Platform-reported ROAS rarely matches your real business KPI because each ad platform measures the conversions it can see and attributes them to itself. When several platforms reach the same customer, each can claim the same sale, so the sum of reported results can exceed what actually happened. The number to manage the business with is the one measured on your side, against your own sales or leads, not the sum of what each platform says.
Why ad platforms report more conversions than you had
Ad platforms tend to over-report in total because each one attributes conversions from its own point of view, with its own rules.
A customer sees a video on one platform, clicks a social ad on another and searches before buying. Each platform sees its own touchpoint and can count the conversion. None of them is necessarily wrong by its own rules. The problem is that the rules differ, and nobody inside any single platform is responsible for the total.
The four reasons platform numbers diverge
Four mechanisms explain most of the gap between platform-reported results and the advertiser’s KPI.
| Reason | What happens |
| Overlapping attribution | Several platforms claim the same conversion |
| Different attribution windows | Each counts conversions over a different period after the click or view |
| View-through and modeled conversions | Conversions are credited without a click, or estimated when the signal is missing |
| Different definition of the goal | The platform optimizes its event, the business measures revenue, margin or qualified leads |
None of these is hidden, all are documented in each platform’s settings. But taken together, they make the sum of platform reports a poor proxy for the business result.
Which number to trust for budget decisions
For budget decisions, the number to trust is the KPI measured on the advertiser’s side, backed where possible by incrementality measurement.
That means your own sales, leads or revenue, from your own systems, as the source of truth. Platform data remains useful for optimizing inside each platform, it’s the signal each one works with. But deciding how much to put into each channel requires a view the platforms don’t have: what each one adds on top of the others. That’s what incrementality measures.
Why this gap is a decision problem, not only a reporting problem
The ROAS gap isn’t only a measurement issue, because budget follows the numbers, and inflated numbers pull budget in the wrong direction.
If every platform looks profitable in its own report, the natural move is to spend more in all of them. Budget gets allocated by who reports best, not by who contributes most. Our view is that this is one of the most expensive habits in media buying, and it’s structural: each walled garden optimizes and reports toward itself, as we explain in walled gardens in advertising. Someone has to decide across them, against one shared KPI.
How Mainkore approaches the gap
Mainkore is autonomous advertising decision intelligence that decides across the channels it’s connected to against the advertiser’s KPI, not against each platform’s own objective.
It reads platform signals as inputs, alongside the rest of its 200+ variables per decision, and allocates between channels as one connected decision. How the KPI it commits to is defined and measured is covered in how a guaranteed KPI is measured.
FAQ
Because each platform attributes conversions it can see to itself, with its own rules and windows. When several platforms reach the same customer, the same sale can be counted more than once.
No. It’s useful for optimizing inside each platform. It’s a poor basis for deciding how much budget each channel should get.
Your own KPI measured on your side, such as sales, revenue or qualified leads, supported by incrementality measurement to see what each channel actually adds.
Measurement and accountability are among the eight criteria in our media buying models guide. Want to see how your budget would move against one shared KPI? Talk to us.
Mainkore. One KPI, yours, across every channel.
