In our previous look at the new rules of Meta advertising, we introduced you to Andromeda. This unified AI engine has changed how your Facebook and Instagram ads find your customers. We discussed how ad creative has become the primary lever for targeting and why a simplified account structure is now essential.
But as we move deeper into 2026, simply knowing about Andromeda isn’t enough. To outpace your competition, you need to understand how to feed this engine the right data and creative signals to maximise your return on advertising investment.
Shift from keywords to personas
Previously, when setting up ad campaigns in Meta, marketers spent a lot of time defining interest groups. Interest groups allow advertisers to reach users based on their activities, hobbies, likes, shopping habits, and behaviours. Today, Andromeda’s retrieval engine is far more interested in the intent behind a user’s behaviour. As detailed in the Andromeda Playbook, the engine is now capable of predicting not just who’s interested in a product, but who’s ready to purchase based on real-time signals across the Meta ecosystem (Facebook, Instagram, Messenger, and its Audience Network).
That means your marketing should focus less on broad demographics and more on customer personas. For example, if you’re a property staging company, don’t just target real estate. Create content that speaks specifically to the challenges of maximising value for a deceased estate or styling a modern townhouse for the young professional market. Andromeda will recognise these nuances in your copy and video, then find the exact users currently in those specific life stages.
Fuel the engine with quality data
Andromeda is only as good as the information you give it. While the AI learns faster than ever, it still relies on high-quality feedback loops. This is where many brands are currently falling behind.
To make Andromeda work well, you should prioritise these three things.
The feedback loop
Ensure your conversions are being tracked accurately. If the AI doesn’t know a sale happened, it can’t find more people like that buyer.
Diversified creative
Since Andromeda relies on your ad creative to identify your audience, uploading only one medium, such as images, limits your potential reach. You need a mix of professional brand videos, authentic user-generated content, and clear, benefit-driven graphics to help the AI test different segments of your market.
Privacy-first signals
With Australian privacy standards evolving, relying on tracking pixels is no longer a complete strategy. Implementing Meta’s Conversions API (CAPI) allows you to share data directly from your server, giving Andromeda a clearer, more reliable picture of your success.
Move to a content-led ad strategy
The technical barriers that once defined digital advertising are starting to vanish. However, while anyone can set up a broad campaign and let the AI run, not everyone can do it effectively. While launching a campaign may seem more accessible, achieving influence and a positive return on investment has become significantly more challenging.
Yes, Andromeda is a powerful tool, but it lacks a human heart. It can identify the person most likely to click, but it can’t explain why your brand matters or why a client should trust your company over a cheaper competitor. Automation can handle the delivery logistics, but it can’t craft the soul of your message.
To succeed in this landscape, your strategy must pivot from technical execution to high-level storytelling. By diversifying your creative—moving between polished brand narratives that showcase your aesthetic authority and raw, authentic customer experiences—you give the AI the human context it needs to build a genuine connection.
