Time to second order (T2) is the elapsed time between a customer’s first and second purchase, read at the cohort level as a median or distribution shape rather than as an aggregate average across mixed-tenure customers. The right read is the median (or the full distribution) of a fixed acquisition cohort — every customer who placed a first order in, say, March 2026 — measured against how long each took to come back. An aggregate average across all customers in a window mixes vintages and is skewed by a survivorship effect: cohorts old enough to have produced a second order at all weigh more in the average than the younger ones still waiting. The cohort form is the operator-useful read; see cohort analysis for the broader discipline.
T2 matters because it is readable months before LTV stabilizes. A 12- or 24-month LTV curve needs 12 or 24 months of customer aging; a cohort’s median T2 is visible inside 60 to 90 days. Two cohorts with identical first-month revenue per acquired customer but a median T2 of 45 days versus 180 days produce very different downstream economics, and the gap shows up early enough to change next quarter’s acquisition plan rather than next year’s.
For many DTC brands in replenishment-cadence categories — skincare, supplements, pet food, coffee — the T2 distribution is often bimodal. A fast-repeaters cluster lands inside roughly one consumption cycle (30 to 60 days for a 30-day-supply product), and a long tail converts weeks or months later, typically on lifecycle email or paid retargeting. The shape diagnoses which retention lever is doing the work.
Three shapes the T2 distribution takes
-
A heavy fast cluster
means product fit is pulling customers back on its own. -
A heavy long tail
means lifecycle marketing or retargeting is recovering customers the product alone did not. -
Both clusters thin
means either the product or the acquisition targeting is the binding constraint, and lifecycle work is layered on top of a leaky base.
Category sets the structural floor. Consumables target a median T2 inside one consumption cycle; considered or durable goods (furniture, mattresses, appliances) have structurally longer T2 that is set by the purchase cycle, not by retention quality. Benchmark T2 against the category baseline, not against an absolute number.
The same 90-day repeat purchase rate can come out of very different T2 distributions, and the shapes imply different operator moves. T2 is what anchors the retention conversation in the months before LTV can resolve.