The direct-to-consumer model that relied on heavy ad spend is broken. Apple's iOS 14 update made ad targeting ineffective by removing access to third-party data, while Shopify flooded the market with competitors, making customer acquisition costs unsustainable.

Judging marketing on a daily spend vs. daily return basis is a major error. Data shows a typical purchase cycle is 3 weeks to 3 months. This time lag, not a drop in ad effectiveness, is why ROAS appears to dip when you ramp up spending. Align your measurement with this reality.

Cookie deprecation blinds ad platforms like Google and Meta to on-site conversion quality. Marketers can gain a significant performance edge by creating a feedback loop, pushing their attributed first-party data (like lifetime value and margins) back into the platforms' AI systems in near real-time.

New measurement tools are moving beyond probabilistic models (guessing based on IP/device) to deterministic view-through attribution. By using first-party data like platform logins, marketers can now directly match an ad impression to a purchase, solving a major measurement challenge.

Many marketers mistakenly use attribution models for precise instructions. Instead, they should be used directionally to understand which channels are generally performing better, without treating the data as absolute truth that dictates every specific action.

A common attribution error is assigning all sales to paid marketing activities. In reality, most brands have a strong "baseline"—sales that would occur even without marketing. Accurate measurement requires modeling this baseline first, then attributing only the incremental lift from campaigns.

To move beyond last-click attribution, small businesses should add a simple metric to their daily tracking: impressions. By analyzing the relationship between impression spikes and the subsequent rise in clicks days or a week later, they can start to see the true top-of-funnel drivers of their business, revealing which channels are building crucial initial awareness.

Ben Thompson posits that Apple's platform restrictions on iOS were a blessing in disguise for Facebook. Prevented from building a true platform, Facebook was forced to double down on being an app, leading to the perfection of its highly lucrative, full-screen advertising model.

Direct attribution models are flawed because platforms like Google and Facebook use tracking pixels to claim credit for sales that would have occurred anyway. Smart marketers are returning to older methods of measuring lift from campaigns rather than relying on misleading platform data.