Why it matters
AOV is easy to compute from storefront or data warehouse order tables. Rising AOV from bundles, upsells, or free-shipping thresholds can improve return on ad spend (ROAS) on first purchase without improving lifetime profit. Heavy discounting, bracketing (buy multiple sizes to return most), and one-time bulk buys inflate AOV while hurting net revenue after returns.
Ad platforms often learn on first-order purchase value. If campaigns optimize toward high-AOV first orders, you may acquire customers who never repurchase or who return product after the platform already credited revenue. That is why growth teams pair AOV with repurchase rate, refund rate, and customer lifetime value (LTV) before trusting merchandising or media wins.
For performance marketing, AOV explains part of CPA and ROAS movement but does not replace predicted lifetime value (pLTV) when value realization is delayed.
Average order value (AOV)
AOV is frequently the default value sent to ad platforms before pLTV maturity:
- First-party data in your data warehouse supplies order lines, discounts, returns, and customer IDs.
- Many teams start with purchase value equal to first-order revenue (AOV at user level on the conversion event).
- Churney models pLTV to forecast value beyond first order, then sends calibrated scores directly to ad networks via Meta CAPI and the Google Ads Conversion API.
- Value-based bidding shifts optimization from "who places a large first basket" to "who is worth more over the prediction horizon."
- Validate with calibration, cohort maturity, and incremental ROAS vs AOV-only business as usual (BAU).
AOV is an input and a short-window signal. pLTV is the activation-grade value object when repeat and returns matter.
Gross AOV:
AOV (gross) = Gross order revenue / Order countNet AOV (preferred for signal design when returns matter):
AOV (net) = (Gross order revenue - Returns - Discounts) / Order countGuardrails:
Exclude or tag gift orders, wholesale, and test transactions per your policy.
Segment new vs repeat customers; blended AOV hides mix effects.
Pair AOV trends with repurchase rate and refund rate before changing value sent to ad platforms.
Category variants
| Model | How AOV shows up |
|---|---|
| Ecommerce / DTC | Mean gross order value; net AOV after returns and discounts is closer to signal truth for value models. |
| Subscription app | First invoice or plan price as "order"; upsells and annual plans skew AOV vs monthly LTV. |
| SaaS / PLG | Contract or first-payment value as order analog; expansion revenue lives outside first "order" AOV. |
Common mistakes
- Using gross AOV when net revenue is the economic truth. Returns and promos make gross AOV look strong while margin collapses.
- Treating AOV as LTV. One order average says little about repeat, churn, or subscription length.
- Sending static AOV on every conversion to ad platforms. Flat values remove the benefit of value-based bidding.
Advertiser lens
| Role | What they ask | What good looks like |
|---|---|---|
| Head of Performance / UA | Did this campaign bring high-AOV buyers? | AOV segmented by campaign and new vs repeat; paired with CPA and ROAS. |
| VP Growth / CMO | Should we push bundles to raise AOV? | Merchandising tests with net revenue and repeat rate, not gross AOV alone. |
| Marketing Analytics / Data Science | Is AOV a good value proxy for bidding? | Correlation of first-order AOV to D90 LTV; case for pLTV when correlation is weak. |
| Data Engineering | Can we compute net AOV reliably? | Return and discount tables joined to orders in the data warehouse; consistent order definition. |
| Finance / Procurement | Does higher AOV improve payback? | Contribution margin per order and cohort payback, not headline AOV. |
FAQ
What is average order value (AOV)?
AOV is total order revenue divided by number of orders in the period. It describes average basket size in currency terms.
How do you calculate AOV?
AOV equals total revenue from orders divided by order count. Use a consistent definition of "order" (gross vs net, shipping included or excluded) and document it.
What is the difference between AOV and LTV?
AOV is per order. Customer lifetime value (LTV) is total value across all orders and time with the customer. LTV incorporates repeat, churn, and refunds; AOV alone does not.
Is AOV a good signal for ad platform optimization?
It is a reasonable starting point when repeat is rare and returns are low. When repeat, subscriptions, or refunds matter, pLTV usually outperforms AOV as a value signal.
Why can ROAS rise while AOV falls?
More conversions at lower basket size can increase attributed revenue faster than AOV declines, or mix may shift toward campaigns with higher volume and lower AOV.
Should platforms receive gross or net AOV?
Net revenue (after returns and discounts) is closer to economic truth for value modeling. Many teams start gross for operational simplicity, then migrate as calibration exposes gap.
How does AOV relate to pLTV?
First-order AOV is often a feature in pLTV models. pLTV predicts total expected value; AOV is one observable at the anchor purchase event.
Not the same as
| Term | Difference |
|---|---|
| Customer lifetime value (LTV) | Multi-order, multi-period value; AOV is single-order average. |
| Predicted lifetime value (pLTV) | Forward-looking modeled value per user; AOV is historical mean per order. |
| Conversion value (platform) | Value parameter on an ad event; may equal AOV on purchase but includes modeling choices. |
| Contribution margin | Profit after variable costs; AOV is revenue, not margin. |
