C
Curly Horse
Advertisers who run a single pixel across multiple Meta ad accounts and multiple target markets need to measure ROAS for one market in isolation, for example US-targeted campaigns only, with both spend and attributed revenue scoped to that market. Today that cannot be done in a single view.
Spend can be filtered by campaign targeting geography (target_country). Attributed revenue cannot: target_country is a Facebook ad-platform dimension and does not exist on the Pixel attribution model, so it has no effect on Pixel Conversion Value. The only geography available on the revenue side is buyer location (shipping or billing country), which answers "where were my buyers" but not "which campaigns' targeting drove this revenue." Users can therefore filter spend by campaign geography and revenue by buyer geography, but cannot produce a clean per-market ROAS that ties both sides to the same campaign-targeting definition within one account.
Proposed solution
  • Provide a supported way to scope attributed revenue / Pixel Conversion Value by campaign targeting geography, so per-market ROAS can be built in one view.
  • Design consideration: the only campaign-geography signal available today is ad delivery country, which is imperfect, since many campaigns deliver to several countries and not every attributed touchpoint carries a resolvable campaign identifier. A naive "target country on revenue" would over-count multi-country campaigns and under-count uncredited touchpoints. Ideally support a clear notion of a campaign's intended target market, or make the delivery-based limitations explicit, so the resulting figure is trustworthy.
  • As a guardrail regardless of the above: when a filter dimension does not exist on a metric's source model, fail loud rather than silent. Surface an error, or grey the dimension out as unavailable for that metric, instead of returning unfiltered data that looks filtered.
Created by kellie@polaranalytics.co