Leading vs lagging engagement indicators

Engagement measurement traditionally relies on lagging indicators — the eNPS distribution, the UWES composite, regrettable attrition rate, and exit-interview themes. Each is real, defensible, and slow. By the time the number changes, the experience producing it has been running for months.

Leading indicators describe the conditions upstream of those numbers. In the Frustration Condition framework, the leading indicators are: the count of statements per Architecture cluster, the median age of an unresolved cluster, the share of clusters that have received a closed Three Doors decision, and the recurrence rate of the same Architecture across consecutive cycles.

The pairing matters because operators tend to ask "What changed?" of lagging indicators (which produces recency bias) and "What is changing?" of leading indicators (which produces structural insight). A balanced engagement read uses both: the qualitative leading read for next-quarter decisions, and the quantitative lagging read for governance and benchmarking.