At a Glance

  • Meta’s Andromeda engine is powering a new era of ad personalisation through SKU-level targeting.
  • “Genome Ads” align ad delivery with individual preferences, not just broad segments.
  • Clean product feeds, modular creatives, and strong first-party data are now performance essentials.
  • Brands already using these tactics are seeing 20–30%+ gains in ROAS and sharper attribution.
  • If you’re not feeding the AI the right ingredients, you’re not getting the right results.

We’ve all seen the shift — ads today feel less like billboards and more like curated suggestions. What used to be broad interest targeting is now laser-focused at an almost eerie level. One shopper sees olive-green running shoes in their exact size. Another, browsing the same site, gets hiking boots for trail terrain. The magic behind this? A new class of AI-powered campaigns we’re calling “Genome Ads.”

Much like a genome uniquely defines a person, these ads match the exact product to the individual based on behaviour, preferences, and contextual data. And thanks to new AI infrastructure like Meta’s Andromeda engine, it’s no longer a future trend — it’s the current playbook for high-performing ecommerce advertisers.

Let’s unpack what makes this shift so powerful, and what you need to do to stay ahead of the curve.

Part of the Series: Intelligence-Led Advertising

This blog is part of our Intelligence-Led Advertising series – a 3-part exploration of how AI is reshaping the way campaigns are planned, personalised, and perceived. From dynamic creative to real-time persuasion engines, we dive into what happens when data moves from predictive to adaptive – and what marketers must consider as intelligence scales.

Explore the full series:

Follow the Person, Not the Product Feed: How “Genome Ads” Are Redefining Personalisation
A deep dive into the shift from product-led targeting to person-led customisation — and how hyper-personalised ads are getting rewired.

AI Isn’t Just Predictive — It’s Persuasive: Designing Campaigns That Adapt in Real Time
Learn how AI now powers adaptive creative, predictive messaging, and continuous learning loops across ad campaigns.

The Ethics of AI Advertising: How Far Is Too Far in Personalisation and Persuasion?
As ads get smarter, the ethical line gets blurrier. This post tackles the risks, responsibilities, and real limits of AI-powered persuasion.

From Personalisation to Precision Matching

Personalised ads are nothing new. What is new is the level of precision. We’re moving beyond demographic slices or retargeting cart abandoners. Genome Ads predict what someone wants before they’ve even asked for it — down to the SKU.

Meta’s Andromeda engine is the AI superbrain making this possible. It runs in the background of your Facebook and Instagram campaigns, analysing billions of potential ad combinations in milliseconds to figure out exactly which product is likely to convert for which person. It’s the engine powering Advantage+ campaigns and Shop ads, where Meta’s AI learns from user behaviour, creative interactions, and first-party signals to serve the best match.

The result? More relevance, better engagement, and seriously improved campaign performance. Meta’s internal data shows advertisers using Advantage+ creative features saw an average 22% lift in ROAS—not from changing their strategy, but from letting the AI choose smarter.

It’s Not About More Ads. It’s About Smarter Ones.

What’s important here isn’t scale – it’s relevance. You don’t need 10,000 ads. You need 10 good creatives the algorithm can adapt for 10,000 different people.

To do that, you’ve got to give the AI the right building blocks:

  • Rich product feeds – These are the raw ingredients the system uses. Poorly tagged SKUs, missing metadata, or clunky category structures can stop the AI from choosing your product at all. Your feed isn’t just a checklist; it’s the foundation of your campaign’s intelligence.
  • AI-ready creative assets – One static image won’t cut it. Think of your creatives as a kit: multiple angles, backgrounds, formats, and copy versions that the platform can mix and match based on what works best for each user. The more modular, the more powerful.
  • First-party data – Your customer behaviour, site interactions, purchase history… this is gold for ad engines. Feeding it into Meta (via tools like Conversions API) gives the system better signals and more context to personalise effectively — especially in a cookieless world.

Why This Matters for Your Bottom Line

Let’s be clear – this isn’t just about clever tech. It’s about results.

Early adopters are already reporting 20–30%+ improvements in ROAS, lower CPAs, and sharper attribution clarity. Brands that implement clean data feeds, AI-optimised creatives, and server-side tracking are not only seeing better campaign efficiency, but also uncovering performance they were previously blind to.

We’ve even seen cases where brands thought campaigns were underperforming – only to realise, once they piped data properly, that they were actually crushing it. When you give the AI the right inputs, it starts doing the heavy lifting.

Don’t Fall Behind — Future-Proof Now

This is the direction the industry is headed. Meta, Google, TikTok, Amazon — they’re all building toward more granular, behaviour-led ad delivery. And with AI models only getting smarter, the gap between brands that embrace “Genome Ads” and those sticking to static campaigns is going to widen fast.

So ask yourself:

  • Are you feeding your campaigns the right data?
  • Are your creatives modular and adaptable?
  • Are your tracking and signals set up to capture the full picture?

If not, now’s the time to act. The future of advertising isn’t about chasing trends—it’s about preparing your foundations so your campaigns can evolve intelligently.

Next in the Series:

AI Isn’t Just Predictive — It’s Persuasive

We’ve mastered product-to-person matching — now it’s time to rethink messaging itself. In our next post, we explore how AI doesn’t just predict what customers want, but influences what they do, crafting personalised journeys that adapt in real time.