Why functional KPIs fail
Functional KPIs are necessary but insufficient. Marketing chases MQL volume; sales chases pipeline coverage; customer success chases satisfaction scores. None of the three is accountable for the system-level outcomes the firm actually needs — revenue predictability, profitability, customer lifetime value.
The structural failure mode: each function optimises against its own metrics, the metrics are individually well-met, and the system as a whole underperforms. This is the predictable consequence of measuring the parts without measuring the whole.
What system-level outcomes look like
The four system-level outcomes specified in the empirical research are revenue growth, profitability, productivity, and customer experience. These cut across the functions; they cannot be optimised by any single function in isolation.
System-level outcomes have specific properties that make them suitable for executive accountability. They are observable from the firm's perspective rather than from inside any function. They are lagged (producing them requires multi-quarter investment). They are jointly produced (no single function has full control over any of them).
Specific executive actions
Three actions implement the shift. First, mandate a shared revenue performance framework co-owned by the heads of marketing, sales, and customer success. Second, tie executive compensation explicitly to system-level outcomes alongside (not instead of) functional KPIs — approximately 30-50% of variable comp should be system-level.
Third, hold cross-functional quarterly business reviews on system-level performance, not on functional reports stitched together. The QBR should answer: how did the system perform this quarter, what produced the result, and what changes are needed across functions?