Start free

Blended ROAS

Blended ROAS: how to check what Google and Meta really made

Four numbers from one closed week give you the return on your ads, your break-even line and how much the two platforms claim twice.

AdCtrl article cover: Blended ROAS, check what Google and Meta really made. The Google Ads, Meta, Claude and ChatGPT icons circle the AdCtrl mark, with Meta highlighted.

Short answer

  • Blended ROAS is store revenue divided by total ad spend for the same dates. $21,000 of sales on $7,000 of Google and Meta spend is 3.0×.
  • Google and Meta each credit a sale to their own ads inside their own windows: Google's click window defaults to 30 days, and Meta's API reads 7-day click plus 1-day view. One order can be claimed by both.
  • Break-even blended ROAS is one divided by your gross margin. Compare blended ROAS with that line, never platform ROAS: under it, the ads lose money however good each platform's report looks.
  • Subtract store revenue from what Google and Meta claim together. When both sides use the same dates and value definition, what's left is the least credit counted twice or missing from the store. It doesn't say which platform is right.
  • Meta's on-device conversion measurement is the default outside the EU and UK and, as reported, counts from 12 November 2026, so Meta can report more purchases while store revenue doesn't move.

Pull Google and Meta spend and claimed sales in one chat

Start free

3-day free trial · no card needed

Blended ROAS is store revenue divided by all ad spend for the same period. Work it out for a closed week from ad spend, store revenue, both platforms' claims and gross margin. Compare it with your break-even, then subtract store revenue from what Google and Meta claim together to see the least credit they count twice or the store never booked.

That minimum assumes matching dates and value definitions. For Google Ads or Meta against GA4, see why conversions don't match.

We run a performance agency and build AdCtrl, which reads Google Ads and Meta spend and claimed sales in Claude or ChatGPT; the store numbers come from you.

Checked 16 November 2026 against Google Ads Help and Meta's help and API pages1.

What is blended ROAS?

Blended ROAS is the revenue your store took in a period divided by everything you spent on ads in that period, on every platform. The store's sales go on top, not the conversion value Google or Meta report. In the example used through this article, $21,000 of sales on $7,000 of ad spend is 3.0×.

Platform ROAS is the version each ad account shows: the conversion value that platform credits to its own ads, divided by its own spend. Google Ads and Meta each work it out only from what they saw, so it answers a narrower question: how much does this platform think it made?

MER, the marketing efficiency ratio, is the closest cousin. HubSpot's guide calculates it from total revenue and total marketing spend, which can hold more than ads, such as agency fees or influencer costs. Triple Whale's metric docs write it the other way up.

HubSpot's MER guide says

"MER is calculated by dividing total revenue by total marketing spend for a defined period."2

Triple Whale's MER docs say

"MER = Blended Ad Spend ÷ Order Revenue", and for "the inverse ratio (Order Revenue ÷ Blended Ad Spend), use Blended ROAS."3

So check which way up a report writes MER before you compare it with someone else's.

Why do Google and Meta claim more sales than your store took?

Google and Meta claim more sales than your store took because each credits an order to its own ads whenever the shopper met one of them inside that platform's window. Neither checks whether the other already took the credit, so one shopper can be one order in your store and two in your ad reports.

Take one shopper. On Monday they click your Google search ad, look around and leave. On Wednesday they scroll past your Meta ad on Instagram without clicking, come back and buy the same day.

Google's click-through window is 30 days if nobody changed it when the conversion action was created1. Meta's view-through setting counts events within 1 day after an impression of your ad4. Both reports show the sale, and the store shows one order.

Diagram: a shopper clicks a Google search ad on Monday, sees a Meta ad on Wednesday without clicking and buys that day. Google Ads reports 1 conversion inside its 30-day click-through window, Meta reports 1 purchase inside its 1-day view-through window, and the store records 1 order.

One order, two claims: a shopper clicks a Google search ad on Monday, sees a Meta ad on Wednesday and buys that day. Google counts the purchase inside its click window, Meta inside its view window, and the store counts one sale.

Meta's own help page, on comparing attribution models, adds: "If using external analytics tools, evaluate performance in those tools."4 For an online shop, the store's sales report is the closest thing to that.

A duplicate tag can inflate a single platform too, when the same order is sent twice. That's a tracking fault, covered in the duplicate conversions checklist. If Meta shows fewer purchases than the store, the opposite problem, start with Facebook ads not converting.

What does each platform count by default?

Each platform counts by its own windows, and the defaults are far apart. Google waits 30 days after a click by default; Meta counts 1 or 7 days after a click on the ad's link and 1 day after a view. Defaults only apply where nobody changed them, so read the settings in your own accounts.

SettingMetaGoogle Ads
Click window1 or 7 days after a link click30 days by default for a new conversion action
View window1 day after an impression1 day by default
Engagement windowEngage-through: 1 day after any click on the ad that isn't a link click, or a 5-second video playEngaged-view: 3 days by default after someone engages with a video ad, settable from 1 to 30 days
If you don't chooseMeta's API reference lists 7-day click plus 1-day view as its read defaultData-driven attribution for most conversion actions
Date a sale lands onNot stated on the Meta pages cited here; see the Meta section of the discrepancy guideThe click date, unless you report by conversion time
Still rolling outSome accounts still use the older click-through and engage-through versionsNot applicable
Extra countingMeasuring conversions on customer devices, the default setting, not active for people in the EU and UKNot applicable

Attribution defaults, checked 16 November 2026. By default Google waits 30 days after a click and Meta 7, so the same week holds different sales in each report.

Source: Meta Business Help Center (attribution settings; measuring conversions on customer devices), Meta Marketing API Insights reference, Google Ads Help (conversion windows, attribution models, conversion tracking data), read 16 November 20264, 5, 6, 1, 7, 8

Meta's newest change adds conversions without touching the windows. It gives Meta an extra way to match an ad view to a purchase on the shopper's own device, working alongside the pixel and the Conversions API rather than replacing them.

The Events Manager setting, measuring conversions on customer devices, is the default, and Meta says it lets you "See more of the conversions that you drive in your campaign reporting"5. The help page has no start date; PPC Land reports that the Events Manager panel reads "Measurement starting November 12, 2026", and that the change applies to datasets that already send events9.

Screenshot of Meta's Business Help Center page About measuring conversions on customer devices, with the passage highlighted: your attribution window and attribution model don't change; measuring conversions on customer devices affects which conversions can be measured, not how they're credited to your ads.

Screenshot of Meta's Business Help Center, 16 November 2026, highlight added. Meta's own reporting can count more purchases from this setting while your store's sales stay the same.

So a Meta report after 12 November can show more purchases than the week before with nothing changed in your shop. Blended ROAS doesn't move with it, because it never uses Meta's count.

Google shifts sales in time instead. Google Ads puts a conversion on the day of the click, so a click on 28 October that turns into an order on 3 November lands in October's report, while the store books it in November. When comparing with Shopify, Google's help page points to reporting "based on the time the conversion occurred"8.

How do you calculate blended ROAS from four numbers?

You calculate blended ROAS from four numbers for the same closed week: total ad spend, store revenue, what each platform claims, and gross margin. Blended ROAS is store revenue ÷ total ad spend, and the other numbers tell you whether that's enough and how much the platforms overlap.

Here is the example week, made-up numbers used through the rest of this article. Google spent $3,000 and claims $12,000 of conversion value, a platform ROAS of 4.0×. Meta spent $4,000 and claims $16,800 of purchase value, 4.2×. The store took $21,000, and the gross margin is 40%.

  • 3.0× blended ROAS in the example: $21,000 ÷ $7,000
  • $28,800 claimed in the example by Google and Meta together
  • 1.37 over-claim ratio in the example: $28,800 ÷ $21,000
  • 2.5× break-even blended ROAS in the example, at 40% margin

Both platforms report above 4×, and the store says 3.0×. Neither platform is lying by its own rules. Each counted the sales it could see in its own window, and some of those were the same sales.

You need read access to both ad accounts, someone who can open the store's sales report (usually the owner or a store admin) and a gross margin figure from whoever does the books. Pick a week that ended at least two days ago, since Google says reporting can lag by 24 to 48 hours8.

  1. Pull Google Ads spend and conversion value for the week.

    Manual: in Google Ads, set the date range to last Monday to Sunday, open Campaigns and add cost and conversion value, using the columns that report by conversion time so each sale falls on the day it happened.

  2. Pull Meta spend and purchase value, with the windows named.

    Manual: in Ads Manager, set the same dates and show amount spent and purchase conversion value. Note each ad set's attribution setting, since results follow it. If you pull it through the API, action_attribution_windows sets the windows and its default option means 7d_click plus 1d_view.

    With AdCtrl: write "Show last Monday to Sunday's spend and claimed conversion value for Google Ads and for Meta, one total per platform."

  3. Take store revenue for the same dates.

    Manual: in your store's sales report, take total sales for the same Monday to Sunday, in the same time zone as the ad accounts. Keep one definition every week, for example with refunds taken off and without tax.

  4. Write down gross margin.

    Manual: the share of a sale left after the cost of the goods and payment fees, from your bookkeeping.

  5. Do the arithmetic.

    Manual: blended ROAS = store revenue ÷ (Google spend + Meta spend). Claimed total = Google value + Meta value. Over-claim ratio = claimed total ÷ store revenue. Break-even blended ROAS = 1 ÷ margin.

    With AdCtrl: put the store revenue and margin in the same message, and the assistant does the arithmetic on the numbers it read.

  6. Repeat it next week.

    Manual: keep one row per week in the same sheet, with the four numbers and the three results.

When the store export and Analytics disagree, keep the blended calculation on the store's revenue and use the Shopify GA4 purchase check to investigate the Analytics gap.

What blended ROAS do you need to break even?

Break-even blended ROAS is 1 divided by your gross margin. At a 40% margin that's 1 ÷ 0.40 = 2.5×: every $1 of ads has to bring in $2.50 of sales before the week's sales cover the ads and the goods sold. The ROAS calculator works out break-even with your own costs, and POAS too.

In the example, a blended ROAS of 3.0× sits 0.5× above break-even. The week's contribution after ads is $21,000 × 0.40 − $7,000 = $1,400, the gross profit left once both platforms are paid. Platform ROAS of 4.0× and 4.2× suggested much more room than that.

Stat card: break-even blended ROAS is 2.5× at a 40% gross margin, and the example week's blended ROAS of 3.0× sits above it.

Example numbers: break-even blended ROAS is 1 ÷ gross margin, 2.5× at a 40% margin, and the example week's 3.0× sits above it.

Blended ROAS counts every sale in the store, including repeat customers and people who found you without an ad. HubSpot's guide says the same of MER, which should include all revenue in the period, paid and organic2.

A shop with a strong base of returning customers can sit above break-even while the ads themselves lose money. A new shop with almost no organic sales gets a number close to what its ads really earn. So we read blended ROAS as the most the ads could have made, and watch how it moves when spend changes.

How much are Google and Meta double counting?

Google and Meta are over-claiming by at least the gap between their combined claims and your store revenue, when both sides use the same dates and value definition. In the example they claim $28,800 against $21,000, so at least $7,800 is credit counted twice or missing from the store. For every $1 the store took, the platforms claimed $1.37.

We line both sides up first. If a platform's conversion value includes tax or shipping and your store revenue doesn't, part of the gap is a difference in measurement. So is a sale that sits in one week in one report and the next week in the other, such as Google's click date against your order date. Once those match, organic sales only make the double-counted or unbooked credit larger.

The number doesn't say which platform is right. Meta claims 80% of the store's week and Google 57%, and either one could be counting the shared orders. Google itself expects discrepancies "often up to 20%" between Google Ads and third-party analytics8; that's one platform against one tool, not two platforms added together.

Our rule of thumb for reading the ratio: no level is safe on its own. Above 1.0, the platforms claim more than the store sold in total.

Store revenue also holds sales no ad touched, so a ratio near 1.0 can still hide a lot of shared credit. What we act on is the trend: a ratio that rises week after week while store revenue stays flat is the signal to run a test.

What would we do with the example numbers?

With the example numbers, we'd leave both budgets where they are. A blended ROAS of 3.0× against a 2.5× break-even means the week's sales covered the ads and the goods sold, with $1,400 of gross profit left, but not by much. Platform ROAS above 4× is no reason to raise spend, because the store doesn't see that 4×.

With Google's dates and both value definitions lined up, the 1.37 ratio says at least $7,800 of the week's credit is shared or missing from the store. We'd write the four numbers down for the next few weeks in the same sheet. If the ratio climbs while store revenue stays flat, or blended ROAS drifts toward 2.5×, we'd run a pause-and-watch test on the platform with the larger claim.

Which platform should you cut? Run a pause-and-watch test

Cutting a platform on its own ROAS means trusting the report that over-claims. A pause-and-watch test judges a channel by the store instead: pause one platform's campaigns, or one region, for a fixed period, and watch store revenue and blended ROAS. Leave the other platform's budget alone while you do.

With AdCtrl, a pause is a status change you see as a preview and approve in the chat, and turning the campaign back on is a separate approval.

If store revenue barely moves while spend drops, most of that channel's claims were sales you'd have had anyway or sales the other platform also claimed. If the gross profit you lose (the revenue drop × margin) is bigger than the spend you saved, the channel was paying its way.

One test isn't proof. Promotions and seasonality move store revenue too, so we'd keep everything else steady and compare with the same weeks before the pause. For how long and how big a fair test needs to be, see is Google Ads worth it.

We'd pause Meta first if

  • Meta's split by window shows much of its claim comes from views rather than clicks
  • Meta claims the larger share of store revenue week after week
  • Your store's own reports show a large share of sales from returning customers

We'd pause Google first if

  • The search terms report shows most conversion value comes from searches for your brand name
  • Google's claim keeps rising while blended ROAS stays flat

We wouldn't pause yet if

  • Blended ROAS is under break-even and the duplicate conversions checklist hasn't been run
  • A sale or holiday falls inside the test period
  • You can't keep the other platform's budget steady for the whole test

If the decision is how to split spend between ad sets in one Meta campaign, the Meta campaign allocation guide shows how to compare reported results and test a different split.

Use the weekly paid media review to turn a budget decision into a dated handoff for the people managing the accounts.

Run the check with AdCtrl

AdCtrl reads the platform half of the check in one chat: Google Ads cost and conversion value by campaign and date, and Meta spend, purchases, purchase value and purchase ROAS with the 1d_click, 7d_click and 1d_view windows. You give it the store revenue and margin.

AdCtrl report card with example data for last week: Meta spend, purchase value and purchase ROAS per campaign, adding up to $4,000 of spend and $16,800 of purchase value.

Real report card template with example data, not a real account: last week's Meta spend, purchase value and purchase ROAS per campaign. The store's revenue isn't on the card, because it comes from you.

In the example, Meta's $16,800 splits into $13,100 within 7 days of a click and $3,700 within 1 day of a view, so about 22% of Meta's claim came from views.

In AdCtrl, write: only reads your accounts
Show Meta spend and purchase value per campaign for last Monday to Sunday, with 7d_click and 1d_view as separate windows, and the share of each campaign's purchase value that came from 1d_view.

It has no store connection, so it can't read Shopify or any orders, and it doesn't deduplicate sales between Google and Meta or give a unified ROAS. The over-claim figure is arithmetic on what the platforms report and what you tell it.

  • Spreadsheet: Two platform exports and a store export, side by side each week

    (Manual)

  • Triple Whale: Connects to Shopify; its docs build Blended ROAS from order revenue and ad spend3, 10

    (Store data)

  • AdCtrl: Reads both platforms' spend and claimed value; store revenue comes from you

    (Platform data)

Checklist before you trust the number

Run these checks each week before you move budget on the strength of one blended ROAS figure. Any no is the first thing to fix.

  • Spend, store revenue and both claims cover the same Monday to Sunday.
  • The store report and both ad accounts use the same time zone.
  • The week ended at least two days ago.
  • Google's conversion value is reported by conversion time.
  • You know which Meta attribution windows the claimed value uses.
  • The weeks you compare sit on the same side of Meta's November 2026 counting change, or you compare blended ROAS only.
  • No duplicate purchase tag is inflating one platform.

Frequently asked questions

Is a 1.5 blended ROAS good?

It breaks even only if your gross margin is at least about 67%, because 1 ÷ 1.5 = 0.667. At a 40% margin, 1.5× loses money: $1,000 of ads brings $1,500 of sales and $600 of gross profit. Judge any blended ROAS against your own break-even, not a range from someone else's shop.

Why did Meta report more purchases in November 2026?

Meta started measuring conversions on customer devices as the default outside the EU and UK, which adds purchases to campaign reporting with the same windows. PPC Land reports that advertisers "can turn it off at any time in the settings of each dataset" in Events Manager9. Your store's sales don't change with it, so check blended ROAS before you raise Meta budgets on a higher purchase count.

Which Meta attribution window should I report?

For this check we'd read 7-day click and 1-day view separately. The 1-day view share shows how much of Meta's claim rests on people who saw an ad without clicking it, which is the part we'd expect to overlap most with a Google click.

Does blended ROAS work for a lead business?

Only if the revenue on top is closed-deal revenue from your CRM for the same period. Leads can close weeks later, so use a longer period than a week. To send those sales back to Google, see offline conversions.

Can AdCtrl read my Shopify orders?

No. There is no store connection: it reads what Google Ads and Meta report, and you give it the store revenue in the chat.

What to do next

Take last week and write down four numbers: ad spend on both platforms, store revenue, what each platform claims, and your gross margin. Work out blended ROAS, break-even and the over-claim, then do the same next week in the same sheet; the trend is the number to watch. If blended ROAS sits close to break-even, run a pause-and-watch test and judge it by the store.

To go deeper, why conversions don't match covers reconciling Google Ads and Meta with GA4, and the ROAS calculator works out break-even and POAS. To pull both platforms' side of the check in Claude or ChatGPT, see how to connect AdCtrl and start the free trial.

Sources

  1. About conversion windows, Google. Accessed 16 November 2026.
  2. Marketing efficiency ratio: How to calculate and improve yours, HubSpot, 2 January 2026. Accessed 16 November 2026.
  3. MER (Triple Whale metrics docs), Triple Whale. Accessed 16 November 2026.
  4. About attribution models and attribution settings, Meta. Accessed 16 November 2026.
  5. About measuring conversions on customer devices, Meta. Accessed 16 November 2026.
  6. Ad Account, Insights (Marketing API reference), Meta. Accessed 16 November 2026.
  7. About attribution models, Google. Accessed 16 November 2026.
  8. Understand your conversion tracking data, Google. Accessed 16 November 2026.
  9. Meta Pixel will count more conversions from Nov 12 unless advertisers opt out, PPC Land, 27 October 2026. Accessed 16 November 2026.
  10. Mapping Shopify Fields to Triple Whale, Triple Whale. Accessed 16 November 2026.

Keep going