marketing-funnel-diagnosis
Skill0 starsChanged 42 days agoMITView on GitHub
Use when conversion falls, acquisition channels disagree, organic landing pages disappear, or a funnel dashboard suggests a bottleneck before measurement, cohort maturity, and traffic mix have been checked.
In this repository66 other files it gives its agents.
- AGENTS.mdAGENTS.md
- postgres-locking-joinsSkill
- ai-bug-patternsSkill
- compressed-ndjson-boundariesSkill
- http-resume-integritySkill
- linux-io-latency-triageSkill
- macos-disk-space-recoverySkill
- minimal-sufficient-evidenceSkill
- modularize-large-filesSkill
- patch-reproducer-validitySkill
- prompt-cache-boundariesSkill
- proton-drive-selective-backupSkill
- transactional-outbox-deliverySkill
- conventional-commitsSkill
- git-advanced-mergingSkill
- git-bisect-forensicsSkill
- git-data-recoverySkill
- git-hooksSkill
- git-rerereSkill
- git-reset-mechanicsSkill
- git-revisions-syntaxSkill
- git-rewriting-historySkill
- git-submodulesSkill
- git-worktreesSkill
- github-ssh-keysSkill
- npm-publish-ci-pitfallsSkill
- sapling-stacked-pr-workflowSkill
- go-concurrency-patternsSkill
- go-data-idiomsSkill
- go-error-handlingSkill
- go-interfaces-embeddingSkill
- go-pprof-profilingSkill
- marketing-geo-experimentsSkill
- marketing-incrementalitySkill
- marketing-positioning-testsSkill
- python-descriptors-mroSkill
- python-logging-cookbookSkill
- python-scaling-regression-testsSkill
- python-unicode-gotchasSkill
- rust-compile-time-invariantsSkill
- rust-deterministic-concurrency-testsSkill
- rust-error-identity-boundariesSkill
- rust-regex-search-contractsSkill
- rust-runtime-bug-classesSkill
- rust-tokio-actor-deliverySkill
- rust-tokio-cancellation-safetySkill
- sales-discovery-qualificationSkill
- sales-mutual-action-plansSkill
- sales-pipeline-forecast-calibrationSkill
- sales-value-negotiationSkill
- ai-triage-evaluationSkill
- container-image-patchingSkill
- dependency-artifact-integritySkill
- dnssec-sentinel-diagnosticsSkill
- iam-policy-change-replaySkill
- rbvm-applicabilitySkill
- rbvm-asset-evidenceSkill
- rbvm-guarded-remediationSkill
- rbvm-priority-decisionsSkill
- rbvm-risk-exceptionsSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 28
If organic entries fall after a template/release change, inspect discovery before rewriting offers.
- Very specific: line 54
Do not use noindex or robots blocking to force canonical selection.
≈ “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.
Files
Marketing Funnel Diagnosis
A funnel rate is conditional on entry rules, identity, event order, and observation time. Changing traffic composition can move aggregate rates while segment performance stays constant. Diagnose measurement before changing campaigns or pages.
Lock the contract
Record counting unit, entry event, ordered milestones, intervening-event rules, deduplication, conversion window, timezone, exclusions, and channel assignment. Each numerator must come from its denominator's cohort. Use equally matured entry cohorts; recent entrants have less opportunity to convert.
Diagnosis sequence
- Reconcile event and business totals. Check releases, consent changes, duplicate events, bot filters, identity merges, missing channels, and reporting lag. Preserve unknown attribution rather than silently discarding it.
- Show counts beside rates: entries, step completions, absolute drop-offs, step conversion, final conversion, and time-to-convert. A denominator surge can lower a rate while completions rise.
- Compare equivalent cohorts within channel, device, geography, or buyer type. Prespecify plausible segments; report small counts and uncertainty.
- Separate mix from within-segment change. Use fixed, non-overlapping segments and old denominator shares for the analyzed rate:
standardized_new = sum(old_share * new_rate). Compare this with both observed aggregate rates. This is descriptive standardization, not a causal counterfactual. - Investigate the localized problem using error logs, customer feedback, and relevant usability evidence. Return competing explanations, counterevidence, and the smallest discriminating check. Confirm business impact through purchases, qualified pipeline, or retention.
Worked mix example
Channel A converts at 10%; B at 2%. An 80:20 traffic split yields 8.4% overall; 20:80 yields 3.6%. Reweighting the unchanged rates by the old mix returns 8.4%: no within-channel deterioration.
Organic-entry branch: canonical evidence
If organic entries fall after a template/release change, inspect discovery before rewriting offers. Canonical declarations express a preferred duplicate/content-superset URL, not a redirect or proof of engine selection.
- Sample affected and unaffected URLs across templates and search engines. Record final URL/status, response-header canonical, original HTML head, rendered head, crawl/index directives, sitemap/internal-link targets, and capture time. Check whether distinct content accidentally shares one canonical; legitimate duplicates can share one. Derive intended mappings from content, not URL shape alone.
- Compare signals: canonical tags should not contradict headers, rendered HTML, redirects, or sitemaps. Verify target content and availability; preserve unavailable evidence as unknown. Do not make every URL self-canonical: genuine duplicates need consolidation.
- In authorized Search Console indexed reports, compare user-declared versus Google-selected canonical and last-crawl time. A live test cannot predict selected canonical. A pre-fix crawl cannot establish post-fix recovery.
- Propose the smallest template correction, preserving duplicate mappings; test representative outputs. Report separately: configuration corrected, post-fix indexed selection observed, acquisition recovered. Recrawl requests guarantee neither indexing nor recovery time; campaign/production changes need authorization.
Example: two distinct product pages both declare the homepage. A naive “canonical exists” check passes; an intended-mapping check flags both. Correct output removes that defect but does not prove Google adopted it or sales improved. Run the original offline snapshot scenarios from the repository root:
python3 skills/marketing/marketing-funnel-diagnosis/reproduce.pySources and scope
- Amplitude: funnel interpretation: ordering, windows, entry cohorts, and drop-offs; inspect your own tool's counting semantics.
- Microsoft: metric interpretation pitfalls: ratio decomposition and telemetry bias; observational funnels still cannot establish causation.
Canonical evidence references, provenance, and test limits: SOURCES.md.
Pitfalls
- Joining monthly signups to monthly purchases mixes cohorts and conversion lags.
- The largest percentage drop may have little recoverable business value.
- Averaging segment percentages without denominator weights distorts totals.
- Optimizing a local step can attract worse-fit buyers or shift failures downstream.
- Do not use
noindexor robots blocking to force canonical selection; these change indexing/crawling.
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
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.

