Definition
The net economic value of AI-triggered retention interventions after accounting for intervention costs, false positive costs, and the baseline departure rate of non-intervened employees.
The Measurement Gap
Organizations invest in attrition prediction models and then deploy retention interventions (stay conversations, compensation adjustments, project reassignments, development opportunities) targeting employees flagged as flight risks. The implicit assumption is that these interventions produce positive ROI: the cost of intervening is less than the cost of the departures prevented. But this assumption is almost never tested because no standard HR metric isolates the incremental value of AI-triggered interventions from the baseline retention that would have occurred without them.
Turnover rate measures overall departures but cannot attribute changes to specific interventions. Retention rate among intervention targets measures a biased sample. These employees were selected because the model predicted they would leave, so their retention could reflect model error (they were never going to leave) rather than intervention effectiveness. Cost-per-hire and replacement cost metrics measure what turnover costs but not whether interventions prevented it. The result is that organizations can spend millions on retention programs triggered by AI predictions without any evidence that the AI-triggered targeting produced better outcomes than random intervention or no intervention at all.
The false positive dimension is particularly costly and invisible. Every retention intervention directed at an employee who was never going to leave (false positive) consumes budget and management attention without generating any retention value. Without measuring false positive costs alongside avoided turnover, RIVI can appear positive even when the intervention program is a net loss.
Decision Supported:
Are our AI-triggered retention interventions generating positive economic returns after accounting for false positives and intervention costs?
Formula
Plain language: Calculate the total avoided turnover cost from employees who were flagged, received interventions, and stayed. Subtract the total cost of all interventions and the cost of false positive interventions. Divide by total intervention cost.
Notation:
RIVI = (Avoided_turnover_cost − Total_intervention_cost − False_positive_cost) / Total_intervention_cost
Variables:
| Variable | Definition | Unit |
|---|---|---|
| Avoided_turnover_cost | Estimated replacement cost for flagged employees who received interventions and remained, adjusted by the probability they would have left without intervention | Currency |
| Total_intervention_cost | Direct costs of all retention interventions deployed (compensation adjustments, bonuses, development programs, management time) | Currency |
| False_positive_cost | Intervention costs spent on employees who were flagged but were never at genuine risk of departure (determined by post-hoc analysis or control group comparison) | Currency |
Proposed formula, pending validation pilots.