We tracked customers who reached their 3rd visit by specific dates, then measured what percentage went on to make 4th, 5th, and 6th visits within the remaining days of December. This approach controls for "time available" and eliminates selection bias.