Blended ROAS Hides Rising Acquisition Costs. Compute CAC and Payback Instead
Blended ROAS mixes new customers with repeat buyers who would have purchased anyway. It can hold steady while your acquisition engine slows down underneath it.
What blended ROAS measures
Total revenue includes repeat purchases from customers you won months ago. As your repeat base grows, blended ROAS drifts up on its own, even when every new ad dollar is buying a customer at a higher price. It is a comfort number. It tells you the business as a whole earns more than it spends on ads, which is worth knowing. It does not tell you where the next ad dollar should go.
Why repeat buyers distort it
Picture two months of the same store with the same ad spend and the same revenue per order. In the first month most orders come from new customers. In the second, the repeat base has grown, so more orders arrive from people who came back on their own. Blended ROAS goes up in the second month, because those orders did not need an ad to win them. Nothing about your advertising improved. Your retention did, and the ratio gave your ads the credit.
That is the whole problem. Blended ROAS rewards you for customers you already have, and it keeps rewarding you while the cost of a new one climbs.
A sample client with a respectable ROAS
The sample agency client, Alder & Pike Supply, spent $209.2K across four channels and brought in $470.1K, an overall return of 2.25x. Most dashboards would stop there and call it fine.
Two other lines from the same audit tell a different story:
- Cost per acquisition rose 14%, from $37.22 to $42.27, across 4,949 conversions.
- Click-through fell 27%, from 1.1% to 0.8%.
Divide the revenue by the conversions and each conversion is worth about $95 in revenue. At the earlier $37.22 cost, that same $95 would have returned about 2.55x. At $42.27 it returns 2.25x. The account is still above break-even, and it is also getting more expensive to win each customer, which is what you need to catch before the cushion is gone.
One average, very different channels
The sample marketing export, from the sample client Ridgeway Coffee, shows the same effect inside a single account. It spent $78.3K, brought in $216.8K and logged 2,055 conversions at $38.10 each. Blended, that is 2.77x.
Underneath, the channels range from 0.69x to 6.87x:
| Channel | Return on ad spend |
|---|---|
| Klaviyo Email | 6.87x |
| Google Ads | 3.20x |
| Blended, all channels | 2.77x |
| 0.69x, on $11.1K of spend |
Pinterest returns 69 cents for each dollar. The blend hides that, because Klaviyo Email's 6.87x pulls the average up.
The split that shows what is happening
first_order_contribution = new_customer_revenue − variable_costs
payback_months = months until cohort cumulative contribution ≥ CAC
These three numbers tell one story: what a customer costs, what their first order contributes, and how long the rest takes to arrive. Read them together:
- CAC below first-order contribution means the customer pays for themselves on day one.
- CAC above first-order contribution, with payback in a few months, is healthy if you have the cash to wait.
- CAC rising while blended ROAS holds steady means the repeat base is masking decay.
An ad export gives you cost per conversion. True CAC needs one more column, which is whether the order came from a first-time or returning customer, and you get that from your order history.
What to do when the two disagree
When blended ROAS is steady and CAC is rising, do three things in this order:
- Find the channel where cost per conversion moved the most. It is usually one or two.
- Look at whether that channel's creative or audience changed, since click-through falling as cost rises is the usual sign of tired ads.
- Move a small share of budget toward the channel with the best return and watch CAC for a month before moving more.
The sample audit does the same thing at larger scale. It prices a move of Pinterest budget toward Google Ads, and it leaves out the part of the move that goes beyond anything Google Ads has run at, because a return you have never seen at that spend is a guess.
Minimum setup
- Tag each order as first or repeat, using the customer ID and order date.
- Pull monthly ad spend for each channel.
- Divide spend by new customers for CAC, and compare it to the contribution on those first orders.
That is enough to compute CAC, payback and the new-versus-repeat split every month. Multi-touch attribution can wait. This cannot.
To size the limit for one channel, the break-even ROAS calculator turns your costs into the return you need, and the CAC ceiling calculator shows the most you can pay for a customer. The cohort LTV guide covers the payback side from a single order export. ProfitFalcon reads that export and splits new from repeat for you, and the Triple Whale comparison lays out how it differs on price and data access.
Upload your store export and see this for your own products, then decide.
Start 14-day trial14 days free. No card. Cancel anytime. Or open the sample store first.
BizFalcon AI