Ad platforms apply different attribution windows, counting methods, and statistical models, producing conversion counts that overlap and do not sum to actual order counts. This is structural, not a bug: each platform can only see its own data and applies a methodology that maximizes its own reported contribution.
The Incentive Problem
Each platform defaults to the most generous attribution window it can justify:
- Google Ads defaults to a 30-day click attribution window
- Meta uses 7-day click and 1-day view (reduced after iOS 14.5)
- LinkedIn defaults to 30-day click and 7-day view
When a customer clicks a Google ad on Monday, sees a Meta ad on Wednesday, and converts on Friday, all platforms that touched the journey claim credit. The conversion is counted multiple times across dashboards. Each platform can only see its own data; the overlap is invisible inside any single dashboard.
Beyond Double-Counting
Ad spend cannot be joined with actual revenue inside platform UIs. Each platform’s self-attributed ROAS is biased toward its own contribution and does not account for margins, customer lifetime value, or post-conversion events like refunds.
What Becomes Possible in the Warehouse
The data warehouse is the only place where all channels meet on neutral ground. Once ad data from every platform sits alongside CRM, ecommerce, and web analytics data, several analyses become possible that are not possible inside platform UIs.
Cross-channel reporting: instead of manually combining numbers across platform dashboards, all channels are queryable from a single table.
True ROAS: joining ad spend with actual revenue from a CRM or ecommerce platform produces a consistent revenue denominator across all channels, rather than each platform’s self-reported figure.
Multi-touch attribution: combining click-level ad data (via gclid and UTM parameters) with web analytics events and CRM conversions in one environment makes it possible to assign credit across touchpoints using first-touch, last-touch, linear, or data-driven models.
Budget optimization: with all spend and performance data in a single environment, reallocating budget across channels becomes a data query rather than a manual spreadsheet exercise.
LTV analysis: connecting acquisition source to long-term customer value requires joining ad data with CRM data — an operation only possible in the warehouse.
The Blended ROAS Principle
Platform-specific attribution data should not be compared directly across channels. Each platform uses different windows, counting methods, and statistical models; reconciling their self-attributed conversions produces misleading results.
Blended ROAS — total revenue divided by total ad spend — sidesteps the attribution disagreement. It asks what revenue was earned per dollar spent across all channels, without requiring agreement on per-channel credit.
UTM-based attribution models (built from warehouse-level attribution queries) use a consistent methodology across all touchpoints and are more reliable than any single platform’s self-reported numbers.
Scale
The average marketing department uses 91 tools. More than half of marketing leaders report disappointment with their analytics output. The global data warehousing market reached $34 billion in 2024 and is projected to more than double by 2033, driven in part by marketing analytics needs.