7 Signs You Can't Trust the ROAS Number on Your Dashboard

The ROAS number on your dashboard is the one most teams optimize to, and it is also the one most likely to be quietly lying. Before you scale a campaign because the platform says it's a winner, run through these seven signs you can't trust your ROAS — each is a concrete symptom that the figure is inflated and shouldn't be steering your budget. None of this means ROAS is useless; it means knowing exactly when to stop believing it.

For the full mechanism behind these signs, read why your ROAS doesn't match your profit; for what to do about it, the reported vs true ROAS framework.

1. Your Platforms Claim More Revenue Than You Actually Banked

This is the single most damning sign, and the easiest to check. Add up the revenue Meta, Google and TikTok each report for the month, then compare it to the real revenue in your order system. If the platforms together claim more than you actually made, the difference is pure over-counting — the same order billed to multiple channels.

What to check instead: real revenue from your order system, deduplicated to actual orders.

2. The Dashboard Improves While Your Margin Falls

A "winning" account whose contribution margin keeps shrinking is the classic tell. Reported ROAS climbs because budget has flowed to campaigns that are good at claiming sales — retargeting, branded search — while the prospecting that actually creates incremental demand gets starved. The screen looks healthier as the business gets less profitable.

What to check instead: contribution margin trend from your P&L, week over week.

3. ROAS Swings When You Change the Attribution Window

If flipping from a 7-day-click to a 1-day-click window cuts a campaign's reported ROAS in half, the number was never absolute — it was a function of the window. A figure that depends entirely on a setting you chose cannot be the basis of a profit decision.

What to check instead: run the same campaign on the strictest window (1-day click) as a sanity floor, and treat the gap to the looser window as borrowed credit.

4. View-Through Is Carrying a Direct-Response Campaign

Open the breakdown. If a meaningful share of a direct-response campaign's conversions are view-through — credited to an impression nobody clicked — the ROAS is padded with sales that may have happened anyway. View-through has a place in brand measurement; counted as performance ROAS, it manufactures wins.

What to check instead: click-through-only ROAS for direct-response campaigns.

5. A Big Share of Conversions Are Modeled, Not Observed

Since the privacy shift, platforms fill measurement holes with modeled conversions — statistical estimates rather than tracked events. A modeled conversion is a guess, and the guess is not neutral: it tends to flatter the campaigns the algorithm wants you to keep funding. The higher the modeled share, the less the number is a count.

What to check instead: the observed-vs-modeled split in your conversion reporting, where the platform exposes it.

6. Your Channels Each Look Profitable but the Blend Doesn't

Every platform's dashboard shows a healthy ROAS, yet your blended, profit-based number is mediocre. That contradiction is the over-count and the cost gaps stacking up. Individually each channel grades itself well; together, deduplicated and after costs, the story changes.

What to check instead: blended ROAS — contribution margin ÷ total ad spend across all channels — as the tiebreaker.

7. Nobody Can Reconcile the Number Against the Bank

If a founder or client asks "why does the dashboard say 4× and the bank say 1.5×" and your team can't answer in one sentence, the number isn't trusted — and untrusted reporting is worse than none, because it erodes confidence in everything else you show. Reconcilability is itself a trust signal.

What to check instead: a blended number you can defend line by line against the P&L.

When ROAS Is Still Worth Trusting

To be fair to the metric: inside a single account, in a single attribution window, reported ROAS is a fast and reliable relative signal. Ad set A beating ad set B on the same platform is real information, and you should keep optimizing creatives and ad sets on it. The discipline isn't to abandon ROAS — it's to stop reading a relative signal as an absolute profit measure. Tune the engine with reported ROAS; steer the car with the blended number.

How a Cross-Channel View Makes the Signs Visible

Most of these seven signs hide because the data is scattered — five tabs, three currencies, three windows — and nobody assembles the blended picture until month-end. A cross-channel view fixes the visibility, not the attribution. Wevion brings every connected platform's spend and reported results into one screen, normalized to one currency at the day-of-transaction rate, so signs 1, 6 and 7 stop being month-end surprises and become a number sitting in front of you. Data syncs roughly every 15 minutes.

To be honest about the boundary: Wevion gives a clearer cross-channel picture and is launch-plus-analytics on one screen, but it does not show "profit" as a live feature and does not perfectly solve attribution — no tool does. For the FX and matrix mechanics, see the cross-channel analytics features guide; for where a dedicated attribution tool still fits, the Triple Whale alternatives guide.

The Bottom Line

If your platforms claim more than you banked, your dashboard improves while margin falls, ROAS swings with the window, view-through carries your direct-response campaigns, conversions are mostly modeled, the channels look good but the blend doesn't, or nobody can reconcile the number against the bank — you can't trust the ROAS on your screen as a profit measure. Use it to compare ad sets; judge budgets on a blended, profit-based figure. Wevion makes the inflation visible by putting every platform's spend and reported results on one screen with day-of-transaction currency and launch in the same place — starting at a permanent free tier (€0), then Starter at €99/mo, Pro at €499/mo, Plus at €1,499/mo (€1,199 annual, billed yearly at −20%), and Enterprise as a custom plan, with a 14-day trial on every paid tier that coexists with the free plan. For the wider workspace this sits in, the ads management platform hub maps the rest.