To understand what your advertising returns, compare platform-reported ROAS with recorded sales, then connect those sales back to your campaigns. In a recent Supermetrics and Azalea webinar, the speakers showed how that connection can change which campaigns look worth funding.
Your ad accounts tell a reassuring story. Campaigns are converting, return on ad spend (ROAS) looks healthy, and the monthly report suggests your budget is working hard.
Then finance shares the sales figures. The story gets less reassuring.
Which number should guide next month's budget? And how do you explain why they disagree without spending the meeting defending your spreadsheet?
That was the problem behind Your ROAS is lying to you, our October 27 webinar with Azalea, a revenue-connected marketing analytics platform. Co-founders Ren Agarwal and Aniket Deshpande joined Supermetrics Senior Account Manager Meghan Hodge and host Joanna LeBlang to explain why the gap appears and demonstrate a way to investigate it.
The webinar showed how the two platforms work together. Supermetrics brings campaign data into a warehouse, and Azalea matches it to recorded sales. You can then see which campaigns deserve a closer look before you move budget.
Watch the full webinar here.
Key takeaways
- Ad platforms can credit the same purchase, so adding their reported revenue together can overstate the sales behind your return.
- In the webinar’s example, the same $50,000 in ad spend yielded a ratio of 4.2x using revenue claimed by the platforms and 1.6x using total recorded sales.
- Supermetrics supplies the marketing data foundation. Azalea adds the sales matching and reporting that help you examine campaigns against recorded revenue.
- Matching a campaign to a sale doesn't establish that the ad caused it. Incrementality testing answers that additional question.
- Verdict: Before deciding which campaigns deserve more budget, check how their reported results connect to sales. For a practical example, see how to connect ad spend to CRM or ecommerce revenue.
Why doesn't platform-reported ROAS match your sales figures?
Platform-reported ROAS reflects the revenue each advertising platform credits to its ads. Your sales system records the purchases that actually happened. Those figures answer different questions.
Imagine someone discovers your product through a social ad, watches a video later that week, then searches for your brand and buys. Depending on their attribution settings, more than one platform may claim credit for the purchase.
Each platform can report a conversion according to its own rules. Add those figures together, though, and one customer starts looking suspiciously like several.
This is why marketing attribution matters when you compare results across channels. The ROAS calculation itself is straightforward: revenue divided by ad spend. Deciding which revenue belongs in the numerator is where the disagreement starts.
It also explains why a good-looking ad report can coexist with disappointing business growth. A campaign may receive credit for purchases from returning customers who were already likely to buy. Another may introduce new customers but receive less credit because they convert later through a different channel.
Your platform report still helps you understand performance within that platform. However, it needs a connection to your wider business results before you use it to tell the whole story.
How can you check the size of your ROAS gap?
Compare the revenue your ad platforms claim with the revenue your sales system records, using the same business scope and a clearly defined reporting period.
Ren opened with a simple illustrative example:
| Revenue basis | Revenue | Ad spend | Revenue divided by spend |
|---|---|---|---|
| Revenue claimed across three ad platforms | $210,000 | $50,000 | 4.2x |
| Revenue recorded in the sales system | $80,000 | $50,000 | 1.6x |
The difference is 2.6 ROAS points. The platforms collectively claim $130,000 more revenue than the business recorded.
That's a useful warning light. It doesn't tell you which platform overclaimed, or how much of the recorded revenue advertising caused. Dividing total sales by total ad spend gives you a blended efficiency measure that can include organic sales and repeat purchases.
Before interpreting the gap, check that you're comparing like with like. Agree on currency and revenue treatment, including refunds. Check whether the ad report assigns conversions to the interaction date or purchase date, and allow for purchases that arrive after the initial click.
See how to account for conversion lag when comparing ROAS.
Once you've established that a gap exists, the next question is more useful: which campaigns connect to the sales you've actually recorded?
How do Supermetrics and Azalea connect ad spend to sales?
Supermetrics brings marketing data into a shared data warehouse. Azalea combines that foundation with sales records and applies the matching logic that connects campaign activity to revenue.
Say you want to investigate a campaign that looks excellent in an ad platform but contributes little in your sales report. You need its spend and campaign identifiers alongside the relevant customer activity and transactions. Comparing two totals won't explain the difference.
In the setup shown during the webinar, Supermetrics brings together paid media data with web analytics and CRM or first-party data. Scheduled transfers keep that information available in a warehouse such as BigQuery or Snowflake, where teams can analyze it alongside sales records.
Azalea then handles the sales join and revenue-matching logic, with dashboards for exploring the results. The connection depends on the identifiers and journey data available; putting two datasets in the same warehouse doesn't automatically link every purchase to an ad.
How do you compare campaigns that use different labels?
Even after you connect the data, you need to know what each campaign was trying to do. Comparing a prospecting campaign with a retargeting campaign gets difficult if their names don't identify the audience consistently.
Aniket demonstrated Azalea's PRISM framework, which uses predefined campaign tags:
| Tag | What it describes |
|---|---|
| Presentation | The format |
| Reach | The audience |
| Intent | The objective |
| Segment | Who the campaign targets |
| Message | The copy theme |
Teams choose from agreed values rather than inventing a new label each time. Azalea's UTM builder uses those tags to create tracking URLs.
This makes a question such as “How do our prospecting campaigns compare across channels?” easier to answer. You're grouping campaigns by a shared definition instead of hoping everyone used the word “prospecting” in the same way.
The same principle underpins good campaign naming conventions: decide what you need to compare before the results arrive.
What changes when you compare campaigns against recorded revenue?
Connecting campaigns to recorded sales can change their ranking, giving you a different starting point for budget decisions.
During the demo, Aniket compared Google prospecting with a TikTok sales campaign. Using the platforms' own figures, TikTok appeared stronger. Using the demo's sales-linked figures, Google prospecting came out ahead.
| Demo campaign | Platform-reported ROAS | Sales-linked ROAS shown in the demo |
|---|---|---|
| Google prospecting | About 6.4x | About 14.5x |
| TikTok sales | About 11x | About 6.9x |
These are illustrative webinar demo figures, not customer results.
The report you open changes which campaign looks worth funding. TikTok leads on platform ROAS; Google prospecting leads on the sales-linked view. Platforms can also under-credit campaigns. Prospecting often introduces customers who buy later through another channel, so its recorded revenue can exceed what the platform claimed.
Aniket also compared prospecting with retargeting. Again, the apparent winner changed depending on the revenue basis. That matters because retargeting reaches people who have already interacted with your business. A high reported return deserves a closer look at which customers bought and what happened earlier in their journey.
Before moving budget, check matching coverage and whether the campaigns reached comparable audiences. Then use the comparison to design a controlled spending test.
Azalea also demonstrated an Ask Azalea chat interface for questions about the connected warehouse data, such as ranking campaigns by ROAS. The value is being able to follow up on the comparison. A ranking becomes more useful when you can ask what sits behind it, a principle that also applies to chatting with your marketing data in other tools.
Does sales-linked reporting replace MMM or multi-touch attribution?
Sales-linked reporting can sit alongside marketing mix modeling and multi-touch attribution. Each approach helps answer a different question about performance.
In the Q&A, Ren explained that Azalea can complement a team's existing measurement methods.
Marketing mix modeling (MMM) estimates how channels and other factors contribute to business outcomes over time. Multi-touch attribution (MTA) assigns credit across recorded customer interactions. Connecting those interactions to sales lets you inspect the transactions behind your campaign reporting.
These aren't mutually exclusive approaches. MMM can use sales as its outcome, and MTA can link touchpoints to recorded purchases. The useful question is whether your particular setup reconciles with sales and makes its assumptions clear.
There's a further question that even a matched transaction can't settle on its own: would the customer have bought without seeing the ad?
Ren addressed that directly: “Nobody can prove that from attribution alone.”
A controlled incrementality test can help you measure how many additional sales your ads generated. A properly designed test compares outcomes for a group exposed to advertising with a suitable control group. Simply comparing two campaigns' ROAS doesn't isolate the effect of advertising.
Our guide to AI reporting and attribution explains where incrementality testing fits alongside other measurement methods. Use sales matching to understand the recorded revenue behind a campaign, then choose the additional method that answers your next question.
Why does reliable ROAS reporting need an owner?
Reliable reporting needs someone who can maintain the connection between marketing and sales as your business changes. Building the first dashboard is only the start.
Ren put the problem neatly: “Marketing analytics is an engineering product owned by a non-engineering team.”
Marketers depend on a system that needs technical upkeep. A new sales stage can change a calculation. A new channel can require different fields. If the person who built the model leaves, the dashboard may remain while the knowledge behind it disappears.
The webinar grouped these issues into three problems: trusting the revenue number, finding the technical skills to maintain it, and making the reporting usable. Fixing the first won't help for long if nobody owns the other two.
Every Azalea account includes a forward-deployed engineer who builds and maintains the revenue matching and stays involved after implementation. In the plan Ren outlined, the first two weeks map the existing setup and identify where marketing and sales fail to connect. Weeks three to six focus on integrations and revenue matching, followed by ongoing maintenance.
The wider lesson applies to any reporting setup: establish who owns metric definitions and who changes them when the business changes. That ownership deserves a place in the buying decision alongside the dashboard itself. It’s also a theme in our 2026 Marketing Data Report.
What should you bring to your next budget conversation?
Bring campaign spend alongside the sales you can match to it, and come prepared to explain which budget decision you want to test.
Put platform-reported ROAS and sales-linked ROAS side by side. If they point to different winners, show where the rankings change and flag any gaps in the sales data. That gives you a concrete starting point for defending your ad spend.
For B2B campaigns, bring pipeline influence and win rate by source. Pipeline influence shows which opportunities a campaign touched. Win rate tells you what share of closed opportunities from each source became customers. A campaign can appear in plenty of deal journeys while the opportunities it sources rarely close. That’s worth investigating before you give it more budget.
Watch the Supermetrics and Azalea webinar to see how the campaign comparison works. To see this on your own data, book an Azalea demo.
Frequently asked questions
-
No. Platform ROAS remains useful for understanding results under that platform's attribution rules. Use it alongside sales-linked reporting and other paid campaign metrics, rather than treating it as a complete account of business performance.
-
No. ROAS compares revenue with advertising spend. Profit also depends on costs such as producing and delivering the product. A campaign's ROAS needs that commercial context before you decide whether to scale it.
-
No. Aggregate totals can flag a discrepancy, but identifying overlapping purchases requires suitable transaction or event records and matching rules. The totals alone don't reveal which purchases appear more than once.
-
For the setup demonstrated, you need advertising data and recorded sales, plus the relevant web analytics or CRM records that help connect them. The implementation team needs to check which identifiers are available and where matching breaks down.
-
The webinar described pipeline influence and win rate by source as additional reporting views. For B2B teams, these depend on connecting campaign activity with the relevant CRM records and agreeing on what counts as sourced or influenced pipeline.