Beyond the Last Click: Rethinking Meta Creative Performance

The core idea here is that Meta’s Andromeda update has decoupled targeting from audience settings, moving it entirely into the creative itself. Most advertisers struggle because they optimize for last-click conversions, a metric that hides the multi-stage journey a prospect takes before buying. By focusing only on the final sale, teams often kill the ads that build brand trust and educate the customer. The strategic advantage now lies in creative diversification and leading indicators, such as hook and hold rates, which show how the system is actually consuming your messaging. For CMOs and agency owners, the goal is to stop tinkering with granular audience segments and start trusting the algorithm to route diverse creative to the right users over time.

The Hidden Cost of Fast Solutions

Modern media buying is plagued by the tinkering reflex. When an ad does not show immediate conversions, the standard response is to pause it. Thomas McNaught argues that this is a mistake. In the current Meta ecosystem, ads act as a cohesive system; some are designed to educate and hook, while others capture the final sale. By killing ads that lack immediate conversions, you starve the top of the funnel that feeds bottom of the funnel success.

You can have an ad that is doing amazing to get you that top of the funnel traffic is feeding all the other ads and just because it does not have a lead because it was not the last attributed ad. You could be making a very bad decision by switching that off.

-- Thomas McNaught

The system requires patience. McNaught notes that leaving ads untouched for three to four weeks allows the algorithm to learn the sequencing required to move a prospect from awareness to action. This creates a competitive moat: most teams lack the discipline to wait, meaning those who do allow the algorithm to stabilize their performance will consistently outperform those who constantly optimize their way into failure.

Creative as the New Targeting

The era of carpet bombing with hundreds of ad sets is over. Meta’s Andromeda update has shifted the burden of targeting from the advertiser to the creative itself. The algorithm now reads the creative to determine the audience, meaning the primary lever for performance is no longer interest-based targeting, but the diversity of your messaging.

The targeting is done for the creative mainly in what we are seeing in the creative of what we are shown and how we are diversifying that message... so that is the sort of split that we are looking at.

-- Thomas McNaught

This shift requires a change in how performance is measured. Instead of looking at a single conversion metric, the team at Tier 11 uses a hit rate system. By identifying ads that rank in the top tier for hook rate, hold rate, watch time, and click-through rate, they can identify which creative assets are actually driving the system, even if they are not the ones getting the final credit.

Where Immediate Pain Creates Lasting Moats

The most non-obvious dynamic discussed is the use of AI-generated creative in sensitive sectors like healthcare. While many marketers view AI as a cheesy shortcut, McNaught highlights its utility in anonymizing patient stories while maintaining human-like emotional resonance. The system does not care if the creative is AI-generated; it cares if the creative provides a good user experience. If the hook and hold rates are high, the algorithm rewards the advertiser with lower CPMs.

This creates a paradox: the discomfort of adopting AI creative or the patience required to let ads run without interference creates a lasting advantage. Most competitors will continue to focus on manual audience management and last-click metrics, leaving the high-leverage, algorithmic-friendly strategies to those willing to embrace the complexity of the new system.

Key Action Items

• Audit your hit metrics: Stop evaluating ads solely on last-click conversions. Over the next month, identify your top 25% of ads by spend and rank them by hook rate, hold rate, and watch time to see which assets are actually assisting your funnel.
• Implement a hands-off policy: For new creative sets, commit to a 3-4 week testing window where no manual adjustments are made. This allows the algorithm to learn the audience sequencing without being reset by constant tinkering.
• Diversify creative formats: Ensure your ad account contains a mix of video, static, and carousel assets. Use video for top-of-funnel education and images for bottom-of-funnel conversion.
• Prioritize lead quality over volume: If budget is limited, optimize for the highest-intent action you can afford to track, rather than chasing high-volume, low-quality leads that do not convert to revenue.
• Re-evaluate AI creative: If you are in a sensitive industry, use AI-generated animation to anonymize storytelling. Focus on the feeling of the creative rather than the production method; if it holds the audience attention, the system will reward it.
• Shift from targeting to messaging: Stop trying to build complex audience segments like 180-day retargeting. Instead, build audience segments through your creative messaging and let the algorithm do the heavy lifting of distribution.