Analytics Is a Mess — Embracing the Exploratory Phase

Benn Stancil argues that the messy, exploratory phase of analytics is not a bug to be fixed but a natural and necessary part of the process. The illusion of linear dashboard development leads teams astray.

Core mental model

There is no capital-T truth in metrics. “There is no correct win rate waiting to be unearthed; one version isn’t true while another is false.” Every metric definition is tautological — it measures exactly what it says, no more, no less. The real work is choosing which definition serves the decision.

The messy middle

When we look at companies with mature data practices, we only see the final, stable metrics and dashboards. This creates a survivorship bias: we assume dashboard development should be linear (blueprint, foundation, frame, finish). In reality, the first steps are uneven and uncomfortable — and that is normal.

New tools don’t fix messiness; they arrive right as the exploratory process is consolidating. The cleanup was happening anyway. The tool just gets the credit.

Practical implication

Rather than trying to stifle the exploratory phase, plan for how to tidy it up later. Work in sandboxes. Reserve space for polished final drafts. Separate exploratory analysis from production analytics.

Connects to ghosts in the data, data’s big whiff, creativity faucet (the messy phase IS the wastewater), analytics as craft.

Open questions

• How do you communicate to stakeholders that the “messy phase” is valuable, not wasteful?
• What’s the right ratio of sandbox work to production work for a data team?