Why Two 'ROAS' Numbers Never Match

You open Meta Ads Manager and it says 4.2x ROAS. Then you check Shopify or GA4 and it says 2.1x. Same week, same store, same ad spend. Finance asks the obvious question: which one is real?

Both are. That's the annoying part.

Blended ROAS vs channel ROAS isn't a case of one dashboard lying and the other telling the truth. They're built to answer different questions, using different math, different attribution logic, and in most cases, different definitions of "revenue" entirely. Once you see the mechanics, the gap stops feeling like a bug and starts looking like exactly what you'd expect.

This isn't a Meta problem, or a Shopify problem, or a "your analytics tool is broken" problem. It's a measurement design problem. Every ecommerce brand running paid ads on more than one platform runs into it eventually.

What Blended ROAS Actually Measures

Blended ROAS is the simplest number in ecommerce marketing: total revenue divided by total ad spend, full stop.

Take a week where your store did $150k in revenue across every channel, and you spent $30k combined across Meta, Google, and TikTok. Blended ROAS is $150k / $30k = 5x. No attribution windows, no per-platform pixel, no argument about who gets credit for the sale. Just store-wide revenue over store-wide spend.

That simplicity is the whole point. Blended ROAS is the number you show a board or a CFO because it can't be gamed by one platform's generous attribution settings. It's also the best number for tracking overall marketing efficiency over time. Is your total ad spend as a percentage of revenue trending up or down quarter over quarter? Blended ROAS answers that cleanly.

What it can't do is tell you anything about where the revenue came from. A 5x blended ROAS could mean Meta is carrying the business and Google is dead weight, or the reverse. The number is flat and honest, but it's blind to channel mix. If your job is deciding where to shift next month's budget, blended ROAS alone won't get you there. You need to look one level down.

What Channel ROAS Actually Measures

Channel ROAS is revenue attributed to one specific platform, per that platform's own tracking, divided by that platform's own spend. Meta's reported ROAS uses Meta's pixel and Meta's attribution model. Google's uses Google's. TikTok's uses TikTok's.

The catch is that "attributed to" means something different on every platform. Meta defaults to a 7-day click / 1-day view window. Google often stretches to 30-day click attribution. TikTok has its own version, and it's changed more than once. That means each platform is counting conversions that happen inside a different time frame, using a different definition of what counts as "influenced by an ad." Widen the window, and more conversions get pulled in. More conversions pulled in inflates the reported ROAS, even with zero change in actual sales.

Then there's the self-attribution problem. Every ad platform is financially and structurally motivated to claim credit for a sale. If a customer saw ads from three different channels before buying, there's a real chance all three platforms report that single sale as their own conversion. Nobody's lying exactly, they're each just measuring their own slice and reporting it as if it's the whole picture.

None of this makes channel ROAS useless. It's exactly the right number for in-platform decisions: which ad set to kill, which creative to scale, whether to raise a daily budget on a specific campaign. It's a bad number for deciding how much total budget Meta deserves versus Google. That comparison needs a neutral referee, which is where tools like GA4 solutions usually come in.

Where the Gap Comes From: Attribution Overlap and Double Counting

Here's the customer journey that breaks every dashboard: someone clicks a TikTok ad, doesn't buy, sees a Meta retargeting ad three days later, still doesn't buy, then searches your brand name on Google and converts.

Three platforms touched that sale. Depending on attribution windows, all three might claim it. TikTok logs a conversion. Meta logs a conversion. Google logs a conversion, often at a huge reported ROAS because branded search converts at a high rate almost by definition. Add up the revenue each platform claims, and you can end up with a number well above what the store actually made, sometimes by a wide margin.

Blended ROAS doesn't have this problem, because it isn't tracking the journey at all. It just asks: did the store make money this week, and how much did we spend on ads to get there? That sale counts exactly once, at the register, regardless of how many platforms think they earned it.

There's a second layer to this too: view-through attribution. A platform counting someone who merely saw an ad (never clicked) as a conversion is making a much weaker claim than one requiring an actual click. Platforms that lean heavily on view-through credit tend to report cleaner-looking ROAS, not because the ads are better, but because the bar for "credit" is lower.

If you're trying to compare Meta ROAS to Google ROAS to decide budget splits, you're really comparing two different measurement philosophies dressed up as the same metric.

When to Use Which Metric

Use blended ROAS when the audience is finance or the board, or when the question is "is total marketing spend efficient." It's the number for setting overall ad spend as a percentage of revenue, and for spotting a real efficiency trend that isn't tied to any single channel's quirks.

Use channel ROAS when you're inside a platform making tactical calls: which campaign to pause, which audience to expand, which creative variant is actually working. That's the layer where channel ROAS is genuinely the right tool, especially for performance marketers running day-to-day optimization.

The mistake worth naming directly: don't line up Meta's reported ROAS next to Google's reported ROAS and use that comparison to decide how to split next month's budget. Different attribution windows make that a false comparison dressed up as data. A platform with a 30-day click window will almost always look better than one running a 7-day window, regardless of which one is actually driving incremental sales.

The fix is triangulation. Pull in GA4 as a neutral third party, or run actual incrementality tests (holdout groups, geo tests, MMM-style modeling) if you want to know what each channel would look like if it disappeared tomorrow. That's the only way to get real channel-level truth instead of platform-reported channel-level marketing.

How Trivas Reconciles Blended and Channel Views

The reason most brands end up doing this math by hand in a spreadsheet is that their tools don't store the data in a way that lets both views coexist. Trivas pulls Amazon, Shopify, Meta and Google ads, and GA4 into one Redshift-based warehouse, so blended and channel numbers are sitting in the same data model instead of living in five separate dashboards that each define revenue differently.

In practice, that means a founder opens BI reporting and sees blended ROAS at the top, trending week over week, and can drill straight into Meta, Google, or TikTok's channel-level numbers underneath without re-exporting anything or rebuilding a pivot table.

The Wingman AI layer sits on top of that and flags the specific case that actually matters: a channel reporting a ROAS spike while blended ROAS stays flat. That pattern almost always means attribution overlap, not real growth, and it's the kind of thing that's easy to miss when you're only looking at one platform's dashboard in isolation.

Building a Reporting Habit Around Both Numbers

Check blended ROAS first, every week, before you touch anything channel-specific. It's your baseline. If it's moving in the right direction, you're probably fine. If it's flat or dropping, that's your cue to go diagnose which channel is underperforming.

A short checklist worth running monthly:

  • Confirm attribution window settings on every platform. Don't assume they're aligned.
  • Exclude branded search from channel comparisons. It converts high regardless of ad spend and skews Google's numbers upward.
  • Reconcile channel-reported revenue against GA4 sessions and revenue to catch inflated attribution early.

One pattern to treat as a red flag every time: a channel's self-reported ROAS jumps while blended ROAS sits flat. That's not the channel getting better. That's attribution overlap pulling credit from somewhere else in the funnel.

Get Both Numbers in One Place

Blended ROAS tells you if the business is healthy. Channel ROAS tells you where to push budget inside a platform. Neither one replaces the other, and any dashboard that only shows you one is giving you half the picture.

If you want to see where your own numbers land, run them through the ROAS calculator, or start a trial to see blended and channel ROAS sitting side by side on your actual store data.