Meta advertising is becoming increasingly AI-driven. From audience discovery to creative optimisation and ad delivery, AI is taking on a larger role in deciding who sees an advertisement and when.
One of the biggest developments in this direction is Meta Andromeda, its AI-powered ad retrieval system.
On October 28, 2026, Meta announced the next generation of Andromeda, bringing significant improvements to how its advertising system identifies and selects relevant ads for users. The update highlights a shift towards more advanced ad matching, greater personalisation and the ability to evaluate a much larger pool of advertisements.
This development is being widely referred to as Meta Andromeda 2.0.
Let’s understand what Andromeda 2.0 means, what’s new in the latest update and how it affects businesses advertising on Facebook and Instagram.
Jump ahead to:
What Is Meta Andromeda 2.0?
Meta Andromeda 2.0 refers to the next-generation version of Meta’s AI-powered ad retrieval system. It is an industry-used name for the October 2026 update; Meta’s official engineering announcement calls it the next generation of Meta Andromeda.
To understand its role, it’s important to know how Meta decides which advertisements users see.
Every time someone opens Facebook or Instagram, Meta has a massive pool of advertisements that could potentially be shown to them. However, not every advertisement is relevant to every user.
Meta needs to identify which ads are worth considering for each individual before deciding their final placement.
What’s New in Meta Andromeda 2.0?
Meta’s October 28 announcement focuses on making its ad retrieval system more capable of evaluating advertisements and identifying relevant matches.
There are three important developments advertisers should understand.
1. Meta can evaluate a much larger pool of ads
Meta says the new system can process over one billion ads per second, representing a 100x increase in throughput compared with its 2024 Andromeda system.
What does this mean for advertisers?
Meta can evaluate a much larger pool of potential ads when deciding what might be relevant to a particular person. This becomes increasingly important as advertisers produce more creatives using AI tools and the number of ads on the platform grows.
2. Better understanding of individual user interests
The upgraded system is designed to understand different aspects of user interests, behaviour and context more effectively.
This means ad selection can become more personalised. For advertisers, the opportunity is better matching between the message in an ad and the person who sees it.
3. Better performance without requiring advertisers to do more
Meta says the upgraded retrieval system delivers better ad matching and conversion performance within the same processing time.
Importantly, Meta reports that Andromeda has delivered a 2.4% increase in conversions on Facebook since 2025. These are Meta-reported results, not a guaranteed improvement for every advertiser.
Meta Andromeda 2.0 vs. the Previous Andromeda System
The October 2026 announcement builds on the Andromeda system Meta introduced in 2024.
The original system was already designed to improve personalised ad retrieval. The latest update takes this further by improving the system’s ability to evaluate advertisements, understand user interests and operate at a much larger scale.
Here’s a simplified comparison:
| Area | Earlier Andromeda | Next-generation Andromeda |
| Ad retrieval | AI-powered ad selection | More advanced personalised retrieval |
| User understanding | Earlier user modelling | More specialised understanding of user interests and context |
| Ad evaluation | Evaluates eligible ad candidates | More sophisticated matching across a larger candidate pool |
| Processing capacity | 2024 system baseline | Approximately 100x higher production throughput |
| Matching quality | Retrieval-stage candidate selection | Matching quality approaching the ranking stage |
How does Andromeda 2.0 affect Meta advertisers?
This is where we should move from explaining technology to explaining marketing implications.
I recommend covering:
- Creative diversity matters: Different concepts, hooks, formats and messaging give Meta more meaningful creative options.
- Audience targeting is evolving: Advertisers should test broader audiences where commercially appropriate rather than relying on excessive segmentation.
- Campaign structures need careful evaluation: Don’t rebuild campaigns simply because of the update.
- Conversion signals remain important: Meta needs accurate information about the outcomes advertisers actually value.
Common Andromeda 2.0 misconceptions
| Common claim | What advertisers should understand |
| Every advertiser needs 20–30 creatives | There is no universal creative quota prescribed by Meta. |
| Audience targeting is dead | Targeting still matters, especially for business and geographical restrictions. |
| Similar creatives are automatically treated as one ad | Meta’s announcement does not establish this claim. |
| All campaigns need restructuring | There is no reason to disrupt campaigns that are already performing. |
| Andromeda guarantees better results | Platform-level improvements don’t guarantee individual campaign performance. |
How should lead generation marketers prepare for Andromeda 2.0?
Here is the five recommendations:
- Test different creative concepts: Focus on different customer pain points, motivations, objections and value propositions.
- Don’t overcomplicate audience structures: Give Meta room to find relevant users while maintaining necessary business constraints.
- Track the complete lead journey: Look beyond form submissions and understand which campaigns generate qualified leads, MQLs, SQLs and sales.
- Improve conversion signal quality: Send accurate CRM conversion events back to Meta using appropriate integrations such as Conversions API.
- Measure business outcomes: Evaluate cost per qualified lead, cost per opportunity and revenue contribution instead of relying exclusively on CPL.
Final Thoughts
Meta Andromeda 2.0 represents another step towards more AI-driven advertising.
Meta is making its advertising system better at understanding users, evaluating advertisements and identifying relevant matches from an increasingly large pool of eligible ads.
For advertisers, the takeaway isn’t to produce an arbitrary number of creatives, abandon audience targeting or rebuild existing campaigns.
Instead, it is an opportunity to improve the inputs that contribute to better advertising outcomes.
Focus on developing diverse creative concepts, maintaining relevant audience controls, tracking accurate conversion events and measuring campaign performance against actual business results.
To know more: Book a demo
Frequently Asked Questions About Meta Andromeda 2.0
1. Is Andromeda 2.0 an official Meta product?
No. Meta’s October 28, 2026 announcement refers to the next generation of Meta Andromeda. The term Andromeda 2.0 is commonly used to describe this update, rather than being the official name of a new Meta product.
2. What does Meta Andromeda do?
Meta Andromeda is an AI-powered ad retrieval system. It helps identify relevant advertisements from a large pool of eligible ads for individual users before those advertisements move to the next stage of ranking and delivery.
3. What is new in the October 2026 Andromeda update?
The update introduces a more advanced retrieval architecture, improved modelling of user interests and context, and significantly higher processing capacity. Meta reports approximately 100 times the production throughput of its 2024 system.
4. Does Andromeda 2.0 affect Facebook and Instagram ads?
Yes. Andromeda is part of Meta’s advertising recommendation infrastructure and supports personalised ad retrieval across its advertising ecosystem, including Facebook and Instagram.
5. Do advertisers need to change their campaign structure?
No mandatory campaign restructuring has been announced. Advertisers can continue using their existing campaigns while reviewing creative diversity, audience restrictions and conversion tracking.
6. Does Andromeda 2.0 make creative quality more important?
Creative quality and relevance remain important. The improved retrieval system can evaluate a larger pool of advertisements, but advertisers still need to develop compelling messages, relevant offers and creative concepts th
