Recovering Lost Conversions After iOS 14 Update
Recovering Google Ads Conversions After iOS 14

๐Introduction
An ecommerce brand running significant Google Ads spend saw reported conversions drop by 40% after the iOS 14 update, despite sales remaining stable according to their Shopify backend.
โThe Problem
iOS 14 ATT prompt caused the majority of iOS users to opt out of tracking, making browser-side conversion pixels unable to attribute purchases. Google Ads was under-reporting conversions and smart bidding was cutting spend on campaigns that were actually performing well.
๐Identifying the Causes
90% of the brand's audience was iOS mobile users. Their conversion tracking relied entirely on browser-side gtag.js, which was blocked by ATT opt-outs. No server-side tracking or Google Enhanced Conversions were in place to fill the signal gap.
โ ๏ธConsequences for the Business
Google Ads smart bidding interpreted the conversion drop as real performance decline and reduced bids on top-performing campaigns. CPA appeared to rise 60% overnight, causing the client to pause campaigns that were actually profitable.
โ Solution
Implemented Google Ads Enhanced Conversions using hashed email addresses and phone numbers captured at checkout. Set up server-side conversion import from Shopify order data via Google Ads API. Added Consent Mode v2 to model conversions from non-consenting users.
๐Results
Reported conversions recovered to within 12% of actual Shopify orders within 3 weeks. Smart bidding re-calibrated and CPAs returned to pre-iOS 14 levels. ROAS improved by 28% as previously paused campaigns were reactivated with accurate signals.
๐Conclusion
iOS 14 fundamentally changed conversion measurement. Browser-side only tracking is no longer sufficient. Server-side + Enhanced Conversions is now the baseline requirement for accurate Google Ads measurement.
๐กKey Takeaways
Enhanced Conversions require collecting hashed customer data at point of conversion. Server-side conversion import provides the most reliable signal. Consent Mode v2 is essential for modeling and recovering lost data in GDPR markets.