Reading Your Exits as a Selection Signal: Operator Insight
Companion to Part IV. The installment shows the reform that worked and got absorbed, and the people you lose because of it. This is the instrument that finds them before they go.
Structural Recap
By month 12, a VP of Engineering at Northstar corrects the structural variable inside her own span of control. She changes what gets rewarded, protects the early messenger, and holds a real standard on commitments. Within a quarter her org is measurably better, escalation latency inside her teams falls, and her people surface problems sooner. She proves the additive thesis with her own hands: good intention attached to an actual change in reinforcement, produces the result the intention alone can never deliver.
Then the surrounding company, which did not change, reabsorbs it. Not by decision but by gravity, eroding her correction at every edge where her org touches the uncorrected system, until she is working twice as hard as her peers to hold a line the structure is actively dissolving. One fragment survives for reasons that are partly person-dependent and do not map cleanly to the framework, which is the honest residue rather than a clean win. The rest erodes.
What compounds here is selection. When an organization rewards adaptation over the standard long enough, it does not only shape behavior, it decides who stays. The people with the most options leave first, then the ones with the strongest conviction, then the uncredited holders of the informal standard. What remains is a team filtered for adaptation and low friction, and the next leader who tries to reform the company inherits a team the last failure helped select.
Distortion Model (Causal Map)
TRIGGER: A strong leader corrects reinforcement inside her own span. It works locally. The surrounding incentive geometry remains unchanged.
↓
BEHAVIORAL SHIFT: The correction erodes at every boundary with the uncorrected system. The leader absorbs the gap personally, working harder to hold a standard which the structure actively dissolves. Effort rises; visible result stays flat.
↓
REINFORCEMENT: Leadership praises the person for fighting the current, which substitutes for fixing the current. The structural cost of holding the standard stays on the individual. Staying costs more than leaving.
↓
METRIC MOVEMENT: The standard-holder exits. Departures run in order: most options first (most talented), then strongest conviction, then the uncredited people holding informal standards. Board metrics still show growth.
↓
ENTRENCHMENT: Selection encodes the lower standard into the architecture. The team that remains is calibrated to the gap between declared and enforced standards. The capacity to fix the company has walked out the door, and unlike earlier stages, it does not come back when the fix arrives.
The causal root is structural cost landing on an individual instead of on the system. The leader holds the standard at personal expense because nothing structural shares the load, and selection turns that unshared cost into permanent architecture. The instrument moves the cost off the person and reads the exits before they finish telling the story.
The recognition above is the public half. What follows is the instrument: how to find the leader holding a standard against the current before they leave, how to read your exits as a selection signal rather than isolated attrition, and the 30-day sequence to intervene while the team that could fix it is still there.
