PAID MEDIA · 9 MIN READ
Your MER is lying to you. Here’s how to fix the attribution gap.
Published 24 April 2026
Marketing Efficiency Ratio (MER) is the cleanest scale metric an ecommerce business has: total revenue divided by total ad spend. But the inputs driving the MER number, Meta ROAS and Google ROAS as reported in-platform, have been drifting further from reality every year since iOS 14.5.
Why reported ROAS lies
- Meta’s attribution window lost its teeth with ATT. Conversions are modelled, not measured.
- Google’s PMax takes credit for sales that would’ve happened anyway through brand search.
- Both platforms claim overlapping conversions, so your summed “platform ROAS” is structurally inflated.
The three-layer fix
Layer 1: Server-side tagging via GTM Server or Stape
Move your pixels server-side. You’ll recover 15 to 30% of conversion events lost to ad blockers and iOS privacy settings. This is the single highest-leverage change most brands haven’t made.
Layer 2: Conversions API for Meta and Enhanced Conversions for Google
Feed hashed customer data (email, phone, name) back to the ad platforms server-side. Match rates go from 40 to 50% up to 80%+. The optimiser gets better signal, CPAs fall, and reported ROAS gets closer to truth.
Layer 3: Incrementality testing
Run geo-based holdout tests quarterly. Turn ads off in one matched market for 14 days and measure the delta. This is the only way to know what your ads are actually adding on top of baseline demand.
What to report to the board
Stop reporting platform ROAS. Report blended MER, new-customer MER, and contribution margin after COGS and ad spend. The last one is the only number that tells you whether you’re actually building a business.
Book a free audit and we’ll show you how much signal you’re leaving on the table.