Meta Andromeda: How Meta’s AI Ad Engine Changes Facebook and Instagram Advertising
Meta Andromeda is the machine learning system Meta uses to retrieve and shortlist which ads could be shown to each person on Facebook, Instagram and its other apps. Meta describes it as its “proprietary machine learning (ML) system design for retrieval in ad recommendation” (Engineering at Meta). In practice, it shifts much of the targeting work from the advertiser to the algorithm and the creative.
Many advertisers have noticed that narrow audiences, heavy manual segmentation and a few polished ads no longer perform as they once did. This guide is for business owners and marketers running Meta ads, and it explains how Andromeda works, what it changes and the six adjustments I make to campaigns.
Quick Answer: What Is Meta Andromeda?
Meta Andromeda is Meta’s AI ad retrieval engine, which picks a shortlist of relevant ads for each person before the final auction. It works best with broad targeting, simpler account structures and many genuinely different creatives, because creative now signals who an ad is for. It affects every Facebook and Instagram advertiser. Start by auditing how varied your ad creatives are.
Want to know if your Meta account is set up for Andromeda? An AI Marketing Audit reviews your paid social structure and creative.
How Does Meta Andromeda Work?
Meta Andromeda works at the retrieval stage of ad delivery, narrowing millions of possible ads down to a small set worth ranking for each person. Meta’s later ranking models and the auction then choose the ad actually shown, based on predicted results and bid.
Meta reported that Andromeda delivered a 6% recall improvement and an 8% ads quality improvement on selected segments (Engineering at Meta). The practical effect is that the system can match far more ad variations to far more people, which rewards advertisers who give it a wide range of creative to work with.
What Does Andromeda Change for Advertisers?
Andromeda changes where advertisers create advantage. Detailed targeting and many small ad sets matter less, while creative volume, creative diversity and clean conversion data matter more.
| Area | Before | With Andromeda |
|---|---|---|
| Targeting | Narrow interests and lookalikes | Broad audiences, with Advantage+ doing more of the work |
| Account structure | Many ad sets split by audience | Fewer, consolidated campaigns and ad sets |
| Creative | A few polished variations | Many distinct concepts, formats and angles |
| Signals | Clicks and basic pixel events | Clean conversion data via Pixel and Conversions API |
Why I Treat Creative as the New Targeting
I treat creative as the new targeting because, with broad audiences, the ad itself tells Meta who it is for. Having managed and optimised more than 500 ad accounts, I have seen the same pattern repeatedly since these changes: accounts with five near-identical ads stall, while accounts testing genuinely different angles find new pockets of customers.
“Different” means different concepts, not different colours. A testimonial video, a founder explainer, a product demonstration, a problem-led static and a user-generated clip each speak to different people. That is also why I keep campaigns simpler than many advertisers expect: fewer ad sets give the system more data per decision.
6 Ways To Adapt Your Meta Ads for Andromeda
Adapting Meta ads for Andromeda means simplifying structure, broadening audiences, diversifying creative and strengthening data. These six changes apply to most accounts.
- Consolidate campaigns. Merge overlapping ad sets so each has enough conversions to learn.
- Broaden targeting. Test broad or Advantage+ audiences against narrow interest targeting.
- Build creative diversity. Run several distinct concepts across video, static, carousel and user-generated formats.
- Refresh creative regularly. Add new concepts every few weeks to avoid fatigue.
- Fix conversion tracking. Set up the Meta Pixel and Conversions API and optimise for real business outcomes.
- Judge results on business metrics. Measure cost per qualified lead or sale, not clicks or reach.
Want your Meta account restructured for Andromeda? See my social media marketing service.
Who Should Adapt for Meta Andromeda?
Every business advertising on Facebook and Instagram is affected by Andromeda, but the biggest gains go to advertisers with complex legacy account structures or limited creative.
Ecommerce Brands
Stores with large catalogues benefit from more creative angles and Advantage+ shopping campaigns.
Lead Generation Businesses
Service businesses gain when they optimise for qualified leads and feed back offline conversion data.
Agencies and In-House Teams
Teams managing many accounts need creative production processes that keep up with Andromeda’s appetite for variety.
Who It Is Not For
Businesses spending very small budgets may not generate enough conversion data for these changes to show clearly. Focus on one simple campaign and strong creative first.
Frequently Asked Questions About Meta Andromeda
Is detailed targeting dead on Meta?
Detailed targeting still exists, but broad and Advantage+ audiences often perform as well or better because Andromeda matches ads to people using creative and behaviour signals.
How many creatives should I run?
Run several genuinely different concepts per campaign and refresh them regularly. Variety of concept matters more than a fixed number.
Does Andromeda affect small advertisers?
Andromeda affects all Meta advertisers. Smaller accounts benefit most from simple structures that concentrate data into fewer ad sets.
What is the difference between Andromeda and Advantage+?
Andromeda is the underlying retrieval system, while Advantage+ is the set of automated campaign options advertisers choose. Advantage+ relies on systems like Andromeda.
What To Do Next
Meta Andromeda rewards simpler account structures, broader audiences, diverse creative and clean conversion data. Audit your creative variety first, then consolidate campaigns and fix tracking.
Want your Meta ads working with Andromeda, not against it? Talk to Crom about your Meta ads, or see my social media marketing service.
About Crom Salvatera
Crom Salvatera is a Sydney-based AI marketing consultant and Head of SEO, AEO and GEO with 22+ years in marketing and 14 in high-level digital. He created the TLC Method (Tech, Links, Content), has managed and optimised 500+ ad accounts and has helped generate $650M+ in revenue for employers and clients, with brand experience including LEGO, Hasbro and JB Hi-Fi. Connect with him on LinkedIn.
