// Archived case / Question discovery
When the homepage is not the front door.
A GA4 discovery and rebuild for a content brand that surfaced six tracking gaps, challenged a homepage-first editorial assumption, and shipped with training so the team could keep using it.
What happened
The site stopped treating the homepage as the default story.
A high-profile content brand's GA installation went from traffic numbers without answers to a GA4 stack that surfaced six tracking gaps the client did not know existed. Once the data was collected cleanly, the editorial premise reframed: podcast episode pages, not the homepage, were where the most engaged audience actually lived.
The implementation shipped with a Loom training library so the team could operate it without the analyst. Adoption signal: the agency returned with substantive usage questions in October 2022, and came back unprompted two years later in April 2024 with a GTM issue.
Situation
GA was installed, but the business questions were still unanswered.
The content brand had podcasts, newsletter, events, and books. GA was installed, but the team could not reliably answer whether people were listening, what drove newsletter signups, or which pages were genuinely performing. The implicit editorial model treated the homepage as the primary entry point.
Task
Move from data without answers to instrumentation mapped to real editorial goals.
The work had two layers: surface the technical gaps where the data was not being collected cleanly, then rebuild around actual editorial goals: podcast engagement, newsletter conversion, content-level dimensions, and a small-team training model.
DiscoverQuestions
Documented what the data was and was not doing, and what decisions it should inform.
AuditGaps
Compared the implementation to the business questions and surfaced six tracking gaps.
AdoptTraining
Rebuilt the measurement layer and shipped Loom training so the team could operate it.
Documented Q&A with agency and client: what the data was doing, what it was not doing, and what decisions it should inform.
Six gaps surfaced: double-firing pageviews, broken UTMs, no conversion tracking, no podcast engagement measurement, misaligned event taxonomy, and incorrect iframe player events.
GA4 implementation rebuilt around scroll depth, episode engagement, two-minute engaged-session conversion, newsletter signups, content dimensions, and fixed source/medium.
Loom training covered custom dimensions, conversions, explorations, entrances/exits, and real-time reporting.
Action
The rebuild was designed around editorial questions.
I facilitated structured discovery with the agency and client, audited the existing installation against the questions the business actually needed to answer, rebuilt GA4 around podcast and newsletter goals, and created a Loom training library so the team could continue using the implementation without me sitting beside them.
Outcome
The audience model shifted toward podcast episode pages.
- Six tracking gaps surfaced, including double-firing pageviews, broken UTMs, missing conversion tracking, and unmeasured podcast engagement.
- Podcast episode pages reframed the audience model: they were where the most engaged users entered, stayed longer, and took more actions.
- Loom training library shipped so the agency and client could use the new implementation without the analyst.
- October 2022 follow-up produced substantive usage questions, not service questions.
- April 2024 unprompted return with a GTM issue showed the implementation was still in use two years later.
Design lesson
Adoption is the deliverable.
A dashboard can be technically correct and still fail if the team cannot use it. The useful signal here was long-tail adoption: the same agency returned years later because the implementation was still operational, not shelved.