Published October 22, 2026
In an April 2025 test across six ad accounts and about $1 million in spend, Seer Interactive found that Meta counted 87% of conversions as caused by its ads, while GA4 credited 67%. Both tools were working as designed. They use different measurement rules. Neither figure, on its own, establishes what marketing caused.
A DTC brand should measure growth by contribution margin and new-customer revenue from its own store data, check acquisition cost and payback against them, and use holdout tests to confirm what each channel adds. Platform ROAS remains useful for comparing campaigns inside a channel. A CFO also needs to know what the whole budget earned.
ROAS is the right way to compare campaigns inside one ad platform, and the wrong way to judge whether marketing as a whole is making the brand money. Each platform uses its own attribution window and records of who saw an ad. As a result, two platforms can claim the same order, and the combined figures can credit more revenue than the store recorded.
ROAS also leaves out most of what decides profit. It ignores product cost, discounts, shipping, returns and the agency's fee. A campaign can post a strong ROAS on a product that loses money once those costs are counted. A brand that raises prices or runs a sale changes every channel's ROAS overnight without any change in how well the ads work.
The error runs in both directions. Haus reports that judging Meta only on 7-day click conversions undercounts its effect by about 15%, so a platform's number can be too low as well as too high. Some sales never reach an ad platform at all. A shopper who finds a product through ChatGPT or Perplexity can arrive with no referrer and land in direct traffic, where no platform claims it. Use store orders to measure the business, then investigate where platform figures differ.
Show your CFO what ad spend drives with a monthly report built from store orders and finance data, covering contribution margin after marketing, new-customer revenue and acquisition cost, and total marketing spend against total revenue. A CFO already trusts the store and the ledger, so those are the numbers to start from.
| Metric | What it shows | How to calculate it | What to watch for |
|---|---|---|---|
| Contribution margin after marketing | Whether marketing leaves the brand with more money | Revenue minus product cost, discounts, shipping, returns and all marketing costs | Include agency fees and tools, or the margin looks better than it is |
| Marketing efficiency ratio (MER) | Return on the whole marketing budget | Total store revenue divided by total marketing spend | It rises when returning customers grow, even if acquisition is weak |
| New-customer MER | Whether spend is still finding new buyers | Revenue from first orders divided by total marketing spend | A falling trend while total revenue grows means growth is leaning on existing customers |
| New-customer acquisition cost | What each new buyer costs | Total marketing spend, including fees, divided by new customers | Compare it with first-order margin, not with average order value |
| Acquisition cost payback period | How long a new customer takes to repay what they cost | Acquisition cost divided by the monthly gross margin a new customer brings | Longer payback ties up cash, which matters more to a CFO than to a media buyer |
Agree on the definitions with finance before the first report, especially what counts as marketing cost and how returns are handled. That agreement gives marketing and finance a shared basis for reviewing results.
Holdout tests measure what a channel adds by turning it off, or holding it back, for a comparable group of people or regions and comparing the results. The comparison estimates what would have happened without the spend.
In a 2025 survey reported by EMARKETER and TransUnion, 52% of US brand and agency marketers said they use incrementality testing. Meta and Google both offer lift studies, and a brand can run simpler tests on its own by pausing a channel in a set of regions for a few weeks.
For brands spending across many channels, marketing mix modeling estimates each channel's contribution from historical spend and sales. Google's Meridian, an open-source model, made this cheaper to start than a custom build. A model needs enough history and variation in spend to be reliable, so a brand can start with holdout tests and add a model as spend grows.
A simple holdout test holds one channel back in a group of regions and compares store sales there with similar regions where the channel keeps running. The steps below are enough to run one without a vendor.
The gap between the two groups is the channel's incremental contribution, and it is the number to set against the platform's own claim. A brand with thin sales in some regions gets noisy results, so a longer test or a larger holdout gives a safer read than a quick one.
Set marketing targets on contribution margin and new customers, agreed with finance at the start of each quarter, and review the platform numbers against store orders every month. Platform dashboards then become a diagnostic tool, useful for finding which campaign to change, while the targets stay anchored in the business.
Each group reads the numbers it can act on. The team running each channel checks platform figures weekly to adjust campaigns, marketing and finance review the store-level report monthly, and each quarter they agree targets and read the latest holdout results together.
The same discipline applies inside a channel. Our account reporting for Chan Luu covers a single quarter and cannot prove cause by itself, but it shows the point clearly. In the second quarter of 2026, revenue from Chan Luu's automated email flows rose 21.8% while the flows reached 44% fewer recipients. Sends and opens would have made that quarter look worse, because the result that mattered was revenue per recipient.
When the store, the platforms and the finance team disagree, reconcile the numbers before anyone sets a new target. These checks are where to start.
A DTC growth audit should finish this work and show the real split of revenue by channel, with the cost behind each part.
CB/I Digital is a performance marketing and AI search agency that plans paid media, email and SMS, conversion, development and search together for DTC brands and measures them in profit. Your monthly report shows revenue, margin and new customers from your store's data, next to the work we shipped. Each month we compare what the ad platforms claimed with the orders your store recorded, and we use holdout tests to check what each channel really adds.
We agree on targets with you in the plan. Nothing changes in your budget, campaigns or tracking until someone on our team signs off, including work our AI agents prepare.
Our DTC growth services put paid media, email, the store and search on one plan. Get a clear picture of where your store is losing revenue, which channels earn their spend, and what to fix first with our DTC Growth Audit.
A good MER depends on your margins, so set it from the numbers rather than a benchmark. Work out the MER at which contribution margin after marketing is zero, then set the target above that break-even point. Two brands with the same MER can have very different profit if one sells at 70% gross margin and the other at 40%.
No. Keep platform ROAS for decisions inside a channel, such as which campaign, audience or creative to scale. Just do not report it as the return on the marketing budget, and do not add ROAS figures from several platforms together.
Not to start. Holdout tests and a clean monthly report from store data answer the questions a CFO asks first. A model becomes worth the effort when spend covers many channels and small shifts in allocation have a material financial effect.
Mike Le
Co-Founder, CB/I Digital and Conative AI
Mike Le co-founded CB/I Digital in New York in 2007 and has helped brands grow through search, paid media and data ever since. He started the agency's AI team in 2019 and later started Conative AI. He writes about how customers find and choose brands now that AI answers sit between them and the search results.
Follow Mike on LinkedInCopyright © 2026 CB/I DIGITAL INC. All rights reserved | Privacy Policy