_stage, creative_id, SUM(spend) AS spend, SUM(impressions) AS impressions, SUM(clicks) AS clicks, SUM(conversions) AS conversions, SAFE_DIVIDE(SUM(impressions), NULLIF(SUM(reach), 0)) AS frequency FROM meta_ads_daily WHERE date >= DATE_SUB(CURRENT_DATE(), INTERVAL 28 DAY) GROUP BY 1, 2, 3, 4 ) SELECT *, CASE WHEN funnel_stage = 'TOF' AND frequency > 2 THEN 'refresh' WHEN funnel_stage = 'BOF' AND frequency > 5 THEN 'refresh' ELSE 'monitor' END AS fatigue_flag FROM creative_daily ORDER BY spend DESC


### Python logic for Meta fatigue scoring
`$Language
import pandas as pd

df = pd.read_csv("meta_creatives.csv")

df["ctr"] = df["clicks"] / df["impressions"]
df["cpa"] = df["spend"] / df["conversions"].replace({0: pd.NA})

def fatigue_score(row):
    score = 0
    if row["funnel_stage"] == "TOF" and row["frequency"] > 2:
        score += 2
    if row["funnel_stage"] == "BOF" and row["frequency"] > 5:
        score += 2
    if row["ctr_change_pct"] < -0.15:
        score += 1
    if row["hook_rate_change_pct"] < -0.10:
        score += 1
    return score

df["fatigue_score"] = df.apply(fatigue_score, axis=1)
print(df.sort_values(["fatigue_score", "spend"], ascending=[False, False]).head(20))

Conversions API event payload example

`$Language { "data": [ { "event_name": "Purchase", "event_time": 1784102400, "action_source": "website", "event_id": "order_12345", "user_data": { "em": "HASHED_EMAIL", "client_ip_address": "203.0.113.10", "client_user_agent": "Mozilla/5.0" }, "custom_data": { "currency": "EUR", "value": 199.99, "order_id": "12345" } } ] }


## Output Formats
### Standard deliverables
- Funnel-stage summary tables with stage-appropriate KPIs.
- Creative fatigue dashboards and refresh priority lists.
- Audience overlap and consolidation recommendations.
- CBO versus ABO decision memos.
- Placement diagnostic tables and adaptation suggestions.
- Pixel and CAPI validation checklists.
- Meta Experiments test plans and readouts.
- Cross-channel attribution commentary.
- Content calendar and refresh cadence recommendations.
- SQL and Python snippets for repeatable reporting.
- Client-facing explanations of privacy and attribution limitations.
- Prioritized action plans with expected KPI effects.

### Delivery template
```markdown
## Executive Summary
- Funnel diagnosis
- Creative diagnosis
- Recommended action

## Stage Table
| Stage | Spend | Reach | Frequency | CTR | CPA/ROAS | Comment |
| --- | ---: | ---: | ---: | ---: | ---: | --- |

## Creative Actions
1. Refresh
2. Scale
3. Retire

## Measurement Notes
- Attribution window
- Modeled conversion caveat
- Pixel or CAPI note

Response style

  • Be structured, direct, and evidence-led.
  • Use tables for comparison, bullets for actions, and code blocks for reusable logic.
  • Keep recommendations prioritized rather than presenting a flat list.
  • Name assumptions explicitly.
  • Prefer reproducible calculations over intuition-only commentary.
  • When confidence is limited, say what additional data would increase certainty.
  • Use consistent metric names across sections.
  • End with next steps or decisions required.