Meta’s Andromeda update has redefined how brands reach their audiences.
Instead of relying primarily on traditional demographic or interest targeting, Andromeda clusters and optimises ads based on creative signals - how the content resonates emotionally, visually, and contextually with real users.
.png)
This shift toward creative similarity and entity-level clustering means that if your ads look and feel too similar, they get grouped under the same entity ID.
We’re saying, create diversity at the levels Meta actually cares about. E.g:
In case you missed it, we went into detail about how £100M brands scale with Andromeda here:
You should now operate under the theory that there is a perfect creative for each user. Instead of “Which audience fits this ad?”, the question becomes “Which specific ad version will resonate most with this single user right now?”
“Soul-mate theory” is just shorthand for Andromeda’s goal: every ad impression should feel as if the advertiser crafted that exact image and line of copy specifically for the person viewing it, and the system now has the scale and smarts to try.
But what does that actually mean in this context?
❌ Deterministic (Old Way)
The algorithm behaved like a rulebook.
If a user matched certain parameters (e.g., location, age, interest in “running shoes”), the system would follow a fixed chain of logic to choose which ad to show.
✅ Probabilistic (New Way)
The algorithm now behaves like a statistician + gambler + learner.
It doesn’t just look at hard rules; it weighs probabilities that a given ad will perform best based on vast, shifting signals: behaviour, history, context, creative, time of day, etc.
Two identical users might now see different ads, because the system is optimising based on the likelihood of engagement or conversion, not fixed logic.
This creates an algorithm that is:
👉 In plain English
Moving from deterministic to probabilistic means the ad-picking brain went from rigid “if-this-then-that” rules to a smart, ever-learning “best-guess” approach. Instead of guaranteeing the same ad sequence every time, it now plays the odds. Your customers get ads that feel more relevant, and you see better results, and the algorithm gets smarter.
Why it matters for e-commerce brands
Anyone who is involved with growing an e-com brand with Meta needs to understand this:
You need a high volume of distinctly different assets across all stages of awareness phases, underpinned by the psychological messaging frameworks that drive your core customers to convert. By doing this, you are more likely to serve the right ads, to the right people, at the right time.
BFCM has always been a high-stakes moment, but this year, we also have the added layer of Andromeda. In previous years, the standard BFCM playbook was to take a proven ad, run multiple versions of versions of it with a different headline, a different discount percentage, or a different CTA, and let the account find a winner through sheer weight of iterations.
However, under Andromeda, things are going to look a little different this year.
As mentioned above, we sat down with Maxim, a creative strategist at Meta who works exclusively with the platform's top 1% of disruptor brands, to get the inside view on exactly what's changed. In the episode linked, Maxim explained that Andromeda can now read inside the creative itself.
That's the key shift strategists need to plan around this BFCM: a set of ads that only differ by headline, offer wording, or CTA will now be read as the same asset.
And that means that instead of each variant reaching a fresh audience, they end up competing against each other for the same pool of people. This means your reach doesn't multiply with volume the way it used to, it just gets divided.
Maxim was direct about how strict this is in practice: even a UGC video with the same creator in the same setting, running a completely different script, in most cases won't register as genuinely different content. The system is looking for real signal - a different creator, environment, camera angle, or a genuinely different message.
During BFCM when you are trying to expand your audience reach, this is an absolutely crucial element to consider. The volume that actually earns reach this BFCM has to come from genuine creative difference - not from another twenty tweaks on the same ad.
Here's a nuance worth planning around specifically for the peak window: creative volume doesn't need to scale in line with spend the way it might feel like it should.
The logic runs like this:
The same logic runs in reverse outside the peak window - lower intent means lower hit rate and lower spend per asset, so the creative-to-spend ratio tightens back up once BFCM ends.
Two important caveats sit alongside this:
So, instead of asking "how many versions of this ad can we run," the question becomes "how many genuinely different people are we trying to reach, and does our creative actually reflect that."
That's where personas stop being a research exercise and become the entire foundation of your BFCM creative plan (more on personas in the next chapter).
The antidote to Andromeda is building real creative range around real audience differences - each one built for a distinct persona with its own motivator, its own fear, its own reason to buy. A brand with well-defined personas, each explored through a handful of genuinely different creative angles, gives the algorithm something to actually work with.
We've seen this play out directly with brands we work with.
When Pure Pet Food came to us, Meta was underperforming despite genuinely strong product-market fit the issue wasn't demand, it was that creative testing lacked the structure to generate real, compounding learnings. We started by identifying two distinct personas anchoring their audience: the Research-Led Realist, sceptical and looking for proof before switching, and the Worried Wellness Seeker, emotionally driven and needing reassurance rather than a product pitch. Each got creative built specifically for how they think, not a shared message stretched to cover both. Across 521 creative assets shipped in six months, that persona split helped drive CPA down 15% without pulling back on volume - proof that the diversity itself was doing the work, not just the extra output.
.png)
.png)
Voy, a men's health platform, faced the opposite starting problem: an audience that often didn't know (or didn't want to admit) it had the problem Voy's treatments solved. Generic education wasn't enough to cut through. The unlock came from moving into identity-led personas, most notably a "dad persona" built around real-life scenario storytelling and myth-busting messaging that directly reframed how men thought about their symptoms. Layering that persona discipline across Hair Loss and TRT, with a different emotional register for each, produced over 500 creative assets, a 71% increase in outbound CTR, and a 33% increase in purchases - because the creative was finally speaking to specific people with specific hesitations, not a broad, undifferentiated male audience.
And with TRIP, the barrier wasn't awareness at all - it was that a well-known brand was still widely misunderstood, leaving new audiences unsure why they should convert. Rather than more of the same creative, we diversified deliberately: founder-led content to give the brand a human, trust-building voice, and dedicated objection-handling creative to close the understanding gap directly. That shift away from a narrow reliance on a small set of assets, toward a genuinely wider creative mix built for prospecting, helped TRIP scale ad spend 206% during Dry January - their single biggest trading period - while CPA actually dropped 10%.
.jpg)
%20(1).jpg)
%20(1).jpg)
Three different brands, three different starting problems, but the same underlying fix that came from producing creative that was finally built around who was actually being spoken to.
Here's the part that matters most heading into BFCM: this isn't a peak-season tactic, it's a year-round discipline. The brands who show up to BFCM with a wide, genuinely diverse pool of creative aren't building that pool in October and November under pressure.
Instead, they've been running this kind of testing all year - continuously exploring new personas, new angles, new formats, watching what surfaces as a genuine winner, and building a bank of proven creative long before peak demands it.
BFCM isn't the moment to start diversifying your creative and testing new personas and big swings. Instead, it's the moment that discipline pays off hardest. 🚨
The brands who've done this work don't walk into BFCM with one winning ad and a scramble to create 20 variations of it. They walk in with an already-proven spread of creative, each piece already validated against a specific persona, each one capable of reaching a genuinely different pocket of audience rather than competing for the same one. When Andromeda is rewarding real signal diversity, that's exactly the asset a brand with a mature, always-on testing habit already has sitting in the bank - while a brand starting from scratch in November is trying to build that diversity under the worst possible time pressure, with the highest CPMs of the year attached to every mistake.
The payoff shows up most clearly at the bottom of the funnel, which is where BFCM often budget concentrates.
.png)
A narrow creative pool means retargeting the same warm audience with slightly different versions of the same message - a shrinking, increasingly saturated pool that fatigues fast during peak week.
A wide, persona-led creative pool means multiple distinct bottom-of-funnel audiences, each one primed by a different insight, each one still fresh when a competitor's narrower approach has already burned through its warm traffic. Incremental reach isn't just a top-of-funnel metric during BFCM - it's what determines whether you're still finding new, convertible bottom-of-funnel shoppers on day four of the sale, or recycling the same shrinking pool everyone else is also fighting over.
💬 OUR TAKE: The brands who win big this BFCM won't be the ones who reacted fastest to Andromeda in November. They'll be the ones who were already running genuine, persona-led creative testing all year, who've already surfaced real winners across a wide range of audience pools and who walk into peak week with incremental reach already built in, instead of trying to manufacture it under pressure.
Big brands typically operate with hundreds of creatives already live at any time - but as covered above, BFCM doesn't necessarily mean scaling that number further. It means making sure the assets you do run are genuinely distinct, since intent-driven hit rate is doing some of the heavy lifting spend-wise.”
1. Simplify account structure - don't build more campaigns, build more creative diversity inside fewer of them.
The instinct is to build complexity: multiple campaigns, multiple ad sets, granular targeting layers. Fewer, broader ad sets with a much wider range of genuinely distinct creative inside them give the algorithm more room to find and match audiences than the same spend spread across dozens of narrow, manually-targeted campaigns. For brands with the volume to run hundreds of creatives, the win isn't more account structure - it's feeding all that creative into fewer, better-fed ad sets and trusting the system to do the audience-matching work that used to sit in campaign setup.
2. Push Partnership Ads higher in your BFCM budget - and don’t isolate them into their own campaign.
This is the single highest-leverage lever available to brands with the budget to run it properly. Partnership ads consistently deliver lower CPA and higher CTR because they inject a genuinely new signal set - a creator's community, content style, and emotional register - that branded creative alone can't replicate. The mistake almost every large brand makes is siloing partnership ads into a separate campaign to "measure them properly." Instead, fold them directly into your existing BAU ad sets, so they lift the creative signal mix of the whole account rather than competing in a bubble of their own. (More on partnership ads later in the guide).
3. Feed volume with genuine signal variety - hundreds of creatives that share the same DNA is still one asset to the algorithm.
This is the trap big brands fall into most: mistaking creative quantity for creative diversity. Andromeda reads the actual content - talent, setting, camera angle, message - not the file count. Two hundred ads that are all the same creator, same setting, with only headline or CTA swapped will be read as a small handful of underlying signatures, all fighting each other for the same audience instead of reaching new ones. At this spend level, the volume advantage only pays off if every batch of new creative is genuinely engineered around a different persona, angle, or format - not just produced faster.
We connect our creative strategy to the Growth Compass:
Andromeda turns creative into a feedback loop; the Growth Compass ensures that loop feeds back into the wider growth system, not just the ad account.
Learn more about our Growth Compass and how it can accelerate your brand here:
Everything in this resource - persona-led creative, genuine signal diversity, testing structures built for scale - is the strategy. Executing it at BFCM speed, at BFCM volume, while everything else in the business is also demanding attention, is a different challenge entirely.
We've already helped brands do exactly this. When TRIP, the world's leading calming drinks brand, came to us ahead of Dry January - their single biggest trading period of the year.
The account had a clear problem: not enough creative volume or variety to prospect and acquire new customers at the scale the moment demanded. Sound familiar? It's the exact same bottleneck Andromeda punishes hardest during BFCM.
We rebuilt their approach around founder-led, objection-handling creative and a genuine increase in creative diversity - moving away from a narrow reliance on celebrity assets toward a mix built to reach new audiences at the top of the funnel. The results, delivered during TRIP's highest-stakes campaign of the year:
206% increase in ad spend
10% decrease in CPA
717 ads launched
Scaling spend during peak season usually costs you efficiency. TRIP scaled harder and got cheaper - proof that the diversity-led approach isn't just theory, it's what actually holds up when the pressure's highest.
"We brought in Soar to help us with performance marketing during a gap in internal resource. Soar ended up staying much longer and helping us navigate the dreaded Andromeda period. Special shout out to Lucia and Kit who really went above and beyond to support TRIP during a very busy time!"
— Annabelle Rogers, Acquisition Lead @ TRIP
If your BFCM plan is still relying on last year's creative playbook, this is the moment to fix that before peak.
Meta’s Andromeda update has redefined how brands reach their audiences.
Instead of relying primarily on traditional demographic or interest targeting, Andromeda clusters and optimises ads based on creative signals - how the content resonates emotionally, visually, and contextually with real users.
This shift toward creative similarity and entity-level clustering means that if your ads look and feel too similar, they get grouped under the same entity ID.
We’re saying, create diversity at the levels Meta actually cares about. E.g:
In case you missed it, we went into detail about how £100M brands scale with Andromeda here:
You should now operate under the theory that there is a perfect creative for each user. Instead of “Which audience fits this ad?”, the question becomes “Which specific ad version will resonate most with this single user right now?”
“Soul-mate theory” is just shorthand for Andromeda’s goal: every ad impression should feel as if the advertiser crafted that exact image and line of copy specifically for the person viewing it, and the system now has the scale and smarts to try.
But what does that actually mean in this context?
❌ Deterministic (Old Way)
The algorithm behaved like a rulebook.
If a user matched certain parameters (e.g., location, age, interest in “running shoes”), the system would follow a fixed chain of logic to choose which ad to show.
✅ Probabilistic (New Way)
The algorithm now behaves like a statistician + gambler + learner.
It doesn’t just look at hard rules; it weighs probabilities that a given ad will perform best based on vast, shifting signals: behaviour, history, context, creative, time of day, etc.
Two identical users might now see different ads, because the system is optimising based on the likelihood of engagement or conversion, not fixed logic.
This creates an algorithm that is:
👉 In plain English
Moving from deterministic to probabilistic means the ad-picking brain went from rigid “if-this-then-that” rules to a smart, ever-learning “best-guess” approach. Instead of guaranteeing the same ad sequence every time, it now plays the odds. Your customers get ads that feel more relevant, and you see better results, and the algorithm gets smarter.
Why it matters for e-commerce brands
Anyone who is involved with growing an e-com brand with Meta needs to understand this:
You need a high volume of distinctly different assets across all stages of awareness phases, underpinned by the psychological messaging frameworks that drive your core customers to convert. By doing this, you are more likely to serve the right ads, to the right people, at the right time.
BFCM has always been a high-stakes moment, but this year, we also have the added layer of Andromeda. In previous years, the standard BFCM playbook was to take a proven ad, run multiple versions of versions of it with a different headline, a different discount percentage, or a different CTA, and let the account find a winner through sheer weight of iterations.
However, under Andromeda, things are going to look a little different this year.
As mentioned above, we sat down with Maxim, a creative strategist at Meta who works exclusively with the platform's top 1% of disruptor brands, to get the inside view on exactly what's changed. In the episode linked, Maxim explained that Andromeda can now read inside the creative itself.
That's the key shift strategists need to plan around this BFCM: a set of ads that only differ by headline, offer wording, or CTA will now be read as the same asset.
And that means that instead of each variant reaching a fresh audience, they end up competing against each other for the same pool of people. This means your reach doesn't multiply with volume the way it used to, it just gets divided.
Maxim was direct about how strict this is in practice: even a UGC video with the same creator in the same setting, running a completely different script, in most cases won't register as genuinely different content. The system is looking for real signal - a different creator, environment, camera angle, or a genuinely different message.
During BFCM when you are trying to expand your audience reach, this is an absolutely crucial element to consider. The volume that actually earns reach this BFCM has to come from genuine creative difference - not from another twenty tweaks on the same ad.
Here's a nuance worth planning around specifically for the peak window: creative volume doesn't need to scale in line with spend the way it might feel like it should.
The logic runs like this:
The same logic runs in reverse outside the peak window - lower intent means lower hit rate and lower spend per asset, so the creative-to-spend ratio tightens back up once BFCM ends.
Two important caveats sit alongside this:
So, instead of asking "how many versions of this ad can we run," the question becomes "how many genuinely different people are we trying to reach, and does our creative actually reflect that."
That's where personas stop being a research exercise and become the entire foundation of your BFCM creative plan (more on personas in the next chapter).
The antidote to Andromeda is building real creative range around real audience differences - each one built for a distinct persona with its own motivator, its own fear, its own reason to buy. A brand with well-defined personas, each explored through a handful of genuinely different creative angles, gives the algorithm something to actually work with.
We've seen this play out directly with brands we work with.
When Pure Pet Food came to us, Meta was underperforming despite genuinely strong product-market fit the issue wasn't demand, it was that creative testing lacked the structure to generate real, compounding learnings. We started by identifying two distinct personas anchoring their audience: the Research-Led Realist, sceptical and looking for proof before switching, and the Worried Wellness Seeker, emotionally driven and needing reassurance rather than a product pitch. Each got creative built specifically for how they think, not a shared message stretched to cover both. Across 521 creative assets shipped in six months, that persona split helped drive CPA down 15% without pulling back on volume - proof that the diversity itself was doing the work, not just the extra output.
Voy, a men's health platform, faced the opposite starting problem: an audience that often didn't know (or didn't want to admit) it had the problem Voy's treatments solved. Generic education wasn't enough to cut through. The unlock came from moving into identity-led personas, most notably a "dad persona" built around real-life scenario storytelling and myth-busting messaging that directly reframed how men thought about their symptoms. Layering that persona discipline across Hair Loss and TRT, with a different emotional register for each, produced over 500 creative assets, a 71% increase in outbound CTR, and a 33% increase in purchases - because the creative was finally speaking to specific people with specific hesitations, not a broad, undifferentiated male audience.
And with TRIP, the barrier wasn't awareness at all - it was that a well-known brand was still widely misunderstood, leaving new audiences unsure why they should convert. Rather than more of the same creative, we diversified deliberately: founder-led content to give the brand a human, trust-building voice, and dedicated objection-handling creative to close the understanding gap directly. That shift away from a narrow reliance on a small set of assets, toward a genuinely wider creative mix built for prospecting, helped TRIP scale ad spend 206% during Dry January - their single biggest trading period - while CPA actually dropped 10%.
Three different brands, three different starting problems, but the same underlying fix that came from producing creative that was finally built around who was actually being spoken to.
Here's the part that matters most heading into BFCM: this isn't a peak-season tactic, it's a year-round discipline. The brands who show up to BFCM with a wide, genuinely diverse pool of creative aren't building that pool in October and November under pressure.
Instead, they've been running this kind of testing all year - continuously exploring new personas, new angles, new formats, watching what surfaces as a genuine winner, and building a bank of proven creative long before peak demands it.
BFCM isn't the moment to start diversifying your creative and testing new personas and big swings. Instead, it's the moment that discipline pays off hardest. 🚨
The brands who've done this work don't walk into BFCM with one winning ad and a scramble to create 20 variations of it. They walk in with an already-proven spread of creative, each piece already validated against a specific persona, each one capable of reaching a genuinely different pocket of audience rather than competing for the same one. When Andromeda is rewarding real signal diversity, that's exactly the asset a brand with a mature, always-on testing habit already has sitting in the bank - while a brand starting from scratch in November is trying to build that diversity under the worst possible time pressure, with the highest CPMs of the year attached to every mistake.
The payoff shows up most clearly at the bottom of the funnel, which is where BFCM often budget concentrates.
A narrow creative pool means retargeting the same warm audience with slightly different versions of the same message - a shrinking, increasingly saturated pool that fatigues fast during peak week.
A wide, persona-led creative pool means multiple distinct bottom-of-funnel audiences, each one primed by a different insight, each one still fresh when a competitor's narrower approach has already burned through its warm traffic. Incremental reach isn't just a top-of-funnel metric during BFCM - it's what determines whether you're still finding new, convertible bottom-of-funnel shoppers on day four of the sale, or recycling the same shrinking pool everyone else is also fighting over.
💬 OUR TAKE: The brands who win big this BFCM won't be the ones who reacted fastest to Andromeda in November. They'll be the ones who were already running genuine, persona-led creative testing all year, who've already surfaced real winners across a wide range of audience pools and who walk into peak week with incremental reach already built in, instead of trying to manufacture it under pressure.
Big brands typically operate with hundreds of creatives already live at any time - but as covered above, BFCM doesn't necessarily mean scaling that number further. It means making sure the assets you do run are genuinely distinct, since intent-driven hit rate is doing some of the heavy lifting spend-wise.”
1. Simplify account structure - don't build more campaigns, build more creative diversity inside fewer of them.
The instinct is to build complexity: multiple campaigns, multiple ad sets, granular targeting layers. Fewer, broader ad sets with a much wider range of genuinely distinct creative inside them give the algorithm more room to find and match audiences than the same spend spread across dozens of narrow, manually-targeted campaigns. For brands with the volume to run hundreds of creatives, the win isn't more account structure - it's feeding all that creative into fewer, better-fed ad sets and trusting the system to do the audience-matching work that used to sit in campaign setup.
2. Push Partnership Ads higher in your BFCM budget - and don’t isolate them into their own campaign.
This is the single highest-leverage lever available to brands with the budget to run it properly. Partnership ads consistently deliver lower CPA and higher CTR because they inject a genuinely new signal set - a creator's community, content style, and emotional register - that branded creative alone can't replicate. The mistake almost every large brand makes is siloing partnership ads into a separate campaign to "measure them properly." Instead, fold them directly into your existing BAU ad sets, so they lift the creative signal mix of the whole account rather than competing in a bubble of their own. (More on partnership ads later in the guide).
3. Feed volume with genuine signal variety - hundreds of creatives that share the same DNA is still one asset to the algorithm.
This is the trap big brands fall into most: mistaking creative quantity for creative diversity. Andromeda reads the actual content - talent, setting, camera angle, message - not the file count. Two hundred ads that are all the same creator, same setting, with only headline or CTA swapped will be read as a small handful of underlying signatures, all fighting each other for the same audience instead of reaching new ones. At this spend level, the volume advantage only pays off if every batch of new creative is genuinely engineered around a different persona, angle, or format - not just produced faster.
We connect our creative strategy to the Growth Compass:
Andromeda turns creative into a feedback loop; the Growth Compass ensures that loop feeds back into the wider growth system, not just the ad account.
Learn more about our Growth Compass and how it can accelerate your brand here:
Everything in this resource - persona-led creative, genuine signal diversity, testing structures built for scale - is the strategy. Executing it at BFCM speed, at BFCM volume, while everything else in the business is also demanding attention, is a different challenge entirely.
We've already helped brands do exactly this. When TRIP, the world's leading calming drinks brand, came to us ahead of Dry January - their single biggest trading period of the year.
The account had a clear problem: not enough creative volume or variety to prospect and acquire new customers at the scale the moment demanded. Sound familiar? It's the exact same bottleneck Andromeda punishes hardest during BFCM.
We rebuilt their approach around founder-led, objection-handling creative and a genuine increase in creative diversity - moving away from a narrow reliance on celebrity assets toward a mix built to reach new audiences at the top of the funnel. The results, delivered during TRIP's highest-stakes campaign of the year:
206% increase in ad spend
10% decrease in CPA
717 ads launched
Scaling spend during peak season usually costs you efficiency. TRIP scaled harder and got cheaper - proof that the diversity-led approach isn't just theory, it's what actually holds up when the pressure's highest.
"We brought in Soar to help us with performance marketing during a gap in internal resource. Soar ended up staying much longer and helping us navigate the dreaded Andromeda period. Special shout out to Lucia and Kit who really went above and beyond to support TRIP during a very busy time!"
— Annabelle Rogers, Acquisition Lead @ TRIP
If your BFCM plan is still relying on last year's creative playbook, this is the moment to fix that before peak.
.svg)
.jpeg)
