Examples
Examples
Try the starting example
- Included cohort: employee ID, start date, exit date or none
- A,2026-01-01,2026-01-11 B,2026-01-01,2026-01-21 C,2026-01-01,none D,2026-01-21,none
- Observation cutoff
- 2026-01-31
- Elapsed tenure days to query
- 20
- Kaplan–Meier retention at each observed tenure day (%)
- 10: 75; 20: 37.5; 30: 37.5
- Retention estimate at requested tenure (%)
- 37.5
- First tenure day retention reaches 50% or below
- 20
- Observed exits on/before cutoff
- 2
- Right-censored employees at cutoff
- 2
- Largest observed tenure (days)
- 30
Try another scenario
- Included cohort: employee ID, start date, exit date or none
- A,2026-01-01,2026-01-11 B,2026-01-01,2026-01-21 C,2026-01-01,none D,2026-01-21,none
- Observation cutoff
- 2026-01-31
- Elapsed tenure days to query
- 10
- Kaplan–Meier retention at each observed tenure day (%)
- 10: 75; 20: 37.5; 30: 37.5
- Retention estimate at requested tenure (%)
- 75
- First tenure day retention reaches 50% or below
- 20
- Observed exits on/before cutoff
- 2
- Right-censored employees at cutoff
- 2
- Largest observed tenure (days)
- 30
Calculation method
Calculation method
Kaplan–Meier retention at each observed tenure day (%)
At each distinct tenure time multiply survival by (1−exits/at-risk). Exits precede censor removal at tied times.
Retention estimate at requested tenure (%)
Use the step value through the requested tenure; return undefined beyond the last observed duration.
First tenure day retention reaches 50% or below
First observed time whose step survival is ≤0.5; otherwise not reached.
Observed exits on/before cutoff
An exit is an event only when its date is on/before cutoff.
Right-censored employees at cutoff
No exit or an exit after cutoff is censored at cutoff.
Largest observed tenure (days)
Maximum exit/censor duration measured in elapsed calendar days.
Symbols and units
- a
- Included cohort: employee ID, start date, exit date or none
- b
- Observation cutoff
- c
- Elapsed tenure days to query
Use the units shown. A list keeps the order in which you enter its values. The symbol ^ means “raised to the power of”.
Assumptions & limits
Assumptions & limits
- A pooled right-censored cohort estimate, not a forecast or causal employee assessment. Assumes comparable cohort membership, non-informative censoring and no unrecorded exits; no confidence intervals, covariate adjustments or extrapolation are supplied.
Are these inputs private?
These calculations run in your browser. A shared link contains the values you choose to share. Only questions submitted to the separate AI finder are sent to OpenAI.
Further reading
NIST: Kaplan–Meier product-limit estimator for censored observations ↗
Background reading; this does not imply endorsement or independent review of SumFizz.