Why Meta Creative Diversity Matters More After Andromeda

Meta Andromeda is changing what it means to produce more creative for Meta ads.

For years, growth teams increased creative volume by taking a winning concept and producing more versions of it: new hooks, new headlines, new opening frames, new aspect ratios, new creators, and new edits.

That strategy still has a place.

But Meta’s advertising system has become much better at understanding the content of an ad and matching different ads to different users. As a result, simply increasing the number of files in an ad account does not necessarily increase the diversity of ideas Meta has available to deliver.

The more useful question in 2026 is not: How many ads are we producing?

It is:

How many meaningfully different creative concepts are we giving Meta?

That distinction is becoming increasingly important as Andromeda plays a larger role in Meta’s ad retrieval system.

What is Meta Andromeda?

Andromeda is Meta’s next-generation ad retrieval system.

Ad retrieval happens before the final auction and ranking process. Meta may have tens of millions of ads that are technically eligible to be shown, but evaluating every one of them in depth for every impression would be computationally impractical.

Instead, Meta first has to identify a much smaller group of potentially relevant ads.

According to Meta, Andromeda reduces a pool of tens of millions of eligible ads to a few thousand candidates that can then move into subsequent ranking stages.

Meta says the system uses deep neural networks to model relationships between people and advertisements and dramatically increases the capacity of its previous retrieval architecture.

That matters because creative is increasingly part of how Meta determines relevance.

An ad is not simply an uploaded file competing independently of everything around it. Meta’s machine learning systems are increasingly capable of understanding the content, context, and characteristics of creative.

For advertisers, that makes creative diversity considerably more important.

Why Andromeda matters for Meta creative testing

The practical consequence of Andromeda is not that advertisers should stop producing creative variations.

It is that creative volume and creative diversity should no longer be treated as the same thing.

Imagine a brand produces 30 ads from one piece of UGC footage.

The team changes:

• the first three seconds
• the headline
• the subtitles
• the background music
• the CTA
• the aspect ratio

From a production perspective, the team has created 30 ads.

From a creative strategy perspective, it may still have created only one fundamental concept.

The customer, problem, visual environment and reason to buy are all the same.

Those variations can absolutely produce different performance results. A stronger hook or edit can make an ad substantially better.

But they do not give the account the same breadth as 30 genuinely different creative ideas.

That is the distinction growth teams need to start planning around.

Creative variation vs. creative diversity

Creative variation means changing the execution of an existing idea.

Creative diversity means introducing a meaningfully different idea.

For example, imagine a premium cookware company.

One ad says:

“The pan you’ll own for the next 20 years.”

Another says:

“Why restaurant-quality searing starts with the pan.”

Another says:

“The wedding gift they’ll still be using in 2046.”

Those ads may promote exactly the same product, but they represent different reasons to care.

The first is about durability and value.

The second is about performance.

The third is about gifting.

Compare that with making three versions of the first ad using different opening hooks.

That is creative variation.

Both are useful, but they solve different problems.

Creative variation helps optimize an idea

You might test:

• hooks
• headlines
• creators
• pacing
• opening visuals
• CTAs
• copy length

The question is:

How do we make this concept perform better?

Creative diversity helps discover new opportunities

You might test:

• different customer motivations
• different customer profiles
• different product benefits
• different use cases
• different formats
• different visual approaches

The question is:

What other ideas could work?

Strong Meta creative testing needs both. The harder question is how much of each you actually need as spend increases. We break that down in ⁠How Many Ads Do You Really Need to Run on Meta as You Scale?, including how to think about angle, hook, format, and iteration coverage.

What Meta means by creative diversification

Meta itself has increasingly emphasized creative diversification across its advertising products.

This makes sense within the direction of its recommendation systems.

If Meta is attempting to find the best combination of person, advertisement, placement, and moment, giving the system genuinely different creative options increases the number of ways a brand can potentially appeal to different users.

That does not mean every ad needs to be radically different.

It means an account should not rely entirely on minor iterations of the same handful of concepts.

A healthy creative library should contain both breadth and depth.

Breadth comes from testing new concepts.

Depth comes from iterating on concepts that work.

Most accounts are much better at depth.

A winning ad appears, so the team produces ten more versions of it.

That makes sense until almost the entire creative pipeline becomes variations of previous winners.

At that point, the account may be producing a lot without learning very much.

The four dimensions of Meta creative diversity

A useful way to build a more diverse Meta creative testing program is to deliberately vary creative across four dimensions.

1. Buyer motivation

Start with the reason someone would buy.

A customer might care about:

• saving money
• saving time
• improving performance
• convenience
• status
• comfort
• appearance
• durability
• simplicity

One product can often satisfy several motivations.

Testing them separately gives Meta different messages to match against different users.

For a mattress company, for example, one ad might focus on back support while another focuses on cooling and another focuses on durability.

Same mattress.

Three different reasons to buy.

2. Buyer profile

The same product can also mean different things to different customers.

A protein supplement could be positioned differently for:

• competitive athletes
• people trying to build muscle
• busy professionals
• older consumers
• beginners returning to fitness

You do not necessarily need to create narrow manual audiences for each profile.

The creative itself can communicate who the product is for.

This is an important shift as Meta continues moving advertisers toward broader targeting and increasingly automated audience selection.

3. Creative format

Format is another source of meaningful separation.

Instead of producing ten versions of the same UGC execution, a brand might test:

• UGC testimonial
• founder video
• product demonstration
• editorial static
• comparison ad
• customer review
• motion graphic
• carousel
• before-and-after
• lifestyle photography

The goal is not to use every format simply because it exists.

The format should change how the product story is communicated.

There is a limit to how far that variation should go. Creative can look meaningfully different without becoming unrecognizable as your brand. ⁠Do You Need Branded Ads on Meta as You Scale? covers what should stay consistent as creative volume expands.

4. Customer context

Finally, change the situation in which the product appears.

A luggage company might show the same product through:

• business travel
• international travel
• weekend trips
• packing
• airport transit
• hotel arrival
• gifting

Context can create significant creative separation without requiring the brand to invent an entirely new value proposition.

Why minor ad variations still matter

There is a temptation to interpret the Andromeda conversation as “variations are dead.”

That is too simplistic.

A small creative change can absolutely produce a large performance difference.

The first three seconds of a video matter.

The creator matters.

The headline matters.

The product demonstration matters.

The job of variation testing is to improve the execution of an idea that has potential.

The mistake is using variation testing as a substitute for concept testing.

If a brand has one winning creative concept, it should explore that concept.

But it should also continue introducing new concepts.

Otherwise optimization slowly turns into concentration.

A better Meta creative testing framework

Instead of planning production around a raw number of ads, start by planning the hypotheses you want to test.

For example, rather than saying:

We need 20 new Meta ads this month.

A growth team could define:

4 buyer motivations

Performance, convenience, value, and status.

3 customer profiles

First-time buyer, enthusiast, and gift buyer.

4 creative formats

UGC, product demonstration, static comparison, and founder video.

Then choose the most promising combinations.

You do not need to produce every possible combination.

The framework simply forces the team to think about diversity before production begins.

A batch might look like:

Now six ads represent six actual creative hypotheses.

That gives the account considerably more information than six hook changes applied to the same video.

How many Meta ad concepts should you test?

There is no universal number of concepts every advertiser should launch.

And there is no public Meta guidance saying brands need eight, ten, twelve, or any other specific number of “Entity IDs.”

The appropriate number depends on factors including:

• monthly Meta spend
• conversion volume
• creative production capacity
• testing budget
• campaign structure
• how quickly ads receive enough delivery to evaluate

A brand spending $20,000 per month should not necessarily follow the same creative cadence as a brand spending $2 million per month.

The better principle is proportionality.

As spend increases, creative exploration should increase with it.

But the additional creative should introduce new hypotheses, not only additional files.

How should brands change their Meta creative strategy?

There are four practical changes I would make.

Stop reporting creative volume by file count alone

“We produced 80 ads this month” tells you very little.

Track the number of new concepts alongside total assets.

If 80 assets came from six concepts, you should know that.

Separate concept testing from variation testing

Create separate production goals for each.

For example:

60% new concepts.

40% iterations of existing winners.

The exact ratio will vary, but separating the categories prevents iteration from consuming the entire production pipeline.

Build creative briefs around differences

Every brief should make clear what is different about the concept.

Define the:

• audience
• motivation
• promise
• format
• context
• proof mechanism

You do not need to change all six.

Changing two or three can be enough to create a meaningfully different creative direction.

Measure concept concentration

Look at your last 30 ads.

How many are actually based on the same three ideas?

This is one of the simplest audits a growth team can perform.

A large account can have hundreds of active ads and still have surprisingly little conceptual diversity.

If part of that production is being handled externally, the same distinction should shape how you evaluate a partner. ⁠What to Look for in Meta Ad Testing Services covers how to tell whether a creative partner is actually producing distinct concepts or simply increasing the number of variations.

What Andromeda means for high-volume creative production

Andromeda does not make high-volume creative production less important.

It makes the composition of that volume more important.

Growth teams still need enough creative to:

• replace fatigued ads
• explore new customer motivations
• test new formats
• support promotions
• enter new markets
• scale spend

But producing more of the same idea eventually provides diminishing strategic value.

The production challenge is therefore shifting.

The old question was:

How do we make more ads?

The more useful question now is:

How do we produce more distinct ideas without slowing production down?

That is ultimately an operational problem as much as a creative one. More concepts mean more briefs, more review cycles, more formats, and more decisions to manage at the same time. ⁠Why Scaling Meta Creative Suddenly Feels Harder Than It Used To breaks down why creative production becomes an operating-system problem as accounts grow.

How Campfire approaches Meta creative testing

Campfire is built for teams that need to increase Meta creative output without committing to a fixed agency retainer.

Creative is produced on a per-ad basis, allowing teams to scale production up or down depending on testing needs.

The production process can support both sides of the testing cycle: developing distinct new concepts and producing additional variations around ideas that are already performing.

The objective is not simply to increase the number of ads entering the account.

It is to increase the number of useful creative experiments the account can run.

See how Campfire helps brands scale Meta creative production.

The takeaway

Meta Andromeda makes one distinction increasingly important:

Creative quantity and creative diversity are not the same thing.

Twenty different files can still represent one creative idea.

And one product can support dozens of genuinely different ideas when you change the buyer, motivation, context, promise, or execution.

Growth teams should therefore stop asking only how many ads they are producing.

Ask how many new hypotheses they are putting into market.

Then use variation testing to make the winners better.

That gives Meta more creative options while giving your team something even more important:

More opportunities to learn what actually makes people buy.

FAQ