Average Customer Lifetime Value in Ecommerce: Benchmarks
There is no single average customer lifetime value in ecommerce — the honest range runs from under $30 to over $500 per customer depending on category, margin, and purchase frequency. Postscript’s 2026 benchmark across 17,000+ Shopify stores puts SMS subscriber lifetime value between $25 and $553 from the 25th to the 90th percentile — a 22× spread inside one dataset. The useful question is not “what is average” but “what should MY number be.” Here is how to benchmark it properly.
Why the spread is so wide. CLV multiplies four inputs — order value, purchase frequency, relationship length, and margin — and every category weights them differently. A supplement brand with $45 orders and monthly replenishment can out-earn a furniture brand with $900 orders and a five-year repurchase cycle. Margin widens the gap further: two stores with identical revenue-per-customer and a 25-point margin difference are not in the same business. That is why copying a published “average” is useless; your benchmark has to come from your own cohort math. (The full formula with a worked example lives in our customer lifetime value guide.)
Pick the right horizon before you compare anything. CLV is quoted over wildly different windows, which is how apples get compared to orchards. A 90-day CLV is an acquisition metric — it tells you what a new customer returns fast enough to fund ads. A 12-month CLV is the planning metric — solid enough data, near enough horizon to act on. A three-year CLV is a valuation metric, useful for board decks and dangerous for budget decisions when your store is younger than the window. When someone quotes you an “average CLV,” ask the window first; the number is meaningless without it.
Benchmark against structure, not averages. Three comparisons beat any industry table. Your 12-month CLV against your CAC — the ratio should clear roughly 3:1 on margin. Your subscriber CLV against your non-subscriber CLV — run the math on both groups, because subscription customers should be a multiple, not a rounding error; if they are not, the program leaks, and subscription churn math shows where. And your owned-channel customers against paid-only customers — people on your email and SMS lists reliably out-spend those you can only reach by paying an auction again.
Three measurement mistakes that inflate the number. First, counting revenue before returns and refunds — apparel and footwear brands can overstate CLV by double digits this way, since return rates concentrate in exactly the categories with the highest gross figures. Second, skipping the margin adjustment: a customer who generates $400 of revenue at 35% margin is worth $140, and every acquisition decision made against the $400 figure overpays. Third, survivorship in aging cohorts — if you compute “average lifetime” only from customers who stayed, the churned majority disappears from the math and the relationship length stretches flattering. The clean discipline: net revenue, margin-adjusted, full-cohort denominators, fixed horizon. It produces a smaller number than the one in most pitch decks, and it is the only version that will not mislead a budget.
What actually moves the number. In our client work the biggest single-quarter CLV movements come from automation and channel orchestration, not acquisition: for Spoonful of Comfort, owned channels carried 46.1% of total BFCM revenue after we restructured email and SMS together — repeat behavior you can see directly in cohort value. Subscription programs compound it further, converting purchase frequency from a probability into a schedule; that is why subscriptions sit inside our retention stack rather than beside it.
Get your number in two minutes. Put your AOV, purchase frequency, and margin into the free customer lifetime value calculator. You will get a defensible CLV to hold against your CAC and your category — and the baseline every retention program you fund from here should be measured against. Rerun it quarterly; a benchmark you never revisit is a guess with a birthday.
