A cohort starts with a rule that can be applied before you know the outcome. Opening-cohort retention follows employees present on the first reporting date. Early-tenure attrition follows a hiring cohort for a fixed window. Recent hires have not completed that window, so counting them as retained makes the rate look better without adding evidence. Keep the observation cutoff in the query. Rehires and transfers also need a stated rule: this course treats one employment episode as one record, and a transfer is not a company exit. Compare rates only after their event, eligible population and observation window agree. A difference in definitions is enough to explain a difference in reported rates.
Group the hire event by month after applying the employee and date filters.
Group the hire event by month after applying the employee and date filters.