A researcher wants to estimate the effect that a marketing intervention had on customer purchases while accounting for time-ordered observations and the timing of events until a customer churns. Which temporal modeling approaches are being combined?