What is Meta Andromeda, and how does it impact your advertising strategy?
If you’re a Meta advertiser, you’ve surely heard the growing chatter around Meta’s new machine-learning system for ad personalization. But you’d be forgiven if you’re not sure what that actually means for your advertising.
The truth is that the official documentation about Meta Andromeda is convoluted and technical. But it would be a mistake to dismiss what Andromeda means for the present and future of Meta advertising.
At the same time, advertisers have learned that it’s smart to stay skeptical. Meta’s announcements often sound polished and futuristic, while the real-world rollout tends to be imperfect and messy.
In this post, I’ll cut through the technical details and highlight what advertisers actually need to know about Andromeda:
- The technical details of Andromeda
- Creative diversification and your ad strategy
- What it means for campaign construction
- How it affects performance
Here’s what Andromeda is, why it matters, and what it means for how you run ads today…
The Technical Details of Andromeda
Meta defines Andromeda in technical terms that sound impressive but aren’t exactly advertiser-friendly. Here’s how Meta defined it in a March 2025 blog post:
Meta Andromeda is an innovative end-to-end hardware, software, machine learning co-designed system introduced in 2024, with Meta Training and Inference Accelerator (MTIA) and NVIDIA Grace Hopper Superchip. This more efficient system enabled a 10,000x increase in the complexity of models used for ads retrieval, the first step in the ranking process where we narrow down a pool of tens of millions of ads to the few thousand we consider showing someone.
If your eyes glazed over reading that list of hardware and software terms, you’re not alone. This isn’t going to mean much to most people.
At its core, Andromeda is about retrieval, the first step of ad delivery where the system decides which ads have a chance to be shown. Andromeda doesn’t rank the winning ads, it picks the short list.
Meta uses an analogy to explain it further:
Imagine having a personal concierge who knows your tastes so well that they don’t just understand that you covet shoes, but that you like to wear red flip flops at the beach.
On one hand, this doesn’t sound all that different from how I envisioned ad audience selection worked. In years past, I’ve used the example of engaging with an ad for a dress shoe and then getting hammered with ads for a similar style of shoe in my feed.
Meta’s analogy leans less on the recent engagement that required such an experience and more on “knowing you.” Andromeda is supposed to move beyond broad categories (“you like shoes”) and toward finer personalization (“you like these shoes in this context”).
Why did Meta need this? Meta says retrieval is the bottleneck.
That first step is massive. Meta says retrieval under Andromeda processes hundreds or even thousands of times more ads than the later stages. The old system couldn’t handle the surge of new ad variations created by Advantage+ and generative AI. Andromeda is their answer to that scale problem.
Bottom line: Andromeda is Meta’s new retrieval engine. It takes millions of ads, narrows them to a few thousand candidates, and does so in a way Meta claims is faster, smarter, and more personal than before.
Creative Diversification and Your Ad Strategy
So, Meta’s hardware is faster and smarter. Great. But the key for advertisers is if and how this impacts their approach to ad creative.
It does. From Meta’s April 2025 blog:
With the rise of AI-enabled advertising tools, the focus has shifted from niche targeting to creative diversification as the best lever to find the most relevant audiences.
So, what does “creative diversification” mean? It’s not about pumping out near-duplicates with minor tweaks. It’s about creating a range of ads that differ in meaningful ways.
Meta links this shift directly to Andromeda:
As businesses upload more and more creatives to support a diversification strategy, Meta Andromeda works behind the scenes to power more complex models that allows Meta to pick the right creative to deliver more personalized ads that are relevant and interesting.
In other words, Andromeda is designed to handle the explosion of creative options. The more meaningful variety you provide, the more opportunities Andromeda has to connect the right message with the right person.
Think about it this way: If you only provide a single product demo video, Meta can only try that one angle. But if you create ads that lean into different pain points and customer personas, the system can learn who responds to each and deliver ads that feel more personal.
Creative diversification shows up in a few key areas:
- Concepts and Angles (Problem/solution, pain points, testimonials, product demos)
- Formats (Short video, long video, static image, carousel)
- Personas (Speaking to different customer types)
Creative diversification means building ads for different people and contexts, not just different versions of the same idea. Instead of being married to a single approach, you provide Meta with the variations Meta needs to match the right creative to the right person.
What it Means for Campaign Construction
If you’ve been following my content during the past several months, this development fits perfectly with my recommendation of a simplified approach to advertising. It reinforces the following…
1. Stop chasing the perfect combo of ad copy and creative. There isn’t one. The “winner” for one person may flop for another. Andromeda personalizes delivery so the right person sees the right variation.
2. Targeting largely happens in the ad now. Instead of obsessing over age, gender, or interest levers, focus on ads that speak to the people you want. Keep ad sets broad and let Andromeda match.
3. Use Advantage+ defaults. Placements and delivery automation are built to work with this system. In most cases, Advantage+ campaigns are the cleanest fit.
4. Simplify your structure. Because you do these three things, there are fewer reasons to overcomplicate campaign construction. You need a different campaign for each goal, but you may not need more than that. You will rarely need multiple ad sets for targeting purposes. And you’ll rarely need multiple ad sets just for targeting. The heavy lifting happens at the ad level.
The bottom line: Your time should be spent creating diverse, thoughtful ads. Campaign construction gets simpler while your creative strategy gets deeper.
How it Affects Performance
Will Andromeda improve results? The honest answer: It depends. While Meta says Andromeda has yielded an 8% increase in ads quality, don’t expect miracles.
Look at Andromeda the way you would any other hardware or feature. It should improve performance overall, but it won’t guarantee results.
Better personalization should mean stronger connections. When the system has more creative diversity to work with, it’s more likely to show someone an ad that feels relevant. That can lead to higher engagement and better conversion rates.
But, performance relies on the ads you provide. If you’re not getting results, it’s not because “Andromeda doesn’t work.” It’s because you haven’t successfully convinced your potential customers to convert.
The main thing to understand is that while Andromeda has the potential to make ads feel more personal at scale, your creative and strategy still decides whether that potential turns into performance.
What About Ad Limits?
There was a time when Meta recommended having no more than six ads in an ad set. They said that after six, there was no clear benefit.
Back in February, this recommendation disappeared. While Meta never came out and said it, this is very likely tied to Andromeda.
I keep hearing of advertisers finding success with far more than six ads now, and it’s not just those with bigger budgets. I’ve experimented with this as well, and I even have an ad set with 50 ads in it.
Whether this “works” is less about the number of ads you provide and more about creating the best possible ads you can. While I’d recommend moving beyond a two or three ad minimum, you’re not guaranteed results by creating 50. Your goal is to provide enough creative diversity, while also providing effective ad concepts, so that you can reach a diverse group of willing customers.
Experiment!
My main recommendation is to take this as a new challenge. When it’s time to create a new campaign, take a fresh approach to ad creation. Instead of your same old approach, focus on creative diversity. That means much more than minor copy and creative tweaks.
You want Meta to have as many good options as possible. That means different concepts, angles, formats, and personas. Try it all and see what hits.
This could be a lot of work. It’s hopefully further motivation to move away from complicated construction in the campaign and ad set so that you can focus that valuable time on the ads themselves.
If nothing works, don’t blame Andromeda. Go back to the drawing board, ask why people aren’t responding to your ads, and try again.
Your Turn
Have you experimented with greater creative diversity in response to Andromeda? What results have you seen?
Let me know in the comments below!
