Meta's advertising platform continues to evolve at pace, with significant changes to targeting capabilities, ad formats, and campaign objectives rolling out regularly. For Australian businesses investing in Facebook and Instagram advertising, staying across these changes isn't optional — strategies that were effective six months ago may now be underperforming, not because the market has changed, but because the platform has.
Advantage+ and the Shift Toward AI-Driven Campaigns
Meta has been aggressively pushing advertisers toward its AI-driven campaign structures, particularly Advantage+ Shopping Campaigns for e-commerce and Advantage+ Audience for broader targeting. These products use machine learning to automate many of the decisions that advertisers previously made manually — audience selection, creative rotation, placement, and bid strategy.
The results have been mixed for Australian advertisers. For businesses with large product catalogues and significant conversion data, Advantage+ Shopping has often delivered improved ROAS compared to manually structured campaigns. For smaller businesses with limited conversion data, the AI has less to work with and can underperform well-structured manual campaigns. The key variable is the volume and quality of conversion data you're feeding the platform — the more it has, the better it performs.
Australian businesses should be testing these formats, particularly if they haven't already, but with realistic expectations. Run them alongside your existing campaigns rather than replacing everything at once, and give the AI sufficient time (and budget) to learn before drawing conclusions.
Privacy Changes and Audience Targeting
The evolution of privacy regulations and Apple's iOS tracking changes continue to affect Meta's targeting precision. Remarketing audiences are typically smaller than they were three years ago, and measurement data can be incomplete or delayed. This doesn't make Meta advertising ineffective — it remains one of the most powerful platforms for Australian businesses — but it requires a different approach to both strategy and measurement.
Businesses that have built strong first-party data assets — email lists, phone number databases, customer lists — are better positioned to weather these changes. Custom Audiences built from first-party data consistently outperform broad interest-based targeting in the current environment. If you haven't been growing your email list as a marketing priority, the Meta privacy landscape is a strong argument to start.
Creative Strategy in 2025
Meta's own data consistently shows that creative is the single biggest variable in ad performance — bigger than targeting, bidding strategy, or campaign structure. In a world where the platform's AI is handling more of the optimisation, the creative becomes even more important as the remaining lever in your control.
Short-form video continues to dominate Meta's engagement metrics. Reels-format creative — vertical, punchy, with key information in the first three seconds — significantly outperforms static images for many Australian advertisers. User-generated style content (authentic, slightly rough, shot on a phone) often outperforms highly produced creative, particularly for direct-response campaigns where the goal is conversion rather than brand building.
The practical implication for Australian businesses is to invest more in creative variety and testing. Produce multiple creative concepts per campaign and let Meta's system determine what resonates with your audience. The winning creative from those tests gives you insight into what messaging and formats your specific audience responds to.
Measurement in a Post-Cookie World
The gap between what Meta's platform reports and what you actually see in your CRM or website analytics has widened as tracking limitations have increased. Australian businesses should use Meta's Conversions API to send conversion data directly from their server rather than relying solely on the pixel — this typically improves reporting accuracy significantly and gives Meta better data for its optimisation algorithms.
For campaign evaluation, use a combination of platform-reported data, your own CRM data, and incrementality tests rather than relying on any single attribution model. The goal is a realistic picture of what's actually working, which may look different from what the platform reports.