Get cohort LTV
Person-grain cohort lifetime value. Each cohort carries one row per elapsed period with the revenue booked in that period, the revenue booked cumulatively, the cumulative revenue per acquired person, and how many of the cohort had paid by then. The parameters are identical to /retention on purpose: the same range and granularity select the same cohorts on both endpoints.
cumulative_ltv divides by every person ACQUIRED, payers and non-payers alike, so it compares directly against a blended acquisition cost. Revenue per payer is cumulative_revenue divided by paying_persons.
Realized revenue only. Nothing is extrapolated, and a cohort returns only the periods it has actually entered — a cohort acquired yesterday returns one or two rows, never a flat twelve-period curve. Each period carries period_complete, which is false for the in-flight period at the end of a young cohort; compare cohorts on complete periods only.
by=channel splits every cohort by the FIRST-TOUCH channel that acquired each person — the same nine buckets /attribution reports (meta, google, tiktok, snapchat, email, referral, organic, direct, other). One channel per person; credit is never shared across touches. The curves move under a channels array and the cohort’s top-level periods array is absent, so a total row and a channel row can never be added together by mistake.
Persons whose acquisition is not mapped are returned in an unattributed bucket and are never dropped from the cohort. The person-to-channel mapping reaches about 365 days back from each person’s first activity and is maintained going forward, so cohorts older than that read heavily unattributed. That is unknown acquisition, not a channel without customers.
Campaign-level and ad-level LTV do not exist. The breakdown is channel only.
Authorizations
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Query Parameters
Inclusive first cohort date, as YYYY-MM-DD or an ISO 8601 instant. Supply with end_date; omit both for the last 90 days of cohorts.
Inclusive last cohort date, as YYYY-MM-DD or an ISO 8601 instant.
Cohort bucket and period unit. Weeks start Monday.
day, week, month Highest period index returned, counted in granularity units from period 0.
1 <= x <= 24total returns one curve per cohort. channel returns one curve per (cohort, first-touch acquisition channel), plus an unattributed bucket. Any other value is a 400 rather than a silent fall back to the total.
total, channel Response
Successful response. Realized revenue per acquired person: nothing is predicted, and only periods a cohort has entered are returned.