Blended ROAS vs. Platform ROAS (and MER): What Is the Difference?

Open Meta Ads Manager and it says 4.2x. Google Ads says 3.8x. Add up the revenue both platforms claim and it’s more than your store actually took in last month. Then you divide total revenue by total ad spend and get 2.6x.

All three numbers are “ROAS.” Only one of them, blended ROAS, matches your bank account.

The usual reaction is to trust the dashboards and over-scale, or to declare the platforms “lie” and ignore the data that tells you which ad to turn off. Both are mistakes. The two metrics answer different questions.

This guide covers what each metric measures, why they rarely match, how to read the gap between them, and a simple rule for which number makes which decision.

What is platform ROAS?

Platform ROAS is the revenue an ad platform attributes to its own ads, divided by what you spent on that platform. It’s the number in Meta Ads Manager, Google Ads, or TikTok Ads Manager, calculated with that platform’s tracking and attribution rules.

The key word is attributes. Meta, Google, and TikTok each decide which conversions to credit to themselves using their own lookback windows and models. None of them can see the other platforms’ ads. So each platform answers a narrow question: “Of the purchases we could observe, how many happened after someone interacted with our ads?” Useful, but not the same as “how much revenue did this spend create?”

What is blended ROAS?

Blended ROAS (blended return on ad spend) is your total revenue divided by your total ad spend across every paid channel, for the same period. It ignores attribution entirely: no windows, no models, no platform claiming credit.

The formula is deliberately blunt:

Blended ROAS = total store revenue ÷ total paid media spend

If your store did $300,000 last month and you spent $100,000 across Meta, Google, and TikTok, your blended ROAS is 3.0x (illustrative numbers). Because the revenue number comes from your store and the spend comes from your invoices, nothing can be counted twice.

You’ll also see this called MER (marketing efficiency ratio). The two are close cousins, and the section on MER vs ROAS below sets out where they differ. What matters is that you define it once and calculate it the same way every month.

Why don’t blended ROAS and platform ROAS match?

Because each platform grades its own homework, and the grading rules overlap. The three biggest causes of the gap are double counting, different attribution rules, and revenue that no platform caused or claimed.

We go deep on the mechanics in our guide to paid media attribution and why Meta, Google, and Shopify don’t agree. The short version:

1. Multiple platforms claim the same order

A customer sees a TikTok ad on Monday, clicks a Meta ad on Wednesday, then searches your brand name and clicks a Google ad on Friday before buying. TikTok, Meta, and Google can each count that purchase, because each one saw an interaction inside its own window. Your store recorded one order.

2. Each platform uses different attribution rules

The windows and models differ, so the same journey gets scored differently:

• Meta credits conversions within an attribution window after a click, a view, or an engagement. Meta recently narrowed click-through attribution to link clicks only and moved social interactions such as likes, shares, and saves into a separate category it now calls engage-through attribution. Which windows are switched on changes how much revenue your reports show.
• Google Ads uses data-driven attribution as the default for most conversion actions, spreading credit across the ad interactions it observed. It also reports conversions against the date of the click, not the date of the purchase in its main columns, so last week’s numbers can keep rising after the week ends.
• Shopify uses last non-direct click as the default model for its marketing reports, with a 30-day window. A view-through conversion that Meta counts will never appear as Meta revenue in Shopify.

None of these is wrong. They are different rules, so they produce different numbers.

3. Some revenue isn’t caused by ads at all

Blended ROAS puts all revenue in the numerator: repeat customers responding to an email, organic search, word of mouth, a creator who posted about you unpaid, people who would have bought anyway. That revenue makes blended ROAS look better than paid media deserves. Meanwhile, platforms also take credit for some of those buyers (the loyal customer who happened to scroll past a retargeting ad before reordering), which makes platform ROAS look better than it deserves too.

Neither metric measures incrementality, the revenue that would not have happened without the ad. That’s why neither one is the whole truth, and why paid media measurement needs both.

Platform ROAS vs. blended ROAS at a glance

	Platform ROAS	Blended ROAS
What it measures	Revenue a platform attributes to itself ÷ that platform’s spend	Total store revenue ÷ total ad spend
Source of revenue	Platform tracking and attribution model	Your store or finance system
Double counting	Yes, across platforms	No
Granularity	Campaign, ad set, ad, creative	Business-wide only
Biggest blind spot	Overclaims shared and non-incremental conversions	Can’t tell you which channel or ad did the work
Best for	In-platform optimization	Budget, scaling, and profitability decisions

When should you use blended ROAS?

Use blended ROAS for decisions about the whole business, such as total budget, scaling and profitability, and use platform ROAS for decisions inside a platform. The mistake is using either number for the other job. (Reconciling Meta, Google and Shopify order by order is a separate problem; our guide to paid media attribution covers which source to trust for which question.)

A simple rule we use: the bigger and more expensive the decision, the more it should rest on blended numbers.

Use platform ROAS for in-platform decisions

Platform ROAS is biased, but the bias is roughly consistent within a platform. If Meta overstates every ad set by a similar amount, it still ranks those ad sets correctly against each other. That makes platform data the right tool for:

• Picking winning creative and killing losers
• Comparing audiences, placements, and campaign structures on the same platform
• Spotting fatigue: an ad whose platform ROAS slides week over week is fading, whatever the absolute number is

Blended ROAS can’t do any of this. It has no idea which of your 40 ads is carrying the account.

Use blended ROAS for business decisions

Blended ROAS reconciles to real money, so it’s the referee for decisions where double counting would be expensive:

• Total budget: whether to spend more next month, or less. Our guide to setting an ecommerce advertising budget builds this from unit economics, with blended numbers as the check.
• Scaling calls: if you raise Meta spend 30% and Meta’s reported ROAS holds but blended ROAS falls, you likely bought attributed revenue rather than new revenue.
• Profitability: your P&L runs on store revenue, not attributed revenue. Whether an ecommerce ROAS figure is actually profitable depends on your margins; see what a good ROAS for ecommerce really is for how to find your break-even.
• Reporting to founders and finance: nothing in blended is modeled, so nobody can argue with it.

Use both for channel allocation, carefully

Splitting budget between Meta, Google, and TikTok is where neither metric is enough on its own. Platform ROAS will overstate every channel, but not by the same amount, so comparing Meta’s 4x to Google’s 4x is comparing two differently-sized exaggerations. Blended ROAS can’t break down by channel at all.

The practical approach is to watch how blended ROAS moves when you change one channel’s spend and hold the others steady. If you cut a channel by 25% and total revenue barely moves, that channel was claiming more than it was creating. If blended revenue drops noticeably, it was doing real work. For larger accounts, controlled lift tests and marketing mix models formalize this. Google’s open-source Meridian is one example of an MMM framework built for this question, though it needs meaningful spend history and analytical skill to run well.

MER vs ROAS: what is the difference?

ROAS divides the revenue attributed to ads by ad spend, for one platform or campaign; MER (marketing efficiency ratio) divides total store revenue by total marketing cost, so it can’t double count but can’t point at a channel either. For an ecommerce brand the practical differences are three:

	ROAS (platform)	Blended ROAS	MER
Numerator	Revenue the platform attributes to itself	Total store revenue	Total store revenue
Denominator	That platform’s ad spend	Total ad spend	Total marketing cost, often including fees, creative and tools
Answers	Which ads and campaigns to fund inside a platform	Is paid media paying for itself overall?	Is the whole marketing budget paying for itself?
Can it double count?	Yes, across platforms	No	No

Many teams use MER and blended ROAS interchangeably; the difference only matters when fees and production are a large share of cost, so write your definition down and keep it. For the general definition and formula, AdBoost Health keeps a MER glossary entry.

What is new-customer ROAS?

New-customer ROAS is the revenue from first-time customers divided by ad spend over the same period, and it shows whether ads are growing the customer base or mostly reselling to people who already buy. It closes the biggest gap in blended ROAS, which a loyal repeat base can flatter.

To calculate it, take the revenue from first-time customers in your store data (Shopify’s customer reports include a New vs returning customers report to start from) and divide it by total ad spend. Illustrative example: a month with $300,000 in revenue, of which $120,000 came from first-time customers, on $100,000 of ad spend, has a blended ROAS of 3.0x but a new-customer ROAS of 1.2x. If your first-order break-even is 2.5x, the ads are acquiring customers at a loss on the first order, which is fine only if repeat purchases reliably earn it back.

Read new-customer ROAS next to new-customer count. Both falling while blended ROAS holds is the classic sign that spend has drifted into retargeting and branded search.

How do you read the gap between the two numbers?

The gap between platform and blended ROAS isn’t noise to ignore. It’s information, as long as you track it consistently.

Track the ratio, not just the levels

Each month, divide the sum of your platform-reported ROAS figures (weighted by spend) by your blended ROAS. Call it your attribution inflation ratio. Say platforms claim a combined 4.0x and blended is 2.5x (illustrative): your ratio is 1.6.

The absolute value matters less than its trend:

• Ratio stable, both numbers rising: healthy growth. The platforms are inflating by a consistent amount and real revenue is climbing with them.
• Platform ROAS rising, blended flat or falling: the warning sign. Platforms are claiming more credit without the store seeing more money. Common causes: spend drifting into retargeting and branded search that harvests existing customers, broader view-through or engagement windows, or overlap growing as you add channels.
• Blended ROAS rising, platform flat: something outside paid media is working (email, organic, PR, a strong product launch), or a platform is under-tracking because of a pixel or conversion API issue. Check tracking before you take the credit.

Once you know your ratio, translate targets

A stable ratio lets you turn a business target into a platform target the media buyer can actually manage to. If you need 2.5x blended to hit your margin goal and your ratio has held near 1.6, the in-platform target is roughly 4.0x. Revisit the ratio monthly and after any big change: a new channel, a new attribution setting, a major sale.

Where does blended ROAS mislead?

Blended ROAS is the honest number, but it isn’t the complete one. Three traps worth knowing:

It rewards a strong existing customer base. A brand with a big, loyal repeat-buyer list can post a great blended ROAS while its paid media acquires almost no one new. The fix is to track new-customer revenue separately: new-customer revenue divided by ad spend tells you whether ads are growing the customer base or just sitting beside it.

It hides channel waste. A great month on one channel can carry a channel that’s losing money. Blended says “fine” while one line item quietly burns budget. That’s why blended is the referee, not the whole scouting report.

It lags and swings with non-ad events. Seasonal peaks, price changes, stockouts, and press all move blended ROAS without any change in ad performance. Compare like periods and annotate unusual months.

Where does platform ROAS mislead?

It flatters retargeting and branded search. These campaigns reach people already close to buying, so they post the highest platform ROAS in most accounts while being the least incremental. Scale them on platform ROAS alone and blended ROAS stalls.

It shifts when settings change. Toggle a view-through or engagement window, or let a platform update its attribution definitions, and your reported ROAS moves without a single extra sale. Always note attribution settings next to any platform ROAS you report.

It can’t see the whole journey. Each platform sees only its own touchpoints. A TikTok campaign that creates awareness people later act on through Google search will look weaker in TikTok’s dashboard than its real contribution, which is why judging upper-funnel channels purely on in-platform ROAS tends to starve them.

What does a practical weekly and monthly routine look like?

You don’t need an attribution platform to run a disciplined version of this. You need two numbers and a habit.

Weekly, inside each platform: use platform ROAS (and CPA) to make creative and ad set calls. Compare ads to each other, not to your blended target.

Weekly, across the business: check blended ROAS and new-customer revenue against total spend. If they move the opposite way from the platform dashboards, investigate before scaling.

Monthly: recalculate your attribution inflation ratio, reset in-platform targets from your blended target, and review any channel whose spend changed meaningfully against what happened to total revenue.

Quarterly, if spend justifies it: run a holdout or lift test on your largest channel, or a geographic test where you pause spend in some regions. Direct evidence of incrementality beats any amount of dashboard debate.

Adjust the inputs to your category. A fashion brand with heavy returns should calculate blended revenue net of returns, or blended ROAS will look healthier than the cash says.

So, which number makes which decision?

Both numbers have a job. Platform ROAS is the right instrument for optimizing inside Meta, Google, or TikTok: it’s granular, fast, and consistent enough to rank ads against each other. Blended ROAS is the right instrument for running the business: it reconciles to real revenue and can’t be inflated by overlapping claims.

Keep each number in its lane, track the gap between them as its own signal, and you’ll stop arguing about which dashboard is lying.

If you’re evaluating partners to run this for you, ask how they report. An agency that only shows you platform ROAS is grading its own homework with the platform’s answer key. Our guide to choosing a performance marketing agency covers the reporting questions that separate the two.