Ecommerce and Shopify ·Explanation
Blended ROAS: compare ad spend against store revenue
Read blended ROAS against the orders your store recorded, compare each platform's claim and check budget pacing before month end. With a worked example.
Your blended ROAS report has a problem: you close one sale, and by Monday both Google and Meta have claimed it. Each dashboard reports a return that says it earned the credit, and each is telling the truth about itself and nothing about the other. Paid budget is the one marketing line that bleeds in real time, and the blended picture is almost always worse than any single platform admits, because the same sale gets counted twice. Reading paid spend across channels well is how you decide where the next dollar goes instead of where the last one went. Below: the blended cost and return read on one honest denominator, the per-channel split that names the dollar to move, and the pacing check that keeps the month inside budget.
Marlowe is a fictional worked example. Its figures illustrate the reading method; they are not customer results or independent benchmarks.
What is paid spend across channels, and why read it as one view?
Paid spend across channels is every paid effort you run, reconciled into one view so you can see where the money goes across channels and what comes back. It is not one efficiency metric. It is the whole paid program read on a single, shared definition of cost, conversion and value, so Google, Meta, search, social and display sit in one place and answer the same two questions: is the money working, and where should the next dollar go?
paid spend across channels := every paid channel's cost and return reconciled into one view, on one shared definition of spend, conversion and value, so the whole program reads together instead of one platform at a time
The reason you need the one view is that each platform grades its own homework. Google reports conversions on Google’s attribution. Meta reports conversions on Meta’s. Both count the same sale, both take full credit, and neither knows the other exists. Stack their dashboards side by side and you are not comparing apples to apples. You are comparing each platform’s most flattering self-portrait. The double-count collapses only when the denominator changes: run total spend against the orders your store actually recorded, and the real cost per result shows up.
Where paid spend sits across the funnel
Paid spend reconciles the paid effort that runs across every band of the funnel: display and most paid social buy attention near the wide top, high-intent and branded search closes near the narrow bottom, and retargeting reaches back to pull people toward the buy. For how those bands work, see the pillar’s funnel section.
The leak that matters hides between channels, not inside one. A single platform can look efficient on its own report while the blended program bleeds, because the same conversion is being claimed twice and your true cost per result is higher than any one dashboard admits. The transition most teams never measure is the hand-off from per-platform reporting to a blended view of cost and return, total spend against the orders the store actually recorded, across the whole program at once. That is the view this piece is about.
What to read across your paid channels
These are the concepts to read across the program, not platform by platform. The efficiency read runs cost against return, then pace against budget. Every concept gets more honest the moment it is defined once and applied to every channel the same way.
| Read | The question | What good looks like |
|---|---|---|
| Spend | What are we paying, and per what? | Cost per click and cost per thousand impressions stable or falling at constant quality. |
| Cost to result | What does one outcome cost? | Blended cost per acquisition below contribution profit per sale, for purchase-scoped conversions, your break-even line. |
| Result | Did the spend produce outcomes? | Conversion rate near 7% for search ads (WordStream 2024), with display and social typically lower; the blended count reconciles to the orders your store recorded. |
| Return | Did the spend pay back? | Return on ad spend above your target. A ROAS of 4 means $4 back per $1 in. |
| Pace | Are we on track to budget? | Projected month-end spend lands inside plan. |
Read every one of these by platform, by campaign and by the action that pays back. Take Ava, the performance-marketing lead at Marlowe, a mid-size clothing brand with a handful of its own shops and an online store. She runs Google, Meta and a display line across the apparel catalog, and the platform split is what makes her whole read work: it is what lets her put Google, Meta and the rest in one column and read return down the rows.
How to read your blended ROAS and per-channel results
Most stacks hand you the data and stop one rung up at description: here is your spend, here is your return per platform. The reads below climb the rest of the way. They diagnose why a blended return lags when every platform looks healthy, point to the decision of which dollar to move, and leave you able to check the outcome next month, whether the reallocation actually paid back. One number decides whether a dollar is worth spending, and one habit decides whether you can trust the comparison.
Blended cost per acquisition is the honest cost of a result. It is total paid spend across all channels divided by the orders your store actually recorded, not the per-platform figure each dashboard reports. Those per-platform figures double-count shared conversions and read lower than reality. When Google tells Ava a sale cost $18 and Meta says $22, her blended figure is almost certainly higher than both, because some of those conversions are the same Marlowe shopper counted twice. Marlowe’s online store runs on Shopify; the orders are the denominator both platforms are arguing over.
Return only means something on one definition. Return on ad spend is conversion value divided by spend, but the value and the cost have to be defined the same way for every channel or the comparison is noise. Read it against a target, not against zero: a ROAS of 3 is healthy for a high-margin coat and a loss-maker for a thin-margin basics tee.
Read both blended ROAS and per-channel return, never one alone. Blended ROAS is total conversion value across all paid channels divided by total spend. It tells you if the paid program as a whole pays back, but it hides the weak channel: a 4.0 blended ROAS can be a 7.0 search channel carrying a 1.5 display channel that is quietly losing money. The blended number is the health check; the per-channel split is the reallocation decision. You need both, because the blend tells you nothing about which dollar to move.
The reallocation read is mechanical once the definition is shared. Where you overspend relative to return is spend against return by platform and campaign. High spend feeding below-target return is the first budget to cut; high return still starved of spend is where the freed dollar goes. Drain the channel earning least per dollar, feed the one earning most, and re-check next month.
Catch pacing while you can still act. Projected month-end spend read against plan mid-month is a budget adjustment. The same number read on the 30th is an apology to finance. Pacing is the cheapest control you have, and only if you read it early.
Scale by the action that pays back. Conversion value by action tells you which outcomes carry the revenue, so you scale those and stop optimizing toward a cheap action that never turns into money. The whole program is only as trustworthy as the single definition underneath it: define cost, conversion and value once, and every read above stops being an argument.
How Maya turns scattered data into answers
Most reporting stops at what happened: here is the spend, here is the return per platform. Maya climbs further. From the data it surfaces an insight, from the insight a diagnosis (why a blended return moved), from the diagnosis a decision (which dollar to move), and from the decision an outcome you can check next month (did the reallocation pay back). That climb up the value pyramid, past description into why, what-to-do and did-it-work, is the whole point, and it only holds if the foundation underneath is solid.
That foundation is three layers, built in this order:
- A marketing data warehouse unifies every platform’s raw exports, Google, Meta, search, social, display, into one cloud database, reconciled so the numbers become comparable across sources.
- A governed Power BI semantic layer defines cost, conversion and value once, so the model cannot quietly invent what they mean or how each channel should be counted. Governed means the definitions are enforced, not that you are restricted.
- A conversational layer, a Model Context Protocol (MCP) server, lets an AI assistant (Claude, ChatGPT, a Slack bot) query that governed layer in plain language. We have written tools for our agents that encode our team’s experience with paid media and its best practices.
The split that makes this work: your prompt steers the answer, the semantic layer enforces what the words are allowed to mean. That is what separates a governed MCP from a chatbot wired to a database, and it is why a cross-channel comparison comes back as a read rather than a guess.
What the unified view changes for paid spend:
- One view across every paid channel. Ava’s spend from Google, Meta and every other source lands in the same governed model, where cost, conversion and value each mean one thing, so every channel reconciles to the same arithmetic. Two raw exports compared side by side are two different counting rules. Maya settles that before she asks the question, and she reads the whole program in one place instead of logging into each platform in turn.
- The blended ROAS read built in. Marlowe’s online store runs on Shopify, and the blended read runs total paid spend against the orders that store actually recorded, so a sale two platforms both claim still counts once. The blended number arrives paired with the per-channel split, so the health check and the reallocation decision land together.
- Read pacing against your plan. Compare spend from the latest daily sync with the budget and assumptions you supply. A pacing read does not establish that scheduled alerts or margin gates have been configured.
- Campaign clusters you define, ten levels deep. Ava doesn’t just want to know how the spring sale push did; she wants to track every campaign she runs, at every level of depth, in one place. In Maya’s UI she builds hierarchical clusters from simple text-matching rules, no code, up to ten levels: level one by channel (Google, Meta, display), level two by objective (prospecting, retargeting), level three by product line (outerwear, shirts), and on down to the individual campaign. Each level previews in real time as she tweaks a rule and refreshes on its own as new spend loads into the model, so nothing goes stale or needs rebuilding. Now she reads spend and return at any altitude, the whole paid program, then one channel, then one objective, then a single campaign, as one consistent hierarchy instead of reconciling a flat export thousands of rows long.
Reading spend and return by campaign cluster
Here are the top levels of that hierarchy (it runs down to the individual campaign at the deepest level), and the read it gives her:
flowchart TD
Paid["All paid · root"]
Paid --> Google["Google · level 1"]
Paid --> Meta["Meta · level 1"]
Google --> GPro["Prospecting · level 2"]
Google --> GRet["Retargeting · level 2"]
Meta --> MPro["Prospecting · level 2"]
GRet --> GRetOuter["Outerwear retargeting · level 3"]
| Cluster | Spend vs last month | Platform-reported return | Pace to budget | What Ava reads |
|---|---|---|---|---|
| Google › Retargeting › Outerwear | +22% | 6.8, strong | on plan | earning hard, keep feeding it |
| Google › Prospecting | flat | 2.1, thin | on plan | break-even, hold and watch |
| Meta › Prospecting | +31% | 1.3, below target | overspending | the leak, first budget to cut |
Meta prospecting is the channel quietly bleeding while outerwear retargeting earns its keep: the answer to her question, plus the dollar to move that a blended-only number would have hidden.
Without Maya: an analyst exports Google Ads, exports Meta Ads Manager, tries to reconcile two different conversion definitions, builds a blended pivot, argues with finance about which cost-per-result is real, and a day later hands over a return comparison nobody fully trusts because each platform counted the same sales. With Maya: Ava asks “Which channel has the best return this quarter, and where am I overspending?” Maya reads spend and return by platform on one governed definition, runs the blend against the orders Marlowe’s Shopify store recorded, flags the channels burning budget below target, and hands back the ranked list with the diagnosis. Minutes, not a day.
The questions a marketer actually asks Maya:
- “How is my paid spend performing across channels?” Spend, conversions and return by platform, with the blend run against recorded orders and the channels below target flagged.
- “Which channel has the best return?” Return by platform, but only once every platform sits on the same conversion and value definition.
- “Where am I overspending relative to return?” Spend against return by platform, campaign cluster and campaign, the reallocation shortlist.
- “What is my blended cost per result, and am I on pace to budget?” Total spend against the orders your store recorded, with spend to date compared with your stated plan.
One honest limit: a shared definition fixes the counting, not the attribution. It does not by itself settle which channel deserves credit when a buyer touched several, so a fair cross-channel return still depends on a consistent attribution model. The blended read sidesteps that argument by using the store’s recorded orders as its denominator; the per-channel split still inherits each platform’s own attribution. See web analytics for where the paid click lands and how the conversion gets counted on-site.
Maya does the assembly and the guardrails. You do the thinking, the allocation, and the creative.
The bottom line
Stop reading each platform’s dashboard as if it told the truth about the others. It tells the truth about itself, on its own attribution, and that is exactly the problem. Put every paid channel on one shared definition of cost, conversion and value, then read return by platform against a target, not against zero and not against another platform’s self-report.
Check the blended ROAS, then move money on the split. Blended return tells you whether the program pays back; the per-channel read tells you which dollar to move. Cut the channels spending hard below target, feed the ones earning above it, and watch the month-end pace projection so the pacing decision happens while you can still make it. Then let Maya assemble the numbers and enforce the guardrails, so your team spends its hours on the budget split, the creative and the offer instead of reconciling exports.
So the question worth sitting with is this: does your stack tell you which dollar to move next, or only that each platform, on its own report, is working?
Glossary
- Paid spend across channels: every paid channel’s cost and return reconciled into one view, on one shared definition of spend, conversion and value, so the whole program reads together instead of one platform at a time.
- Spend: what you pay a platform for delivery, read per click (cost per click) and per thousand impressions (cost per mille).
- Cost per acquisition: spend divided by conversions. For purchase-scoped conversions, compare it with contribution profit per sale; the click-cost ceiling also depends on conversion rate.
- Conversion rate: conversions divided by clicks (or sessions). Near 7% for search ads (WordStream 2024); display and social typically run lower.
- Return on ad spend (ROAS): conversion value divided by spend, on one shared definition of both. Read against a target, not against zero. A ROAS of 4 means $4 back per $1 in.
- Blended cost per acquisition: total paid spend across all channels divided by the orders your store actually recorded. The honest cost of a result, higher than any single platform’s self-reported figure.
- Blended ROAS: total conversion value across all paid channels divided by total spend. The program-level health check; it hides the weak channel, so always pair it with the per-channel split.
- Recorded orders: the orders your store actually logged, read from the commerce platform itself. The honest denominator for blended reads: two ad platforms can claim the same sale, but the store records it once.
- Month-end pace projection: the projected month-end spend at the current pace, read against plan so a pacing problem is caught while you can still act.
- Campaign cluster: a group of related paid campaigns rolled up under a shared rule (a channel, an objective, a product line), read together rather than one campaign at a time. Maya lets you build these as a nested hierarchy from text-matching rules in its UI, up to ten levels, so you read spend and return at any altitude.
Related
- the full-funnel pillar: how every discipline maps to one model
- search ads: the high-intent paid channel that closes at the action stage
- paid social: the awareness-to-action paid channel and its efficiency reads
- web analytics: where the paid click lands and how the conversion gets counted on-site
- customer lifetime value: the return side, what a converted customer is worth over time
Ask this about your own account
Connect your sources, then ask Maya the question this piece just answered, on your own numbers.