The subscription LTV math nobody calculates correctly

Most stores measure subscription LTV wrong. Here’s the model we use across every Recharge build.

Most subscription LTV calculations are wrong by a factor of 1.5–3×. We see it on every account we audit. The error is almost always in churn modeling.

The bad way

Average revenue per subscriber × average tenure. Tenure is computed from current active subs, which biases low because long-tenure customers are over-weighted. The model is also blind to retention cohort changes.

The right way

Cohort-based survival modeling. For each acquisition cohort, compute month-over-month retention. Sum the retention curve. That’s your expected order count. Multiply by AOV.

It’s not hard. It just requires you to stop pulling the single-number reports and start working with cohorts.

If your subscription LTV doesn’t change when you change acquisition channel mix, you’re not modeling LTV.

What this changes

Once we model LTV correctly, almost every client we audit revises their target CAC. Most discover they’ve been overpaying for paid social by 30–60%. A few discover the opposite — they’ve been starving channels with strong long-tail retention.