The attribution problem
Your Meta pixel reports that 68% of revenue comes via Meta ads. Google Ads claims 45%. TikTok claims 22%. Add those up and you get 135%, more than all of your revenue. Every ad platform claims the conversion as its own success, even when the customer touched multiple channels. That is the classic attribution problem.
What is attribution reconciliation?
Attribution reconciliation compares two data sources. Pixel-attributed data is what ad platforms report based on clicks and views. Customer-declared data is what customers themselves say about how they found you. The second source comes from post-purchase surveys, a short question after purchase asking how they found you. That is the only source that cannot be manipulated by ad platforms.
What does the data show?
We consistently see large differences between pixel-attributed and customer-declared data.
- Meta ads are overclaimed by 30 to 40% on average by the pixel.
- Word of mouth is completely invisible in pixel data, but customers mention it in 15 to 25% of cases.
- Organic search is underestimated. The pixel sees the last click, not the research phase.
- Direct traffic is underestimated. Customers who navigate directly after seeing an ad earlier do not get credited to that ad.
Why do ad platforms overclaim?
Ad platforms have a financial interest in claiming as many conversions as possible. Meta uses a 7-day click plus 1-day view attribution window by default. If someone saw your ad (a view, not even a click) and purchases within 24 hours, Meta claims that conversion, even if the customer came through Google after the Meta view.
A concrete example
A Benelux fashion brand with €50,000 per month ad spend ran attribution reconciliation. Meta pixel showed 68% of conversions, surveys showed 31%. Organic search was 9% in the pixel and 28% in the survey. Word of mouth was 0% in the pixel and 19% in the survey. Nearly 1 in 5 customers came via word of mouth, and the brand invested €0 in referral programs. After shifting €8,000 per month from Meta to SEO and referrals, measured return improved.
Layer Insights shows pixel-attributed channel data and customer-declared survey data side by side. We integrate with KnoCommerce, Fairing, Zigpoll, and Grapevine for survey data, and combine this with pixel and CAPI data for the attribution side. You see where the pixel overclaims, which channels are underestimated, and where budget actually returns.