Average CAC is the aggregate number — the single figure that combines all your channels, segments, and product lines into one. It is useful as a high-level health check, dangerous as the only number you act on. Most of the interesting decisions live in the segmented view, not the average.
How to calculate average CAC
Average CAC = Total acquisition spend (period) ÷ Total new customers (same period)
The simplicity is the appeal — and the limitation. Average CAC tells you what acquisition cost in total, but it does not tell you which channels are pulling it up, which segments are pulling it down, or whether the customers acquired this period are the same kind of customer who drove your historical LTV.
Why average CAC is misleading on its own
Three scenarios where the average lies:
- Channel mix shift. If organic referrals (low CAC) shrink and paid social (high CAC) grows, average CAC rises even if no individual channel got more expensive.
- Segment shift. If you start acquiring more low-LTV customers, average CAC can drop while LTV:CAC actually worsens.
- Spend timing. Spend in March acquires customers in March, April, and May. A single-month average CAC distorts in either direction depending on cycle.
How to use average CAC well
- Pair it with channel-level CAC. The average is the summary; the channels are the story.
- Pair it with LTV. Average CAC alone is meaningless without the LTV it is being judged against.
- Track it across cohorts, not periods. Cohort CAC (the spend on customers acquired in March, looking back) is more honest than calendar CAC.
- Watch the trend more than the level. Whether average CAC is going up, flat, or down is more useful than whether it is "good" or "bad" in absolute terms.
What is a good average CAC
Almost entirely category-dependent. Low-AOV impulse subscriptions might target $20–60 CAC. Premium replenishment $80–200. High-AOV curation $150–400. B2B subscriptions can run into the thousands. The number is meaningless without the LTV next to it — see LTV:CAC for the ratio that actually matters.