January 28, 2026/
Master 21 proven client acquisition strategies for freelancers and agencies. Complete with outreach templates, platform recommendations, and conversion tactics.
A deep dive into Meta’s Andromeda algorithm, why Facebook ads performance has shifted, and how advertisers must adapt in the AI-driven era.
Quick Summary:
The Meta Andromeda update represents the most significant transformation in Facebook and Instagram advertising since 2022. Rolled out globally in October 2025, this AI-powered system fundamentally re-engineered how ads are selected, delivered and personalised across Meta’s platforms. Unlike previous updates focused on privacy or tracking, Andromeda shifts advertising success from precise audience targeting to creative diversity and AI-driven personalization.
Key Takeaways:
- Andromeda uses machine learning models 10,000 times more complex than the previous system
- Success now depends on creative variety rather than granular audience targeting
- Advertisers who adapted early report 8-17% conversion increases and up to 16% cost reductions
- Those using outdated strategies experience rising CPMs and declining performance
- The update requires simplified campaign structures with broader targeting and diverse creative assets
What is the Meta Andromeda Update?
What Does Meta Andromeda Actually Mean for Advertisers?
Meta Andromeda is a personalized ads retrieval engine that leverages advanced machine learning systems to enable cutting-edge innovation in the ads retrieval stage. In simpler terms, Andromeda represents the first major stage of Meta’s ad delivery process, filtering millions of candidate advertisements down to a few thousand eligible options for each impression opportunity.
The update was built to address a critical bottleneck in Meta’s advertising infrastructure. As generative AI tools became widespread, advertisers weren’t uploading 5 ads per campaign anymore—they were uploading 50, 100, or letting Meta’s tools generate hundreds of variations automatically. The previous system couldn’t efficiently process this explosion of ad variations.
How Has Meta Changed the Way Ads Get Delivered?
Before Andromeda, Meta’s advertising operated on a targeting-first model where advertisers controlled who saw their ads through interests, demographics, and behaviors. Andromeda’s intelligence doesn’t just ask “who should see this ad?” It now asks “which ad should this person see?”
This represents a complete paradigm reversal. Instead of grouping people into broad audience segments, Meta now predicts which specific ad each person is most likely to respond to based on intent signals, creative quality, and real-time engagement patterns.
What Technology Powers the Meta Andromeda Update?
Andromeda leverages the NVIDIA Grace Hopper Superchip to enable cutting-edge machine learning innovation in the ads retrieval stage. This powerful hardware infrastructure allows Meta to process hundreds of thousands of ads per user impression in real-time—something the previous system couldn’t handle.
According to Meta’s engineering documentation, the updated system is now 4 times more efficient at driving ad performance gains for a given amount of data and compute than its original ads recommendation ranking models.
- All Posts
- Digital Marketing
- Meta- Facebook & Instagram
- Back
- Digital Marketing Updates
- Lead Generation
- Client Acquisition
- Back
- Meta Algorithm Updates
January 21, 2026/
Meta’s Andromeda update is the biggest change to Facebook and Instagram advertising in years. Targeting is no longer the advantage...
January 18, 2026/
Do businesses really need SEO? This blog delivers a brutally honest answer. It explains why SEO is not about rankings...
What Has Changed with Meta Advertising After Andromeda?
Summary: How Has Facebook Advertising Changed in 2025?
The Andromeda update fundamentally transforms six critical aspects of Meta advertising: the retrieval process, targeting methodology, creative evaluation, campaign structure requirements, performance metrics, and the learning phase duration. Understanding these changes is essential for adapting your advertising strategy to the new reality.
How Does Meta’s New Ad Retrieval System Work?
The retrieval stage is where Andromeda operates, acting as an intelligent gatekeeper before ads enter the auction. Meta says retrieval under Andromeda processes hundreds or even thousands of times more ads than the later stages.
Previously, the system used limited rule-based logic to determine ad eligibility. Now, Andromeda replaces much of the rule-based logic of older systems with deep neural networks that analyze user behavior, eligible ads, and creative options in real time.
Does Audience Targeting Still Work on Meta Ads?
The shift from narrow targeting to broad audiences represents one of the most challenging adjustments for experienced advertisers. Meta itself now recommends advertisers adopt broad targeting, use Advantage+ placements, and allow the algorithm to make its own delivery optimisations.
Lookalike audiences haven’t disappeared, but their role has evolved. They now serve as useful signal inputs rather than hard boundaries. The algorithm learns from lookalike seeds but quickly expands beyond them, identifying similar users with comparable behavioral signals.
How Does Meta Evaluate Creative Similarity Now?
One of the most significant but underreported changes involves how Andromeda evaluates creative similarity. There are real cases where ads with entirely different people, scripts, and edits were still grouped as the same ad simply because they shared similar setups—like the same creator type or background environment.
The system analyzes visual patterns, not just text and audio. Meta considers ads similar if their first three seconds are the same, which means slight variations of the same concept no longer provide the diversity the algorithm needs.
What Campaign Structure Works Best After Andromeda?
Complex campaign structures that worked in 2024 now actively harm performance. Running 10+ ad sets with narrow targeting is the old playbook that no longer delivers stable conversions.
Simplified structures with fewer campaigns and broader targeting give Andromeda the freedom to allocate budget efficiently. Reducing segmentation helps the system learn faster and distribute spend based on real-time performance rather than pre-defined audience buckets.
What Performance Results Can I Expect from Andromeda?
Early adopters report mixed but ultimately positive results when properly implemented. When advertisers who did not previously use Advantage+ creative turned on its AI-driven targeting features, they experienced a 22% increase in ROAS.
For businesses leveraging generative AI features, Meta estimates that businesses using image generation are seeing a +7% increase in conversions.
How Long Does the Learning Phase Take Now?
The timeline for ad optimization has fundamentally changed. The old approach was to kill underperforming ads within 24-48 hours, but Andromeda’s learning phase works differently.
New guidelines suggest giving creatives 4-7 days with adequate spend before making elimination decisions. The algorithm needs this time to identify the right micro-audiences for each creative variation.
What Should I Focus on to Make My Meta Ads Work Now?
Summary: What’s the Most Important Thing for Meta Ads Success in 2025?
Success under Andromeda requires a complete strategic reorientation. The primary focus must shift to producing diverse, high-quality creative assets, simplifying account structures, embracing automation, and maintaining consistent creative refresh cycles. These aren’t optional optimizations—they’re fundamental requirements for competitive performance.
Why Is Creative Diversity So Important for Meta Ads Now?
If targeting once defined success, creativity is now the true differentiator. Andromeda’s machine learning systems thrive on creative diversity, but this doesn’t mean cosmetic variations.
It means meaningfully different ideas that represent distinct audience motivations or emotional triggers. Minor changes like headline tweaks or background color adjustments no longer register as different to the algorithm.
The P.D.A. Framework for Creative Variation:
To create genuinely diverse ads, use the Persona-Desire-Awareness (P.D.A.) framework:
- Persona: Are you speaking to a student, parent, or business owner?
- Desire: Does the audience want to save time, reduce costs, or gain confidence?
- Awareness: Are they discovering the problem, comparing solutions, or ready to purchase?
Combining these variables creates ads that differ in tone, imagery, and message, giving Andromeda distinct signals to match with different user segments.
How Many Creatives Should I Run Per Campaign?
The recommended number of creative assets has dramatically increased. Meta now advocates for 10-50 ads per ad set, a stark contrast to the previous best practice of 3-6 ads.
However, balance is critical. Loading too many creatives can backfire. The algorithm needs sufficient budget to properly evaluate each option, so 30-40 creatives in a single ad set with limited daily spend prevents adequate learning.
Best Practice Range: 8-15 genuinely different creative variations per campaign for most budgets.
How Often Should I Refresh My Ad Creative?
Ad fatigue occurs faster under Andromeda’s stricter evaluation criteria. Advertisers should refresh ads every 2-4 weeks to maintain performance and prevent creative similarity penalties.
Monitor your Creative Fatigue and Creative Similarity metrics in Meta’s dashboard. If Creative Similarity is high, meaning you lack diversity, the Andromeda algorithm will punish your account by raising your CPMs because it views the content as repetitive and fatiguing.
What Ad Formats Should I Use for Better Performance?
Don’t rely solely on one format. Successful campaigns now include:
- User-generated content (UGC) videos
- Professional product demonstrations
- Static images with bold visual hooks
- Carousel ads showcasing multiple products or benefits
- Short-form vertical videos optimized for Reels placement
- Testimonial-style content
- Educational or how-to formats
Advertisers using Advantage+ Creative, which generates multiple variations automatically, saw 22% higher ROAS on average.
What’s the Best Campaign Structure for Meta Ads in 2025?
Consolidation is now a performance driver rather than a limitation. The optimal structure for most advertisers:
One product = One campaign = One ad set + Broad targeting + Multiple diverse creatives
Before Andromeda, advertisers created hyper-segmented campaigns with dozens of ad sets targeting narrow audiences. With Andromeda, consolidation wins.
Should I Use Advantage+ Automation or Manual Controls?
Andromeda was specifically designed to make Advantage+ work better. The system now has the processing power and intelligence to make superior automated decisions.
Key Advantage+ features to enable:
- Advantage+ Audience: Allow Meta to expand beyond your defined audience
- Automatic Placements: Let the algorithm determine optimal placement distribution
- Campaign Budget Optimization (CBO): Enable budget distribution across ads based on performance
- Advantage+ Creative: Turn on AI-driven creative optimizations (with caution in regulated industries)
In Meta’s tests, campaigns using Advantage+ with broad targeting saw up to 10% lower cost per lead compared to traditional manual campaigns.
How Important Is Pixel Tracking for Meta Ads Success?
Pixel setup quality and accurate event reporting directly affect Andromeda’s ability to optimize effectively. Poor tracking data leads to higher costs and lower conversion rates because the algorithm makes optimization decisions based on incomplete or inaccurate signals.
Ensure your Meta Pixel is properly implemented, conversion events are firing correctly, and the Conversions API is integrated for maximum data accuracy.
How Do I Make My Meta Ads Perform Better After Andromeda?
Summary: What’s the Step-by-Step Process to Improve My Facebook Ads?
Improving ad performance under Andromeda requires systematic implementation of creative diversity protocols, proper campaign structure, strategic budget allocation, and continuous testing. These actionable steps translate the strategic principles into day-to-day advertising operations.
How Do I Create Diverse Creative for Meta Ads?
Implement this step-by-step framework to maximize Andromeda’s performance potential:
Step 1: Audit Your Current Creative Library
Identify how many truly distinct creative concepts you’re currently running. If multiple ads share similar:
- Visual composition
- Opening hook (first 3 seconds)
- Creator type or setting
- Core messaging angle
They likely register as the same ad to Andromeda.
Step 2: Create Meaningful Variations
Develop 8-15 creatives per campaign that vary across these dimensions:
- Hook Variety: Different opening statements, questions, or visual grabs
- Format Mix: Combine UGC videos, static images, carousels, and demonstrations
- Messaging Angles: Address different pain points, benefits, or use cases
- Visual Diversity: Different settings, people, color schemes, and compositions
- Length Variation: Mix 6-second, 15-second, and 30-second videos
Step 3: Implement Continuous Creative Testing
Don’t wait for performance to decline before refreshing. Establish a regular cadence:
- Add 2-3 new creative concepts weekly
- Retire worst-performing ads after they’ve had adequate learning time (4-7 days minimum)
- Maintain a rotation of proven winners alongside new tests
- Track which creative themes drive the strongest performance
How Should I Allocate Budget Across My Meta Ads?
Effective budget allocation under Andromeda requires flexibility. Rather than fixed testing versus scaling ratios:
When Strong Winners Emerge:
- Allocate more budget toward scaling proven performers
- Maintain 20-30% of budget for ongoing creative testing
- Monitor for fatigue signals (rising CPMs, declining CTR)
When No Clear Winners Exist:
- Shift more spend into testing new creative concepts
- Ensure each creative receives sufficient impressions for meaningful data
- Avoid premature optimization decisions
What’s the Best Campaign Setup for Different Business Types?
For E-commerce:
- One campaign per product category
- Broad targeting (country-level or large region)
- 10-15 diverse creatives per campaign
- Advantage+ Shopping campaigns for catalog products
For Lead Generation:
- One campaign per core offer
- Minimal audience restrictions
- 8-12 creative variations addressing different pain points
- Advantage+ App campaigns where applicable
For Brand Awareness:
- Consolidated reach campaigns
- Maximum placement diversity
- Heavy emphasis on video creative variety
- Regular creative refresh every 2-3 weeks
Should I Separate Testing and Scaling Campaigns?
Separate Testing from Scaling:
Mixing new creatives with proven winners slows testing because Meta favors ads with existing social proof. Consider maintaining:
- Testing Campaign: New creative concepts with dedicated budget
- Scaling Campaign: Proven winners with majority of budget allocation
This separation ensures new concepts receive adequate spend to generate meaningful performance signals.
How Can AI Tools Help with Creative Production?
A high-velocity creative pipeline is essential because Andromeda relies on a constant flow of new inputs to optimize effectively.
AI-assisted workflows help teams:
- Produce more concepts quickly
- Reduce creative production cycle times
- Replace fatigued creatives before performance declines
- Test more variations without proportional cost increases
Why Are My Meta Ads Getting More Expensive After Andromeda?
Summary: What’s Causing My Facebook Ad Costs to Increase?
Rising advertising costs after the Andromeda update stem from specific structural misalignments between your campaign setup and the new system’s requirements. Understanding these causes and implementing targeted solutions can reverse cost increases and restore profitable performance.
What Causes CPM Increases After the Andromeda Update?
1. Do I Have Enough Creative Diversity?
CPMs climbing 20%+ since mid-2025 indicates you’re likely competing inefficiently under the new system. If your creative library consists of minor variations rather than truly distinct concepts, Andromeda has limited options for matching ads to specific user preferences.
Result: The algorithm repeatedly shows similar ads to overlapping audiences, driving up frequency and costs while reducing effectiveness.
2. Is My Campaign Structure Hurting Performance?
Andromeda rewards consolidated campaigns with creative diversity; fragmented structures pay a premium. Running multiple ad sets with narrow targeting prevents the algorithm from efficiently learning and allocating budget.
Each fragmented ad set receives less data, extends the learning phase, and competes inefficiently in the auction. Consolidation provides the volume Andromeda needs for effective optimization.
3. Am I Getting Penalized for Creative Similarity?
If your Creative Similarity is high, the Andromeda algorithm will punish your account by raising your CPMs because it views the content as repetitive and fatiguing.
This penalty mechanism actively increases costs for advertisers who haven’t adapted to the creative diversity requirements. The algorithm essentially charges more to show what it perceives as repetitive, low-value content.
4. Are Competitors Adapting Faster Than Me?
As more advertisers adopt Andromeda-optimized approaches, those using old tactics face increasingly difficult competitive dynamics. The competitive advantage window is closing, and what feels like strong performance today will be table stakes in six months.
Advertisers who adapt establish algorithmic learning and creative testing momentum that compounds over time, while those using 2024 strategies pay progressively higher costs for worse results.
5. Am I Killing Ads Too Quickly?
Advertisers who cut ads too quickly often miss the upside of later optimisation. The extended learning phase means ads that appear underperforming in the first 48 hours may become top performers after 5-7 days once Andromeda identifies their optimal micro-audiences.
Killing ads too early forces you to constantly restart the learning process, inflating costs and preventing the algorithm from stabilizing performance.
How Can I Reduce My Meta Ad Costs?
Immediate Actions:
- Consolidate Campaign Structure
- Merge overlapping ad sets
- Eliminate narrow targeting parameters
- Simplify to one campaign per product/offer
- Expand Creative Library
- Produce 8-15 truly distinct creatives
- Ensure variation in hooks, formats, and angles
- Address different customer motivations and awareness stages
- Enable Advantage+ Features
- Turn on Advantage+ Audience expansion
- Use Automatic Placements
- Activate Campaign Budget Optimization
- Extend Learning Patience
- Give new creatives 4-7 days minimum before elimination
- Allocate sufficient daily budget for meaningful learning
- Monitor performance trends rather than day-one results
Case Study Example:
Karina Gardner, who runs a design bootcamp, saw her costs spike to $86 per conversion. She implemented the protocol by adding just eight new creatives—specifically one carousel and three images, plus AI iterations. Within 24 hours, her cost per result dropped to $13.87.
What Metrics Should I Monitor to Prevent Cost Increases?
Track these key indicators:
- CPM Trends: Sustained increases signal creative fatigue or similarity penalties
- Creative Fatigue Score: New Meta metric indicating when refresh is needed
- Creative Similarity Score: High scores trigger cost penalties
- Budget Distribution: Healthy campaigns distribute spend across multiple creatives
If you’re seeing immediate budget concentration on 1-2 ads within hours, you likely don’t have enough meaningful creative diversity for the algorithm to work with.
What Advanced Strategies Work Best for Meta Ads Now?
When Did the Andromeda Update Actually Roll Out?
Andromeda didn’t flip a switch and immediately change behavior. It learned gradually which ads to suppress and which to promote. This explains why some advertisers saw immediate impact in mid-2024 while others didn’t notice changes until September or October 2025.
Meta deployed the update to different advertisers at different times, and the AI’s learning curve meant initial performance changes appeared as “bad months” rather than systematic shifts.
How Does Individual-Level Optimization Actually Work?
Andromeda moves Meta from audience-level assumptions to individual-level prediction. Rather than determining which audience segment should see an ad, the system identifies which specific ad each person is most likely to respond to.
This architectural change allows Meta to handle much larger creative volumes while making more precise budget allocation decisions. Performance depends less on manual targeting and more on supplying diverse, high-quality creatives.
How Do I Manage the Transition Period?
Expect short-term volatility during adaptation. Early rollouts in 2025 saw CPM spikes, ghost approvals, and erratic budget shifts, particularly for smaller budgets.
However, the pattern is clear: marketers who adapt to Andromeda’s rules reduce costs and improve engagement, while those holding onto old tactics risk higher spend for weaker results.
What’s Coming Next for Meta Advertising?
Andromeda isn’t a temporary algorithm fluctuation that will revert. This is the permanent foundation of Meta’s advertising infrastructure going forward.
The system will continue evolving with more sophisticated hardware and AI capabilities. Looking forward, the Andromeda model architecture is expected to transition to support an autoregressive loss function, leading to a more efficient and faster inferencing solution that delivers a more diverse set of ad candidates.
What Mistakes Should I Avoid with Meta Ads After Andromeda?
1. Should I Stop Creating Multiple Ad Sets?
Creating dozens of narrow ad sets based on interests and demographics actively harms performance under Andromeda. The algorithm needs volume and variety within campaigns, not across fragmented structures.
2. Are Small Creative Changes Enough?
Changing headlines on the same image or making small color adjustments doesn’t provide the diversity Andromeda requires. The system’s visual recognition models identify these as essentially identical ads.
3. How Long Should I Wait Before Optimizing Ads?
Making elimination decisions within 24-48 hours prevents the algorithm from completing its learning process and finding optimal micro-audiences for each creative.
4. What New Metrics Should I Track?
Meta introduced Creative Fatigue and Creative Similarity metrics specifically for Andromeda. Ignoring these signals while they trigger cost penalties wastes budget unnecessarily.
5. Should I Fight Against Automation?
Attempting to maintain manual control through narrow targeting, complex structures, and limited creative variety means competing against the system’s design rather than leveraging its capabilities.
Key Takeaways: What Do I Need to Remember About Meta Andromeda?
The Meta Andromeda update represents a permanent shift in digital advertising infrastructure, not a temporary fluctuation. Success requires fundamental strategic adaptation:
Essential Principles:
- Creative diversity is now the primary performance driver
- Simplified structures with broad targeting outperform complex segmentation
- Advantage+ automation delivers superior results when properly implemented
- Extended learning phases require patience before optimization decisions
- Continuous creative refresh prevents fatigue and cost increases
Performance Benchmarks:
- Adapted advertisers: 8-17% conversion increases, up to 16% cost reductions
- Unadapted advertisers: 20%+ CPM increases, declining ROAS
- Creative refresh cadence: Every 2-4 weeks
- Learning phase: 4-7 days minimum for new creatives
- Optimal creative volume: 8-15 genuinely diverse variations per campaign
The Competitive Window:
The advertisers seeing 8-17% conversion increases and 16% cost reductions didn’t get there by accident. They restructured early, tested aggressively, and built creative production systems whilst others waited to see how it plays out.
As the baseline performance level rises with wider adoption, early movers gain compounding advantages through months of algorithmic learning and creative testing data. The window for competitive advantage narrows with each passing month.
The fundamental question is no longer whether to adapt to Andromeda, but how quickly you can implement the required changes before outdated strategies become prohibitively expensive.
Frequently Asked Questions About Meta Andromeda
Q: Should I still use lookalike audiences after Andromeda?
Yes, but their role has changed. Lookalike audiences now serve as useful signal inputs rather than hard boundaries. The algorithm learns from the seed but expands beyond it to find similar behavioral patterns.
Q: How many ads should I run per campaign with Andromeda?
8-15 genuinely diverse creative variations work for most budgets. Ensure each creative receives sufficient budget for adequate learning—too many ads with limited daily spend prevents proper evaluation.
Q: When should I kill underperforming ads under Andromeda?
Wait 4-7 days with meaningful spend before making elimination decisions. The extended learning phase means early performance may not reflect ultimate potential.
Q: Can I still use interest-based targeting after the Andromeda update?
You can, but it’s increasingly counterproductive. Broad targeting allows Andromeda to find your best customers more efficiently than manual interest selection.
Q: Why did my winning ads from 2024 stop working after Andromeda?
Andromeda learned gradually which ads to suppress and which to promote. Your “winner” ads from April might have worked fine through June, then quietly stopped getting delivery in August.
Q: Is the Andromeda update permanent or will Meta change it back?
Yes. Andromeda is the permanent foundation of Meta’s advertising infrastructure going forward, and the system will only become more sophisticated.
Conclusion: What Should I Do About Meta Andromeda?
The Meta Andromeda update demands a fundamental rethinking of digital advertising strategy. The transition from targeting-first to creative-first methodologies represents more than a tactical adjustment—it’s a paradigm shift in how performance advertising operates on Meta’s platforms.
Advertisers who embrace this change by prioritizing creative diversity, simplifying campaign structures, and leveraging AI-driven automation will discover new levels of efficiency and performance. Those who resist will face escalating costs and diminishing returns as the competitive landscape evolves.
The evidence is clear: adaptation works. Early adopters report double-digit performance improvements while reducing costs. The strategic imperative is not whether to change, but how quickly you can implement the creative production systems, simplified structures, and automated optimization that Andromeda rewards.
Your competitive advantage depends on taking action before the adaptation window closes entirely. The algorithm is already learning, the competition is already optimizing, and the baseline performance level continues rising. The question is whether you’ll lead this transition or be left fighting an increasingly expensive battle with outdated tactics.
The future of Meta advertising belongs to those who recognize that creative excellence, strategic simplicity, and algorithmic partnership have replaced manual targeting precision as the foundations of advertising success.
