General

Meta's Andromeda Algorithm: The Biggest Facebook Advertising Change Since iOS 14.5

Jason Poonia Jason Poonia | | 5 min read
Meta's Andromeda Algorithm: The Biggest Facebook Advertising Change Since iOS 14.5

If your Facebook ad costs went through the roof in July 2025, you're not going crazy. You're experiencing the fallout from Meta's Andromeda algorithm update - the most significant change to Facebook advertising since the iOS 14.5 privacy update devastated tracking in 2021.

As someone who's been managing Facebook ads for New Zealand businesses since 2021, I can tell you this: Andromeda isn't just another algorithm tweak. It fundamentally changed how Meta's creative recommendation system works, and it caught almost every advertiser off guard.

Most businesses have no idea what happened to their campaigns. They just know that strategies that worked for years suddenly stopped delivering results. CPMs skyrocketed. Cost per result went through the roof. And all the tactics they normally used to fix underperforming campaigns made things worse.

If this sounds like your July 2025 experience, you need to understand what Andromeda actually changed and what you need to do differently going forward.

What Is Meta's Andromeda Algorithm?

Meta announced the Andromeda algorithm update on April 22nd, 2025, but most advertisers felt the full impact in July when it rolled out globally to all accounts.

Andromeda is now the first step in Meta's creative recommendation process - the system that decides which ads get shown to which people. This replaced their previous creative recommendation algorithm entirely.

Why Meta Made This Change

The reason for Andromeda is actually pretty straightforward: AI-generated content broke their old system.

Meta reported seeing 10,000 times more ad creatives than ever before once generative AI tools became popular. Their previous system couldn't handle the massive influx of similar-looking, AI-generated ads flooding the platform.

Andromeda was designed to better identify and reward genuinely diverse, high-quality creative content while filtering out the endless stream of generic AI-generated ads.

How Andromeda Changed Facebook Advertising Forever

The Old Creative Recommendation System (Pre-April 2025)

  • Upload a few ad creatives
  • Meta's algorithm found the best performing one
  • That "winning" creative got most of the reach
  • When it fatigued, you'd upload new creatives and repeat

The New Andromeda System (Post-April 2025)

  • Requires diverse messaging across all your creatives
  • No single creative dominates - algorithm rotates between different messages
  • Rewards ads that address specific customer situations
  • Punishes generic, one-size-fits-all messaging

What This Means in Practice

Before Andromeda: You could have 5 similar ads with slightly different hooks, and one would typically win and get most of the budget.

After Andromeda: You need 15+ completely different ads addressing different customer motivations, and the algorithm actively rotates between them to match specific audiences.

The Two Paths Forward (And Why One Is Much Better)

Based on our testing with clients across different industries and spend levels, there are two approaches that work with Andromeda:

Path 1: High-Volume Creative Production

This is Meta's official recommendation - one e-commerce company in their case study went from 3-4 creatives per week to 50+ per week.

The Numbers:

  • 50 creatives per week × 4 weeks = 200 new ads per month
  • Massive time and resource commitment
  • Higher creative production costs
  • Constant campaign management required

Who This Works For:

  • Large businesses with dedicated creative teams
  • E-commerce stores that can use Meta's AI creative tools effectively
  • Companies with significant advertising budgets (200K+ per month)

Path 2: Strategic Creative Diversity

This is what we recommend for most New Zealand businesses - create 15+ unique ads per month, each focusing on a different customer motivation or situation.

The Strategy: Instead of 15 variations of the same message, create 15 completely different messages addressing different reasons people buy from you.

Who This Works For:

  • Small and medium-sized businesses
  • Service-based companies
  • Businesses without huge creative production resources
  • Most New Zealand companies

Why Creative Diversity Beats High Volume

After testing both approaches with clients across different industries, we consistently see better results with the strategic diversity approach:

Better Performance Metrics

  • 30-50% lower cost per result
  • Higher quality leads and customers
  • Longer campaign longevity
  • Less time spent on campaign management

More Sustainable Operations

  • Requires strategic thinking, not just volume production
  • Builds deeper understanding of your customers
  • Creates reusable insights for other marketing
  • Doesn't burn out your creative team

Superior Algorithm Performance

Here's what's fascinating: with the diversity approach, we're seeing smaller, niche audience segments respond strongly to specific messages that historically would have been ignored.

Instead of one dominant ad reaching everyone poorly, we're seeing 3-4 different messages each connecting deeply with their ideal audience segment.

How Different Industries Should Adapt

E-commerce Businesses

Focus on different customer motivations for buying your products:

  • Quality/durability
  • Price/value
  • Convenience/speed
  • Status/lifestyle fit
  • Problem-solving
  • Trust/security

Service-Based Businesses

Address different customer situations that bring people to you:

  • Specific problems they're facing
  • Different stages of awareness
  • Industry-specific needs
  • Different service urgency levels
  • Various budget considerations

B2B Companies

Target different decision-maker perspectives:

  • Financial concerns (CFO messaging)
  • Operational efficiency (COO messaging)
  • Growth opportunities (CEO messaging)
  • Technical requirements (IT messaging)
  • Risk mitigation (Legal messaging)

The Technical Reality of Andromeda

What Changed Under the Hood

  • Andromeda uses multi-vector analysis to understand creative content
  • It maps ads to audience segments based on semantic similarity
  • The algorithm actively prevents any single creative from dominating
  • It rewards genuine diversity over superficial variations

What This Means for Campaign Performance

  • You'll see more even distribution of budget across creatives
  • Smaller audience segments will show stronger engagement
  • Campaign longevity increases with proper diversity
  • Cost per result improves as audience matching becomes more precise

Common Mistakes Advertisers Are Making

Mistake 1: Panicking and Changing Everything

Many advertisers completely rebuilt their accounts thinking they needed to start from scratch. Usually unnecessary and often harmful.

Mistake 2: Just Adding More of the Same

Creating 50 variations of the same basic message doesn't satisfy Andromeda's diversity requirements.

Written by

Jason Poonia

Jason Poonia is the founder and Managing Director of Lucid Media, helping NZ businesses grow online since 2018. With over 6 years delivering results for clients across New Zealand and internationally, Jason combines technical expertise with proven marketing strategies to help businesses attract more customers and build scalable systems. Background in Computer Science from the University of Auckland.