Meta Campaign Structure in the Andromeda Era

A conversation I’ve been having more and more often recently when auditing Paid Social Ads campaigns:

“The structure looks great!… for 2024. Unfortunately, this won’t work in 2026.”

“Why not?”

It’s a fair question. For years, the best Paid Social accounts were built around control. More campaigns. More ad sets. More audience splits. More funnel stages. More rules designed to tell the platform exactly who to reach and when.

That used to make sense.

But the system we are advertising into has changed dramatically. Meta is no longer the same platform it was even two years ago, and campaign structures that once looked strategic can now actively hold performance back. What used to feel organised can now create fragmentation. What used to feel precise can now restrict the algorithm. What used to be considered best practice can now be the thing preventing scale.

Those outside the industry tend to think I’ve become slightly too poetic when I mention… Andromeda!

So I usually have to quickly explain that this is Meta’s new retrieval-based, AI-powered ad retrieval and ranking system. It replaced the previous ad infrastructure and marks the biggest shift in how Meta serves ads in more than a decade.

For those who enjoy a good fact: Meta chose the name as a poetic reference to the sheer scale of the system. In the same way that the Andromeda galaxy contains more than a trillion stars and continues to expand, Meta’s algorithm was built to search through tens of millions of possible ad-to-user matches in real time.

So, in the end, the answer is poetic after all.

Above: A self-created diagram

From Micro-Planets to Galaxy Signal Design

Before this galaxy-level shift, Paid Social managers typically structured accounts around heavy segmentation by audience and placement, with separate ad sets for interests, lookalikes and retargeting. And, naturally, only around five to six ads per ad set.

The logic was understandable. If you wanted to reach new customers, you built prospecting campaigns. If you wanted to nurture warmer users, you built middle-of-funnel campaigns. If you wanted to capture existing demand, you built bottom-of-funnel campaigns. Then you added lookalikes, interest stacks, retargeting windows, placement splits and creative variations for each layer.

The account would often look highly sophisticated on paper. It had structure. It had stages. It had labels that made everyone feel in control.

But Andromeda is far more advanced. It brings together signals across audiences, placements and behaviours to identify conversion efficiency at scale. It needs “a good chunk” of data and “a good chunk” of creatives — so yes, more than five ads per ad set will be required.

Because in this larger galaxy, our role is no longer to “control” performance. It is to design systems that produce strong, clean signals for the algorithm.

This is the mindset shift. We are not building tiny, isolated planets anymore. We are building a connected galaxy of signals.

Consolidation: The Way to Create Strong, Clean Signals

Consolidation is essential if you want to create those strong, clean signals. The algorithm is effectively searching through millions of potential ad-to-user combinations, so your data needs to be concentrated enough for Andromeda to identify patterns clearly. This allows for faster learning and smoother scaling.

Fragmented accounts make this harder. When budget is split across too many campaigns and ad sets, each part of the account is forced to learn from a smaller data pool. Signals become weaker. Learning takes longer. Performance becomes more volatile. Scaling becomes harder because the system does not have enough concentrated information to make confident decisions.

So, rather than spreading budget thinly across several campaigns and ad sets, Meta is increasingly rewarding:

Fewer campaigns

Fewer ad sets

More budget per learning system

Broader targeting inputs

This does not mean there should be no structure. It means the structure needs to serve the system rather than satisfy an old media buying instinct. Consolidation is not about being lazy. It is about removing unnecessary barriers so the algorithm can do the thing it is now built to do.

The best account structures in 2026 will not necessarily be the most complex. They will be the clearest.

Campaign Structure Should Mirror Business Goals, Not Audiences

The first step towards effective consolidation is to structure campaigns around business goals, not audience segments. This is the most scalable way to work because it aligns signals around the final business objective.

So, instead of thinking:

“TOF / MOF / BOF”

You should be thinking:

Acquisition vs Retention

Lead vs Sales

Geographic or operational constraints

This is a simple but important distinction. Audiences are not business goals. Funnel labels are not business goals. Interest groups are not business goals.

A campaign should exist because it supports a distinct commercial outcome. If the business wants to acquire new customers, that may justify a dedicated acquisition campaign. If the business wants to drive repeat purchases from existing customers, that may justify a retention campaign. If different regions have different budgets, shipping constraints, regulations or sales teams, that may justify separate structures.

But splitting campaigns simply because one audience is “cold” and another is “warm” is becoming less useful. Meta can increasingly identify intent, likelihood to convert and relevance without us needing to manually pre-sort users into rigid funnel buckets.

The question is no longer: “Which audience am I targeting?”

The better question is: “What signal am I asking the system to optimise towards?”

Using Broad as an Expanding Universe

If Meta requires consolidated data, then broader targeting becomes the natural next step in that strategy.

We are moving away from manual control, granular segmentation and rigid funnel logic. Whether we like it or not, Meta now has access to far more behavioural and predictive user data than we can realistically process ourselves. This makes strict audience-based targeting less and less relevant.

A strong and diverse creative system is becoming the new targeting layer.

That sentence is worth sitting with, because it changes a lot.

In the old world, targeting was something you set inside the ad set. You told Meta: show this ad to these people, with these interests, in these placements, at this stage of the funnel.

In the Andromeda era, targeting is increasingly expressed through the creative itself. The angle, hook, format, problem, promise and proof point all help the system understand who the ad is for.

A parent might respond to safety and easy cleaning. A renter might respond to instant transformation. A design-conscious homeowner might respond to aesthetics and texture. A time-poor household might respond to practicality and maintenance. These people may all sit inside the same broad audience, but they are qualified by different messages.

Broad does not mean vague. Broad means giving the system room to find the right person, while giving it enough creative signals to understand why that person should care.

Creatives, the New Targeting Layer

Before Andromeda, audiences largely dictated delivery. Now, Andromeda uses creative signals to understand which users should see a particular ad.

The purpose of this retrieval engine is to build a “shortlist” of strong ads — not one universal winner, but a group of highly relevant variations that resonate differently with different users. Your hook and message are what qualify the audience.

This is where creativity and diversity become critical. A “winning” ad for one person might be completely irrelevant to another, even if both people share the same declared interest, such as oily skin beauty products. Andromeda’s aim is to personalise delivery so that the right user sees the right variation.

It is no longer just about quantity. It is about meaningful variety.

This is where many brands get stuck. They hear that they need more creatives, so they produce more versions of the same idea. They change the opening line. They use a slightly different background. They swap the creator but keep the same structure. They test a different colour overlay. Technically, there are more ads in the account. Strategically, the system is still being fed the same concept.

Andromeda does not just need more assets. It needs more distinct signals.

That means more angles. More emotional entry points. More product use cases. More reasons to believe. More objections answered. More formats that feel native to different types of users.

The creative strategy has to become broader and more intentional at the same time.

The More Variety, the Higher the Chances of Finding Winners

Andromeda learns not only from clicks and conversions, but also from engagement patterns and interaction behaviour.

Small tweaks are no longer enough. Changing the copy. Swapping the background. Adjusting overlays. Reusing the same creator in slightly different shots. Meta increasingly reads this as the same ad.

We have moved beyond the “find 7 differences” era. “Different” now means genuinely different:

Different concepts and angles

Different narrative frameworks, such as problem/solution, pain point, AIDA, testimonial-led or contrarian takes

Different formats, including UGC, static, video and carousel

Different personas and funnel stages

This is especially important because the best-performing creative in an account is rarely the best-performing creative for every possible customer. It is simply the creative that found a strong pocket of relevance.

Another concept may unlock a completely different pocket. Another hook may speak to a different pain point. Another format may appeal to a different level of awareness. Another persona may make the product feel newly relevant to someone who previously scrolled past.

This is why creative variety is not just a “nice to have”. It is one of the core levers for scale.

If every ad says the same thing in roughly the same way, the system has limited room to expand. If each ad gives Andromeda a genuinely different reason someone might buy, the account becomes far more scalable.

Scaling Comes From Structured Creative Diversity

When we talk about tapping into different personas, we are also talking about one of the most reliable ways to scale.

Different people buy for different reasons, even if both have shown an interest in “interior decor”. The more motivators or hooks you give Andromeda, the more opportunities it has to find a match.

Meta has even suggested in internal analysis that “different motivators unlock new audiences 89% of the time.” This should be treated with some caution, but it strongly reflects what we see in practice: creative variation drives expansion.

This is the type of thinking that matters now. Not just “what is the product?” but “what role does this product play in different people’s lives?”

One person may buy the rug because they have children and need something washable. Another may buy because they have just moved house and want the room to feel finished. Another may buy because they care about design but do not want something that feels too loud. Another may buy because they have pets. Another may buy because they are tired of replacing cheap rugs that do not survive real life.

The product is the same. The motivation is different.

A strong scaling system will usually include:

3–4 fundamentally different hooks

Distinct creative formats, such as UGC vs polished vs static

Different emotional triggers

Different pacing and storytelling structures

However, before you introduce a completely new hook into an established campaign, you need to make sure it has been tested first. This is why a testing environment still matters and should remain part of your account structure.

Why Testing Still Matters

Consolidation does not mean throwing every idea into one campaign and hoping the algorithm works it out.

Testing still matters because new creative ideas carry risk. A new hook may be brilliant, or it may confuse the market. A new persona may unlock scale, or it may pull delivery away from the strongest buyers. A new format may improve attention, or it may fail to communicate the product clearly enough.

The role of a testing campaign is to create a controlled environment where genuinely different ideas can be evaluated before they are introduced into the main scaling system.

This is where you test:

New hooks

New offers

New formats

New personas

New product angles

New objections

New proof points

The key is to test meaningful differences, not tiny cosmetic variations. A testing campaign should help you learn something useful about what motivates customers. It should not become a dumping ground for minor creative edits.

When a concept proves it can attract attention, drive engagement and convert efficiently, it can then be moved into the consolidated scaling campaign with more confidence.

What This Looks Like in Practice

Old structure:

Campaign 1: TOF
LAL
Interests

Campaign 2: MOF

Campaign 3: BOF

Each campaign has multiple ad sets and narrow budgets.

New structure:

1–2 consolidated scaling campaigns

1 campaign focused on finding new people who will buy from your ecommerce store or submit a lead

1 testing campaign

Broad targeting, or minimal segmentation

Separate ad sets should only be used where there is a genuine difference in audience, objective or, in some cases, budget allocation. For example: distinct industries, significantly different audience profiles, or geographic and operational requirements.

Each ad has a different purpose or hook

Advantage+ placements fully enabled

Retargeting is often folded into the same ecosystem or significantly simplified

In practice, this creates a much cleaner account.

The scaling campaign becomes the main learning engine. It receives the strongest creative concepts, enough budget to learn properly and enough variety to personalise delivery across different users.

The testing campaign becomes the exploration engine. It helps you identify new hooks, new messages and new creative territories before they are added into the main system.

This gives the account both stability and discovery. You are not constantly disrupting your core campaign with unproven ideas, but you are also not relying on the same few winning ads until fatigue eventually catches up with you.

The Real Shift: From Media Buying Control to Creative System Design

The deeper change here is not technical. It is strategic.

For a long time, performance marketers were rewarded for account architecture. The person who could build the most detailed funnel, the most precise retargeting logic and the most segmented structure was seen as the one with the most control.

But control has changed shape.

In the Andromeda era, control does not come from restricting the algorithm. It comes from feeding it better inputs.

That means:

Clearer objectives

Cleaner account structures

Stronger data concentration

Broader targeting freedom

More distinct creative ideas

Better understanding of customer motivations

The media buyer’s role has not become less important. It has become different. The job is no longer just about pulling levers inside Ads Manager. It is about designing a system where strategy, creative, data and machine learning can work together.

The accounts that win will not be the ones with the most ad sets. They will be the ones with the clearest signals and the strongest creative range.

Start Building Galaxies, Not Micro-Planets

The funny thing is that none of this is really just about campaign structure. It is something more poetic. It is about accepting that the system has changed.

For years, we built increasingly complex account structures — based on micro-planets — to help Meta find the right people. Today, Meta is much better at finding those people than we are.

Our role is no longer to build endless audience combinations or separate every stage of the funnel into its own campaign. Our job is to create a galaxy that can feed the system with clear business objectives, enough data to learn from and enough creative variety to personalise the experience.

The irony is that campaign structures are becoming simpler while creative strategy is becoming more sophisticated.

We are spending less time organising audiences and more time understanding people. And honestly, I think that’s a far more interesting job.

Because the brands that adapt fastest will not be the ones trying to rebuild 2024 inside a 2026 system. They will be the ones willing to simplify the structure, sharpen the signals and take creative strategy seriously.

That is the real opportunity of the Andromeda era.

Not more control for the sake of control.

More clarity. More creativity. More useful signals.

A bigger galaxy for the algorithm to explore.