ROAS is the metric most D2C and ecommerce teams check first, and it is genuinely useful. It just answers one narrow question: for the spend attributed to a campaign, how much revenue was attributed back. It says nothing about what that revenue cost to fulfil, whether the customer was new, whether they will return, or whether the attribution behind the number is trustworthy. Growth decisions made on ROAS alone tend to look correct for a quarter and wrong for a year.
What ROAS actually measures — and what it leaves out
ROAS is a ratio between attributed revenue and platform spend. Both sides of that ratio are incomplete. The spend side usually excludes creative production, agency or team cost, discounting, shipping and returns. The revenue side is reported by the same platform that is being judged, at whichever attribution window that platform prefers.
That does not make ROAS useless. It makes it a campaign diagnostic rather than a business metric. Use it to compare creatives, audiences and campaign structures inside one platform. Do not use it on its own to decide whether the brand is growing profitably.
- It is platform-reported, not business-verified
- It ignores cost of goods, shipping, discounts and returns
- It treats a first-time buyer and a loyal repeat buyer identically
- It changes when the attribution window changes, even if nothing else does
CAC and blended acquisition cost
Customer acquisition cost puts the question the right way round: what does it cost to acquire one customer, not what multiple of revenue did a campaign report. Channel-level CAC helps you manage a channel. Blended acquisition cost — total acquisition spend divided by total new customers, across every channel — tells you what the business is really paying to grow.
The gap between the two is where most surprises live. When channel ROAS looks strong while blended cost keeps rising, the channels are usually claiming credit for demand that already existed. Reviewing both together is what stops a brand from scaling a number instead of a business.
Conversion rate, AOV and the store's own performance
Acquisition cost is partly a media problem and partly a store problem. If the product page does not answer objections, or the checkout adds friction, media has to pay to compensate for it. That is an expensive way to fix a conversion issue.
Track conversion rate at the stages that decide the outcome rather than only site-wide: product page to add-to-cart, cart to checkout initiation, checkout initiation to payment. Then read average order value alongside it. AOV moving up while conversion falls is a different situation from both rising together, and the two need different responses.
- Product page to add-to-cart
- Add-to-cart to checkout started
- Checkout started to payment completed
- Average order value by product set and by channel
- Where mobile behaviour diverges from desktop
Contribution and margin context
Revenue is not a decision-grade number in ecommerce. Contribution after cost of goods, shipping, payment fees, discounts and returns is. Two campaigns with the same reported ROAS can contribute very differently once product mix and discounting are applied.
Once you can see contribution by product set and by channel, several decisions become obvious that were previously arguments: which SKUs deserve media support, where discounting is buying volume at a loss, and how much acquisition cost the business can genuinely absorb.
New versus returning mix, repeat purchase and retention
A blended ROAS figure can stay flat while the underlying business quietly changes shape — for example when returning customers carry more of the revenue and new-customer acquisition weakens. Splitting new versus returning revenue is the fastest way to see that early.
From there, repeat purchase rate and time between orders tell you whether the product and post-purchase experience are doing their job. Customer lifetime value is worth building only once those inputs are stable; a lifetime value estimate built on a few months of data will justify almost any acquisition cost you want it to.
- Share of revenue from new versus returning customers
- Repeat purchase rate and typical time to second order
- Which first products lead to a second order
- Lifetime value, once repeat behaviour is stable enough to model
Attribution and source quality
Most reporting disagreements are attribution disagreements. Platforms count conversions they influenced, analytics counts sessions it saw, and the finance view counts orders that were paid for. All three can be internally consistent and still contradict each other.
The practical fix is not a perfect model. It is a single agreed reference for orders and revenue, consistent campaign tagging, and a habit of reading platform numbers as directional signal rather than truth. That is the work we cover on our analytics and attribution page — making the numbers comparable before optimising them.
Reading the metrics together
No metric on this list is meaningful alone. ROAS without contribution hides margin problems. CAC without repeat purchase hides whether acquisition was worth it. Conversion rate without AOV hides mix effects. Attribution without an agreed order reference makes every other number arguable.
A workable review rhythm is to read acquisition cost, conversion, contribution and repeat behaviour in one view, on the same cadence, with the same definitions. Campaign-level ROAS stays where it belongs — inside the ad platform, for creative and audience decisions.
Where to start
Start with definitions, not dashboards. Agree what counts as a new customer, which order source is authoritative, and what costs are included in acquisition. Most brands find their reporting was never wrong so much as inconsistent.
Then instrument the checkout and product-page steps properly, split new versus returning revenue, and add contribution by product set. Once those exist, media decisions stop being a debate about whose number is right.
Frequently asked questions
- Is ROAS still worth tracking for a D2C brand?
- Yes, but for the job it is good at: comparing creatives, audiences and campaign structures inside one ad platform. It is a campaign diagnostic, not a measure of whether the business is growing profitably.
- What is the difference between CAC and blended acquisition cost?
- CAC is usually measured per channel. Blended acquisition cost divides total acquisition spend by total new customers across all channels, so it reflects what the business actually pays to grow rather than what one platform claims.
- Which conversion metrics matter most for ecommerce?
- Stage-level conversion is more useful than a single site-wide rate: product page to add-to-cart, cart to checkout started, and checkout started to payment completed, read alongside average order value.
- Why is contribution more useful than revenue?
- Contribution accounts for cost of goods, shipping, payment fees, discounting and returns. Two campaigns with identical revenue and ROAS can contribute very differently once product mix and discounts are applied.
- When should a brand start using customer lifetime value?
- Once repeat purchase behaviour is stable enough to model. Lifetime value built on a short history tends to justify whatever acquisition cost the brand hopes to afford, which makes it a risky input for budget decisions.
- Why do ad platform numbers disagree with analytics and finance?
- Each system counts a different event with different rules and windows. The practical fix is one agreed authoritative source for orders and revenue, consistent campaign tagging, and treating platform figures as directional signal.
Continue exploring