Sunrun found $80M in operating leverage — with data, not gut calls.

With 15,000 people spread across field sales, customer experience, operations, finance, and technology, Sunrun’s leadership team had no accurate real-time view of how the organization actually spent its time.

Business intelligence dashboards showed only lagging financial data; performance reviews lacked standardization and reflected managerial perception and bias. When the board pressed on OpEx, the instinct was to rely on gut and rough estimates to make cuts.

As CEO, Powell wanted to approach the problem differently — with evidence, not assumption.

Parable connected to Sunrun’s existing workplace tools — Google Workspace, Salesforce, Slack, calendar systems, and HRIS data — passively, without surveys or workflow disruption.

The resulting data model analyzed behavioral patterns by job code, comparing employees within their own roles for higher fidelity, and applied a second layer of team-level analysis to surface overstaffed functions alongside individual-level signals.

The result was statistically validated: confidence scores of 90–95% across operations teams. Recommendations were built to be conservative and defensible — designed for a CEO who needed to act on them with conviction. Crucially, the model examined leadership first, before frontline staff, ensuring the data lens was equitably applied.