Digital Horizons

Navigating Meta's New AI-Powered Ad Targeting: The Andromeda Update Explained

James Walker & Brian Hastings

Meta's Andromeda update has quietly transformed digital advertising, and marketers who haven't adapted are already falling behind. This groundbreaking AI enhancement to Meta's advantage plus ads automation function represents the culmination of Meta's staggering $65 billion investment in artificial intelligence technology.

At its core, Andromeda fundamentally changes how ads find their audience. Rather than advertisers defining demographic parameters, Meta's neural networks determine which creative assets will resonate with specific users based on their behavior, preferences, and position in the buying journey. As we discuss in this episode, this creates a dramatic shift in campaign structure - moving away from multiple segmented ad sets toward a single ad account with diverse creative that addresses different aspects of the customer journey.

The implications for brands are profound. Creative diversity becomes the central pillar of campaign success, requiring marketers to develop content addressing various information needs rather than organizing by funnel stages. Performance measurement also transforms, as users often need multiple ad exposures before converting. Early adopters report 25-30% reductions in acquisition costs, demonstrating Andromeda's powerful optimization capabilities.

This evolution aligns perfectly with changing user behavior, as people increasingly prefer consuming information within trusted platforms rather than visiting external websites. For advertisers, success now depends on in-platform storytelling and creative excellence rather than targeting expertise. Subscribe to Digital Horizons for more insights on navigating these transformative shifts in digital marketing and leveraging AI advancements for business growth.

The Digital Horizons Podcast is hosted by:

James Walker
- Managing Director WHD
Brian Hastings - Managing Director Nous

Speaker 1:

We always take it with a grain of salt when there's a new platform that says it's going to sort of take away all the hassle.

Speaker 2:

The biggest focus on this is making sure you've got enough creative diversity within your ad accounts, because an ad that talks to you may not talk to me. The way users are using the internet is changing, so it's been really rolling out since then, but it's taken a bit of a while of testing and really understanding what we need to do here to make it work. We're back.

Speaker 1:

I just hit the button and I'm like, oh, I've got to speak.

Speaker 2:

Yeah, yeah, welcome back to Digital Horizons. It's been a little while I've been over traveling a little bit, and so we tried one episode that was, I guess, me dialing in from Bangkok, and we felt that didn't really have the same impact as us sitting in a room having a proper conversation about things. So we put things on hold for a little bit and now I'm back and we're back at it.

Speaker 1:

Yeah, it's good to have you back. It's good to be back in the room and to not just be me sitting in the room looking at a computer screen.

Speaker 2:

Yeah, that would have been a strange situation. I mean me doing it on the computer.

Speaker 1:

I was reacting to that and then talking to the camera, but I'm proud of us for actually producing some content during that time. Yeah, we gave it a go.

Speaker 2:

But now that we're back and the conversation that we're going to be having today is around Meta's Andromeda update, which is having a massive impact in the way that we are now operating our managing ad campaigns and having a huge impact in the performance, I suppose I should start by saying this is Digital Horizons. Oh yeah, good plan, we're back, I'm.

Speaker 1:

Brian, this is James. We're the channel for business owners, marketers and AI enthusiasts, I'd say as well. It's definitely the direction we're going. Yeah, we're testing and trialing this sort of stuff to see what's working now and also what we should be focused on in the future. And today, meta's Andromeda, which is effectively a tool added to the ads platform and specifically to enhance the advantage plus ads automation function of Meta. What it does is help tailor the right creative from your bank of creative to the right person, based off their interest or likelihood of actually engaging with that bank of creative, to the right person based off their interest or likelihood of actually engaging with that piece of creative. It's not so much finding the right person who will like your product, but of a big bank of ads, which one is actually going to move the needle or engage them or work the best on that particular person using AI neural networks.

Speaker 2:

Yeah, so the way I'm understanding it is that it's really now got a lot more nuance and understanding as to where people are within their buying journey and which using its AI networks and capabilities. The recent data has come out that I think last year Meta invested 20 billion or something into AI upgrades.

Speaker 1:

You'd hope there's some sort of quality outcomes from that sort of investment, and they've just increased it to 65 billion.

Speaker 2:

So I think that they're really going all in, and we did a piece a little while back around how Zuckerberg is really wanting to really own the whole landscape in terms of push a button, get results in the ad campaign with zero import or any assistance from creative agencies or anything that want to really make it happen all on our own.

Speaker 2:

And I think that this is the first step in that, because it's really taking away any need for any kind of from a media buying point of view any kind of targeting. Your creative is doing the targeting and so, by getting the right creative that's going to speak to your audience, meta is going to know who to put in front of you. Don't need to tell her. All right, I want to show this ad to males, 45 to 55-year-olds that are living in specific regions. Meta is going to know, know, based on all the data that it has on all of its customers and all the different ad campaigns that are currently running, who should be seeing this ad right now.

Speaker 1:

Right. So that's the big point of difference there. It's not just the serving or the best performing ads being served to your defined audience, it's going and finding the right people for each of your ad types.

Speaker 2:

Absolutely, and so that's where really the biggest focus on this is making sure you've got enough creative diversity within your ad accounts, because an ad that talks to you may not talk to me. Yeah, so, if I've got, say, and depending on what part of the user journey I'm in so if I'm looking at want to go on another holiday I'm currently thinking about it I'm not looking at booking flights? Yeah so ads to me that, compared to someone who was just about to start booking flights right now, they're going to be very different in terms of what is going to the messaging that's going to matter to that person at the right time. So if you've got ads that are sort of thinking about all the different parts of the buyer's journey you're going to have, maybe you found a story, so you might have some product features videos. You might have some us versus them. You might show some benefits, all the different things and different angles. So that way you're covered.

Speaker 2:

When someone is thinking about or not, even at the beginning stage, you really want them to discover your ads. You don't need to structure it in a way that you know all right, this is my top of the funnel, this is my bottom of the funnel. This is this you just want to make sure you've got a wide variety of ads in there so that that way, when someone, when your ads are being delivered meta will then know all right, this is the right, the years gone by and the introduction of targeting people based off their interests and their buying signals.

Speaker 1:

But this is a huge leap forward and I think it'd be easy to sort of conflate it or confuse it with what we used to have with programmatic advertising, where it would effectively buy based off your buying signals what other sites you visited? Are you the right type of person to fit within the model of who we want to place ads in front of on a DSP or a demand side platform with display ads? And it would then monitor how many of those people, based on those buying signals, came back to the website and converted and then readjust who it's targeting.

Speaker 2:

Correct, so it's using the first party data for the conversion and rather it's targeting. Correct, so it's using the first party data for the conversion and rather it's basically almost adding labels to users based on their activities that they're doing across the internet, as opposed to the video content that they're consuming, and this is going Oracle, we know, based on what they're watching and what they're doing, then this is the right video to be showing them the right ad to be showing them.

Speaker 1:

And we're seeing from TikTok algorithm and the Meta Reels algorithm the amount of attention-based information these platforms can learn about you, what really keeps you engaged, what really holds you longer. The algorithm is growing so strong from the content you're consuming that they're the best people to know what is going to engage with you, based off the organic content that you're looking at and stalling on and how your algorithm's been building and making assumptions here, that it is using that sort of interest-based or attention-based algorithmic view of us to help with the ad platform or is it all based off ad stuff?

Speaker 2:

I'm sure it's taking into account all the huge amount of data in all different areas, but I think it's just added a whole different layer in terms of what the ability is for targeting the right ad to the right person. Yeah, I think the biggest shift, though, is really the way that we're structuring and setting up the campaign, so I've got a few, I guess, tips and ideas around what can be done to really take advantage of this new changes within the ad platform.

Speaker 1:

Yeah, it's been available since late last year.

Speaker 2:

Yeah, I believe so. It's been really rolling out since then, but it's taken a little while of testing and really understanding. Well, what do we need to do here to make it work?

Speaker 1:

Yeah, and at both our agencies we always take it with a grain of salt. When there's a new platform that says it's going to sort of take away all the hassle, I often think, oh, this is made for entry-level advertisers who just want to click a button. But this one is turning out to be actually quite useful and helping us to target or select creative or serve the right creative in ways that we just couldn't do manually with manual optimization. So let's get into the tips of what you should be doing to utilize this within platform 100%.

Speaker 2:

So I think what you want to be thinking about is that people are less likely to even visit your website right now after watching an ad. I think they're much more likely. What you want to be doing is thinking about the way that you're structuring your ad account. So you really just want single ad account, single ad set, and then think about a whole journey and all the different angles that you can put on there, and just different angles that you can put on there, and just imagine that they never go to your website until they're ready to buy.

Speaker 2:

So think about all right, well, what does my services page look like? What's all the information I can pull out of that to give them that information? Then the next touch point might be all right, well, what's the features and benefits of my service? Next touch point might be some UGC content, giving some credibility to it, showing people how they're the business, all this kind of thing. And so you think about that journey of all the different information, really what websites really are. But you want to put that in a format within your ad campaign so that some people aren't going to give a shit about your founder story. They just want to know the features and benefits. They just want to go all right, cool, that's what I need right now. I want to buy that thing, but other people they really need to be, I guess go through the process and see multiple touch points.

Speaker 1:

Yeah, have an opportunity to connect with the brand and build trust.

Speaker 2:

Correct, correct, instead of expecting that they're going to click off and go to the website because people's behaviors are changing as well where they want the information.

Speaker 2:

Now they don't really want to have to click off and go discover and take it from the self. They want to be fed that in the feed. And so, by treating it as a guess of journey, that's how Meta's looking at it. It gives a full picture and it gives a full funnel flow of your ads in the single campaign. Now challenges that's going to give you then is all right. Well, how do we optimize? Because if you're getting a conversion of CPA on one of your ads, I know that previously the way we would approach it is like we would have benchmarks and be like all right, cool, if an ad's not achieving X results, fucking kill it. But right now that's not going to be the case, because someone might see that founder video, they might see the next video, next video, and it's a sequence of that journey that they're going through. And so when you get to that last ad, you might see all right, well, all these other ads didn't get the conversion, this one did so kill all those ones, but they were instrumental in making that happen.

Speaker 1:

They all work together, have a place to be there to serve, just in case it's part of the journey that that particular customer needs to go on Exactly.

Speaker 2:

And so it changes the whole way now of media buys and how to optimize and manage campaigns. And so the way that we've been finding that is really what should be doing is looking at all right, cool is Meta. Because Meta is really going to self-regulate this as well. Because they might be like all right, cool, we know that this ad is a core piece within this flow to create that conversion, so they're going to continue spending on that ad. If you're running ads and you're seeing all right, cool, last week it spent $500. This week managers are only giving that ad $20. Then that's managers almost saw all right, well, this ad's not working. And then that's where you start looking at Oracle. Maybe we need a new variation or a new version of that ad because it's not hitting the mark. But these other ads are still getting a lot of budget, they're still working, they're consistent or they're increasing. It's showing that Meta's saying, yep, this is playing a much more stronger part of the funnel.

Speaker 1:

It's really quite clever from Meta when you think about their environment. Ads serve them so much revenue, but they also take people away from their platform. So the strategy from Meta helps or seemingly helps the advertisers, but it's also helping Meta keep everyone in that one place. We're not extracting them out for as cheaply as possible to get to a website, capturing that lead data and then never having to pay Meta again. You're having to do every one of those touch points in platform and never allowed as a user or a user of Meta to leave. That's right, that's it, and we have to just keep paying for the opportunity to build that brand frequency and that brand story using these platforms.

Speaker 2:

That's right. But I guess the way that Meta also then, by having all that data to build that brand frequency and that brand story using these platforms, that's right. But I guess the way that Meta also, then, by having all that data within the same campaign, same ad set, giving it the freedom to be able to do that, it really then should be like all right, well, meta's going to have a good understanding of all right. Well, this is a picture of this kind of person and we know if they watch this video, this video, this video, this video and this sequence there sequence, there's a very high chance that they're then going to click off and purchase, yeah, and so it gathers all that data and from the initial test that we're seeing, we've seen like drops of sort of 25 to 30% in overall CPAs, because it's really learning and understanding how to get the most out of the content that is being fed to it, to the right people at the right time.

Speaker 1:

I think, while it's clever from meta and they're sort of trying to capture their audiences in their platform, the way users are using the internet is changing. It's not like everyone's wanting to jump off and browse dodgy old websites. I'm sick of shitty websites, I'm sick of using them and if I can get the answer in a platform I'm comfortable with, I will. So, yeah, I think they're reacting to the changing expectations of users that they want to continue using the platform they've opened. It's a big strain on digital marketers, advertisers, brands, small businesses that for years have been focused on just your site and how your site presents this story organic search, everything being an ad click bring to your website. Then they'll go away, then they'll come back via organic and they'll do a little bit more research All that research happening on your site, as you've said, with these platforms and I presume it's going to pair well with the similar approaches in TikTok, which they don't have this yet, but I'm sure they'll follow suit You're going to have to be building these brand narratives in platform and it will be instrumental to campaigns and brand campaigns not just performance campaigns to be present in that environment.

Speaker 1:

It's just getting so much of our attention, either TikTok or Meta, yeah, like it's where people go to for their news, it's where people go to for product recommendations, and then if I'm jumping out, I'm jumping into chat, gpt-5 for my sort of research, and then, you know, maybe Googling for a local search. It's a whole new wild world when you think of the differences in how we used to structure our digital strategies in the past to now.

Speaker 2:

I'm really excited for what Andromeda is going to bring, being that within our agency, we do have a creative team. I think that creative has obviously been such a more and more important part of the process, and it's something that we're really investing heavily in and making sure that we're able to deliver the best possible. I guess creatives for our clients. It really shows that all right, I think we've made the right decision in focusing on this, because it's what's going to really get the results in the meta campaigns, because it's not now about being who's the best at being able to pick the right audiences and targeting and all this kind of shit. It's really just about all right, can you create creatives that are going to speak to the audience and you really understand what it is that they need and making sure that you hit those points?

Speaker 1:

Yeah, I think there is going to be a massive expectation for brands who want success in the meta platform to be creating and to be bold with their creative and their ads so that they have more and more options to test and, yeah, some of them will fail, but they need to be in there and to be tested to see if they're failing and sometimes failure might not be measured how we used to measure failure. Like you were saying before, it might be just a trust-based ad that is part of a user journey. So I think we're going to see a lot more creative types and brands exploring different creative paths, rather than a similar ad type that you always see from, say, jb Hi-Fi or whoever it might be. There's going to be various different versions of creative paths within the Meta platform and that's something we'll have to keep supporting our clients with and building out.

Speaker 2:

Yep. Well, this has been another, well, a long-awaited week for Digital Horizons, but we're going to be, as we always say, much more consistent. Thank you for your time again watching.

Speaker 1:

We're really enjoying getting back into this and hope you've enjoyed the content. Yeah, thank you.