Your agency's dashboard says 5.9x ROAS this month, yet your bank balance barely moved. This gap between agency-reported ROAS and actual revenue is one of the most common disputes between businesses and the people running their ads, and it is rarely fraud.
It comes down to how each ad platform counts credit for a sale. A publicly filed agency services agreement shows just how baked-in this uncertainty is: the marketer in that contract does not promise a specific return on investment from the campaign it runs.
Performance marketing work has also generated $6.6 million in revenue from $1.1 million in ad spend, a 5.9x ROAS, but even that figure needed reconciling against real order data before either side trusted it. Here's where the mismatch comes from and how to check it yourself.
Too Long? Read This First
- Meta's default attribution window (7-day click, 1-day view-through) inflates ROAS; switching to 7-day click only gives a more honest number.
- A healthy Marketing Efficiency Ratio (MER) for established DTC brands sits between 3.0x and 5.0x, a better scaling signal than any single platform's ROAS.
- As of 2026, expect roughly 2x to 4x on B2B LinkedIn (90-day cycles), 3x to 6x on Meta for D2C, and 4x to 8x on Google Search lead gen.
- CRM tagging of every opportunity by source is the only way to check an agency's revenue claim against your own books.
What Is Platform-Reported ROAS?
Platform-reported ROAS is the return-on-ad-spend figure calculated inside a single channel's own ads manager, such as Meta or Google, using that platform's own attribution model to decide which sales count as its doing.
It is revenue the platform attributes to its ads, divided by what you spent there. Because each platform grades its own homework, it has every incentive to claim as much credit as it can reach.
ROAS is only one of several outcome metrics agencies get judged on; The performance marketing KPIs that actually matter cover the fuller set, including CAC and pipeline value. Treat platform ROAS as a channel-level bidding signal, not a statement about total company revenue.
Why In-Platform ROAS Can Mislead Revenue Reality
In-platform ROAS misleads because the same shopper's journey often gets credited to more than one channel, and the counting window itself is generous by default. Two mechanics drive most of the gap.
1. Attribution window inflation
Meta's default attribution setting, as of 2026, counts a sale up to 7 days after someone clicks an ad and 1 day after they merely saw it, so a shopper who never interacted with the ad can still be credited a day later.
Setting attribution to 7-day click only removes that view-through count and produces a more honest figure. This fixes one platform's padding, but it does nothing for overlap with other channels.
2. Retargeting cannibalisation
Retargeting campaigns almost always report the account's highest ROAS, because they show ads to people who already browsed, added to cart, or bought before. Much of that revenue would have arrived without the ad, so the ROAS overstates what the spend actually added.
Ads That Don't Burn Cash
Get a strategy built for ROI—not vanity metrics.
Incrementality testing, comparing an exposed group against a holdout that sees no ads, is the only way to separate genuine lift from revenue that was coming anyway. The fix is not to distrust every number, but to know which settings and campaign types inflate a figure before comparing it to revenue.
3. Misattribution From Overlapping Channel Claims
Overlapping channel claims mislead revenue reporting because two or more platforms can each log credit for the same purchase. This kind of cross-channel overlap pushes platform-reported conversions 15 to 20% above actual results on average, largely because last-click attribution rewards whichever channel the buyer touched last, regardless of what earlier channels contributed.
If Google Search and Meta both claim the same order, simply adding their reported revenue together double-counts that sale. Multitouch attribution models and incrementality testing give a more reliable read because they split or test credit rather than letting every channel claim the full sale. Treat the summed ROAS across all your channels as a ceiling, not a total, since some of that revenue was claimed twice.
How Marketing Mix Modelling Gives a Truer Picture
Marketing mix modelling (MMM) gives a truer revenue picture by statistically analysing historical spend and outcomes across the whole business, rather than tracking individual users.
It becomes especially useful once privacy restrictions, such as iOS tracking limits and cookie deprecation, make user-level tracking unreliable. The table below compares the main methods for reading true marketing performance.
| Method | What it measures | Best used for | Key limitation |
Platform ROAS | Revenue vs spend inside one channel's own attribution | Daily bid and budget decisions | Double-counts sales other channels also claim |
MER | Total revenue vs total marketing spend, business-wide | Monthly budget and scaling calls | Does not show which channel drove which sale |
Marketing mix modelling | Statistical contribution of each channel from historical data | Strategic channel-mix planning | Needs months of consistent spend history, not real-time |
Incrementality testing | Causal lift from an exposed group vs a holdout | Confirming a channel adds new revenue | Slower to run, and reduces spend in the test group |
MMM will not tell you which ad set to pause tomorrow, but it is the check that stops platform-optimised spend from quietly cannibalising channels it gets no credit for.
How to Independently Verify Reported Results
You verify reported results by comparing platform dashboards against your own order and revenue data, not by trusting a single reporting source. Start with MER, calculated as total revenue divided by total marketing spend, and compare it against the 3.0x to 5.0x range typical of established DTC brands.
Then compare it against channel benchmarks: by 2026, B2B campaigns on LinkedIn Ads typically post a ROAS between 2x and 4x across 90-day cycles, Meta Ads aimed at D2C should hit 3x to 6x, and Google Search Ads built for lead gen should reach 4x to 8x and count as healthy performance.
A performance-first agency should also have already set up server-side tracking, GA4 funnels, and attribution windows matched to your purchase cycle and offline touchpoints, so ask to see that reconciliation, not just the platform export.
Include a clear attribution clause in the scope of work before signing to prevent reporting disputes later. A single reconciled MER trend, reviewed monthly, matters more than any individual campaign’s ROAS screenshot.
CRM Tagging for Source Attribution
CRM tagging solves misattribution by recording each opportunity's real source in your own system, independent of what any ad platform claims. Every opportunity should be tagged by source at creation, because if an agency cannot show the dollar value of the deals it created in your CRM, that is a genuine accountability gap.
For B2B, this means tagging leads at first touch and carrying that tag through to closed-won revenue, so the number you compare to spend is actual cash, not modelled credit. For ecommerce, tagging first-order source alongside repeat-purchase behaviour catches revenue a platform logs as new business when it was really a returning customer.
CRM tagging will not replace platform reporting, but when the two disagree, the CRM's dollar figures are what your finance team should trust.
Conclusion
Agency-reported ROAS is a channel-level bidding metric, not a statement about your total revenue, and the two will keep diverging for as long as attribution windows, overlapping claims and retargeting cannibalisation exist. MER, incrementality testing and CRM tagging are how you close that gap rather than argue about a single dashboard number.
Ads That Don't Burn Cash
Get a strategy built for ROI—not vanity metrics.
If you run one acquisition channel, tightening its attribution settings may be enough. If you run several, MER and marketing mix modelling together give the more dependable read. Before your next budget conversation, pull your own order data and check it against whatever ROAS your agency reported.
Frequently Asked Questions
Why does agency-reported ROAS not match my actual revenue?
Because platform attribution windows count view-through activity and multi-day clicks, and several channels can each claim credit for the same sale, inflating the reported figure.
What ROAS benchmark should I compare against in 2026?
As of 2026, expect roughly 2x to 4x on B2B LinkedIn over 90-day cycles, 3x to 6x on Meta for D2C, and 4x to 8x on Google Search lead gen.
What is MER and how does it differ from ROAS?
MER divides total revenue by total marketing spend across the whole business, while ROAS is calculated per platform; a healthy DTC MER sits between 3.0x and 5.0x.
Can I trust agency-reported ROAS at all?
Yes, for bidding and budget decisions inside one channel. It is not reliable for company-wide revenue decisions because it ignores cross-channel overlap.
How do I fix Meta's ROAS inflation?
Switch Meta's attribution setting from the default 7-day click, 1-day view-through to 7-day click only, which removes the view-through overcount.
What is incrementality testing?
It compares an exposed group against a holdout group that sees no ads, isolating the revenue a channel actually adds rather than revenue it simply claims.
