January 31, 2026 · 6 min read
Marketing Data Governance, the Project Nobody Wanted to Own Until It Broke Something
Everyone agreed data governance mattered in theory. Nobody would own it in practice until a GDPR data subject request took four days to fulfill and should have taken four minutes.
For about eight months, 'we should really sort out data governance' was a sentence said in nearly every quarterly planning meeting at a client I worked with, followed immediately by nobody being assigned to do it. It wasn't that anyone disagreed it mattered. It's that governance work has no obvious owner in most marketing orgs — it's not a campaign, it's not a dashboard, it doesn't move a KPI anyone gets measured on, so it always lost to whatever did.
What finally forced the issue
A customer submitted a GDPR data subject access request — asking the company to produce every piece of personal data it held on them. It took four business days to answer, involving manual searches across a CRM, a marketing automation platform, a data warehouse, two analytics tools, and a customer support ticketing system, because nobody had a map of where personal data actually lived. The legal team had assumed this would be a quick lookup. It turned into a multi-day fire drill with three people pulled off their actual work.
Governance work loses every prioritization fight until the day it's the only thing standing between you and a compliance deadline, and by then it's already too late to do calmly.
What ownership actually looked like once assigned
The fix wasn't a grand governance platform rollout. It started with one person — a senior analytics engineer who already understood the data landscape best — getting explicit, protected time and a mandate to own exactly three things, no more, with everything else deferred to a second phase that may or may not ever happen.
- A single data inventory mapping every system that holds customer PII, kept current as a living document, not a one-time audit.
- A documented, repeatable process for fulfilling a data subject request that any team member could execute in under an hour.
- Clear data retention rules per system, enforced by actual deletion jobs, not just a policy doc nobody automated against.
- Explicitly deferred: broader governance ambitions like a full data catalog or lineage tooling, which were real goals but not the fire that was currently burning.
The next data subject request, three months later, took under two hours. Governance work rarely gets funded proactively because its payoff is invisible until the day it isn't — nobody throws a party for the incident that didn't happen. The lesson that stuck: assign a narrow, concrete owner before the compliance deadline forces you to assign a panicked one.