Fictional Portfolio Project · Marketing Analytics & CRO
REAL IMPLEMENTATION & VERIFIED TESTING: Live WordPress landing page, PulseFit-only GTM configuration, GA4 behavioral events, exactly-once QA, test UTM attribution, and a public interactive dashboard.
SYNTHETIC / HYPOTHETICAL ANALYTICAL DATA: The 3,600-session dataset and every commerce, funnel, channel, segment, and conversion metric are modeled demonstrations—not real users, customers, purchases, revenue, or campaign performance.
PLANNED / NOT IMPLEMENTED EXPERIMENTATION: PF-EXP-001 is a design only. No test was launched and no lift, significance, result, or winner is claimed.
Challenge
Build an end-to-end analytics and CRO demonstration for a fictional portable fitness product while maintaining a strict boundary between genuine implementation evidence and modeled business analysis.
Measurement Strategy
The architecture measures genuine behavior: the primary CTA and resistance-level selection. Commerce events were deliberately excluded because the live page has no cart, checkout, or purchase flow.
Tracking Implementation
GA4 delivery uses isolated PulseFit-only GTM configuration. The initialization sets send_page_view=false, preserving the existing Site Kit page view. Controlled Preview and production QA verified exactly one primary_cta_click and one select_item, expected non-PII parameters, no commerce payloads, and no unrelated Project 01–04 firing.
Synthetic Dataset
A deterministic 3,600-session synthetic dataset supports repeatable analysis by source, device, user type, campaign, engagement, CTA, and modeled funnel stage. Every analytical output retains an explicit synthetic-data disclosure.
Dashboard & Funnel Analysis
The self-hosted interactive dashboard recalculates KPIs, acquisition views, behavioral signals, funnel stages, and segments from the synthetic records across six sections: Executive Overview, Acquisition, User Behavior, Conversion Funnel, Segmentation, and CRO Insights.
CRO Diagnosis
The modeled funnel highlights the transition from offer engagement to the primary CTA as the largest diagnostic loss. This is a hypothetical cue for investigation—not observed business performance.
Hypothesis Prioritization
An ICE backlog separates implemented safeguards, research needs, and possible experiments. Priorities emphasize honest disclosure, measurement integrity, decision clarity, and reduced uncertainty near the CTA.
Experiment Design
PF-EXP-001 proposes moving kit contents, resistance range, and risk-reduction information closer to the hero CTA. Its specification defines rationale, audience, measures, guardrails, QA, and stopping rules; it remains planned and unlaunched.
Verification
Verification covered GTM Preview, GA4 DebugView/Realtime, exactly-once isolation, event parameters, one Site Kit page view, test-specific UTM attribution, zero commerce events, zero PII, and post-publication behavior.
View Live Implementation · View Implementation Evidence
Takeaways
This project demonstrates practical analytics governance: measure only genuine functionality, isolate production changes, distinguish QA traffic from real traffic, label synthetic analysis persistently, and design experiments without presenting plans as results.