Retention ROI and agency-labor savings: the calculator methodology
Retention ROI equals avoided separations at the operator’s own fully loaded cost, plus avoided agency premium valued at the invoiced rate differential, plus the 0.5 premium portion of avoided overtime, minus platform cost and internal effort cost — measured against a frozen baseline and attributed only where a driver moved and an action was logged.
The formulas
- Avoided-separation value: AS = (baseline separations − measured separations) × fully loaded cost per separation, counted only for the confirmed-and-actioned cohort.
- Avoided agency premium: AA = Σ (baseline invoiced agency hours − measured invoiced agency hours) × (effective agency rate − internal loaded rate).
- Avoided overtime premium: AO = (baseline OT hours − measured OT hours) × 0.5 × loaded base rate; only the premium portion counts.
- Census-protection value: reported as a separate labelled scenario line, only against a documented and dated admission hold.
- Net program value: (AS + AA + AO) − platform cost − internal effort cost, where internal effort includes ED, scheduler, HR, and finance loaded hours.
Rules that keep the number defensible
- Freeze the baseline in writing before the first intervention and never restate it.
- Align baseline and measured periods to the same calendar quarters or state the seasonal adjustment.
- Exclude acquisitions, dispositions, and communities in lease-up from both periods.
- Attribute only where a driver measurably moved and an action was logged; report the rest as unattributed.
- Count agency savings as rate differential on avoided hours, never as gross invoice reduction.
- Count overtime savings as the 0.5 premium portion only.
- Charge internal effort and full undiscounted platform cost against the result.
- Publish low, base, and high sensitivity plus a zero-attribution floor case.
- Disclose every confounder: wage actions, market events, leadership change, rate-card change, acquisition.
- Tag every figure by evidence class: A public primary source, B vendor public materials, C operator-reported, D design target not yet operator-proven.
Input specification
Inputs, sources of truth, and rules| Input | Source of truth | Rule |
|---|
| Baseline separations | HRIS separation records | Frozen period, voluntary only, transfers excluded, seasonally aligned |
|---|
| Fully loaded cost per separation | Operator payroll, agency invoices, recruiting spend, onboarding hours | Operator cost lines only; published ranges used to sanity-check, never as the input |
|---|
| Invoiced agency hours and effective rate | Agency invoices reconciled to shifts | Includes guarantees, cancellations, travel, and contract overtime; rate cards not accepted |
|---|
| Overtime hours | Payroll | Only the 0.5 premium portion is savings-eligible |
|---|
| Internal effort cost | Operator estimate of loaded hours | ED, scheduler, HR, and finance time is always charged |
|---|
Frequently asked questions
- How do you calculate retention ROI for a senior care portfolio?
- Sum avoided separations at the operator’s own fully loaded cost, avoided agency premium at the invoiced rate differential, and the 0.5 premium portion of avoided overtime against a frozen baseline, then subtract platform cost and internal effort cost. Attribute only the confirmed-and-actioned cohort where the underlying driver measurably moved.
- How are agency-labor savings calculated without overstating them?
- Savings equal avoided invoiced agency hours multiplied by the effective agency rate minus the internal loaded rate for the same role and shift. The effective rate must include guarantees, cancellations, travel, and overtime on contract hours, taken from trailing invoices rather than the rate card. Hours the operator would have paid internally are not savings.
- Does SeniorCRE publish a payback period or savings figure?
- No. SeniorCRE publishes the method, the formulas, and the guardrails. No completed operator-production pilot is published as of August 7, 2026, and buyers get a back-test on their own history with low, base, and high cases instead.
https://seniorcre.com/workforce-intelligence/retention-roi-methodology