All insights
AnalyticsOperations5 min read
Why your "data team" is a request queue
If every chart goes through a central team, your data team isn't a team — it's a bottleneck with a Slack channel.
Many companies hire a "data team," and within a year that team is fully occupied serving ad-hoc requests from every other team. The bottleneck is the team itself. Strategic data work — the reason you hired them — never happens because the queue is always full.
Self-service is the only escape
The unit of leverage isn't "one more analyst." It's giving the rest of the organization the tools and the model to answer their own questions. That requires a semantic layer, clean canonical models, and tooling non-analysts can actually use.
What the data team should do
- Build and maintain the canonical models.
- Set standards for how metrics are defined.
- Coach other teams on how to answer their own questions.
- Take on only the analyses that require their actual expertise.
If your data team is busy, they're probably doing the wrong work.