Overview
A Hex metric provenance map helps marketers answer a deceptively simple question: where did this KPI come from? This playbook traces one important number through Hex projects, notebook cells, queries, data connections, and recent runs so the team can cite it without crossing their fingers.
Use it when a metric appears in a campaign report, board deck, growth review, or pipeline readout and nobody is fully sure which notebook is the source of truth. The output is a tracker plus a short narrative summary that explains the metric logic in plain business language.
Why you should trust the metric before you quote it
Marketing decisions often move faster than analytics cleanup. A KPI can be copied across notebooks, renamed in a chart, filtered differently for one campaign, or left tied to a stale data connection. That does not make the number useless, but it does make provenance worth checking before the metric becomes a headline.
Hex is built for collaborative analytics notebooks and apps, which means business teams can use the same workspace for exploration and reporting. Hex describes projects as places to combine code, SQL, markdown, visualizations, and apps in one workflow in its project documentation. That flexibility is useful, but it also means the analysis layer needs an occasional map.
Run this playbook when a KPI is important enough to quote externally, use in a forecast, or defend in a leadership meeting. The result is not a technical audit for its own sake. It is a practical answer to which definition is current, which data feeds it, and what needs fixing before the team treats it as canonical.
Step-by-step
- 1Confirm the KPI, the reporting context, the audience, and any known Hex projects or owners tied to the number.
- 2Search Hex for projects, notebook labels, queries, markdown notes, and dashboard text that reference the KPI or close variants.
- 3Inspect the likely source projects to map the metric definition, upstream data connections, filters, date windows, joins, exclusions, and calculations.
- 4Check recent project runs, ownership cues, and connection status so the tracker separates metric logic from freshness or governance risk.
- 5Compare competing definitions side by side when multiple notebooks calculate the KPI differently.
- 6Produce a provenance tracker and a short summary that names the likely source of truth, confidence level, open questions, and recommended cleanup.
Frequently asked questions
What kind of metric is this best for?
It works best for KPIs that influence marketing decisions: pipeline sourced, conversion rate, paid media efficiency, activation, retention, lead quality, or campaign revenue.
Does this replace a data team review?
No. It gives marketing and analytics teams a clean starting point. If the playbook finds conflicting definitions or unclear ownership, a data owner should confirm the canonical version.
What if the KPI appears in several Hex projects?
The playbook compares each definition instead of picking one too early. It will recommend the strongest source of truth when the evidence is clear and flag the decision when it is not.
How often should we run it?
Run it before major readouts, planning cycles, board reporting, or whenever a heavily cited metric starts showing different values across notebooks or dashboards.
