mardab96 Weekly Tracking Signal Readout
Use the shared quality bar in ../references/output-standard.md and ../references/skill-design-principles.md when those files are available.
Use this skill when
- the user shares GTM, GA4, Google Ads, Meta Pixel/CAPI, consent, click ID, data layer or CRM import evidence tied to weekly tracking signal readout.
- the next decision could change conversion actions, tags, triggers, event parameters, imports, consent checks or bidding signals.
- platform counts disagree with backend, CRM or debug traces, and the team needs to know which signal to trust.
Do not use this skill for broad analytics theory. Use it when tracking evidence, platform counts, debug traces or import data must be checked before an operator trusts the signal.
Required input
- GTM container export, tag list, GA4 event list, Google Ads conversion actions, Meta Pixel/CAPI setup or screenshots.
- debug traces, preview screenshots, data layer examples, click IDs, consent state and recent tracking changes.
- platform reports for the same time window so browser, server, GA4 and ad platform counts can be compared.
- what decision the user is trying to make next: verify, debug, deduplicate, import offline conversions or assess readiness.
- If an input is missing, continue with a clearly marked assumption instead of inventing data.
Analysis workflow
- Summarize changes in tags, consent, events, conversion actions, imports and diagnostics during the week.
- Compare key event counts across GA4, ad platforms, backend and CRM for the same window.
- Call out anomalies, duplicates, missing values, match-rate changes, import lag and consent impact.
- Assign a tracking confidence score by platform and event.
- Return approval asks, watch items and one highest-priority fix.
Decision rules
- If the data does not connect to revenue, pipeline, qualified leads or conversion quality, label the recommendation as a hypothesis.
- If platform metrics and downstream data disagree, trust the downstream source for business quality and platform data for delivery mechanics.
- If the issue could be tracking, offer, audience, page or follow-up, do not collapse it into one cause without evidence.
- Do not tell the user to edit live tags, containers, consent settings or conversion actions without an approval step.
Output format
| Finding | Tracking evidence | Signal risk | Recommended fix or check | Confidence |
|---|---|---|---|---|
| Specific diagnostic claim | Data, screenshot, report, note or missing-data marker | Business or signal consequence | Smallest useful next step and owner | High / Medium / Low |
End with:
Decision:fix / test / monitor / ask for data / do not act yetApproval needed:yes/no and what would change if approvedMissing data:only the inputs that would materially change the recommendation
Practical example
User: "Here are GTM screenshots, platform counts and debug traces for weekly tracking signal readout. Is this signal safe to trust?"
Assistant should: use the supplied evidence, run the workflow above, produce the skill-specific rubric or diagnostic table, and stop at approval-ready recommendations.
Guardrails
- Do not make changes to live campaigns, pages, tags, containers, CRM fields or customer messages.
- Do not claim performance impact without evidence.
- Mark missing data clearly.
- Keep recommendations practical for a performance operator, founder or owner with a real advertising problem.