Audience to Content: Driving the Next Step in Targeting

The paradigm shift in targeting on social media powered by the fast-paced evolution of AI

The Creative Gold Rush is an Immense Opportunity

Social media advertising has entered a new era. Where advertisers once defined precise targeting inputs, platforms now decide who sees what - a shift most visible on Meta over the past year. 

In late 2024, Meta introduced Andromeda: a machine learning system that matches the right ads with the right users, at the right time, in the right place. The result - higher ad quality, stronger personalization, and better outcomes for both users and advertisers. 

For marketers, this means content is now the targeting. Once, ads followed the targeting. Today, targeting follows the ads. 

This change demands a new mindset: integrating personas, concepts, and formats strategically to ensure relevance. We must return to our marketing roots - producing content that communicates meaningfully, not mechanically. 

Winning in this environment requires creative breadth over creative volume - tailoring content to audiences, aligning with the algorithm, and letting diversity drive discovery. 

Those who adapt quickly - and pair AI-driven efficiency with human creativity - will outpace the competition. The advertisers who embrace this shift early will define the next era of performance. 

Implications for Brands: What this shift means in practice 

The shift from input-based targeting to creative - based delivery changes how brands plan, produce, and evaluate content. 

To stay relevant - and ahead - advertisers should focus on six key actions: 

1. Diversify creative output: Move beyond microterations. Build distinct assets with different messages, visuals, and emotional hooks to reach new “audience pockets.”

2. Think in concepts, not formats: Creative diversity isn’t just about volume - it’s about variety. Test new narratives, tones, and value propositions, not just layout or color tweaks.

3. Reconnect brand and performance: Brand equity fuels performance. Balance long-term resonance and short-term conversion to create sustainable growth.

4. Design for awareness levels: Develop content for audiences at different stages - from unaware, to problem-aware, to ready-to-buy - and let Meta's algorithm find the match.

5. Combine AI with human creativity: Use AI for speed, scale, and insight - but rely on human judgment for storytelling and emotional connection.

6. Simplify structure - amplify learning: With algorithms handling delivery, focus less on complex campaign setups and more on iterative creative testing and insight loops. 

THE GOOD OLD DAYS OF “INPUT TARGETING"

One of the core strengths of advertising on social media has always been the ability to target a select audience very specifically, based on predefined platform data inputs available when setting up campaigns and defining a desired audience. 

On Meta, this has always been referred to as “interest targeting” – an option for advertisers to select keywords representing the actual interests of each user on the platform when building an audience. If you sell hiking boots, targeting users interested in “hiking” would produce a highly relevant audience interested in that specific category, making the ad experience more relevant and driving a better business outcome for both the advertiser and Meta. The same goes for TikTok and Snapchat. 

On LinkedIn, we uniquely have the option to target by job attributes - such as which companies users work for, their job experience, and their job title. In contrast to Meta, TikTok, and others, LinkedIn has always been a powerhouse when it comes to targeting by professional attributes - because it’s the only platform where we rigorously and continuously update our professional lives. All this could be summed up as “input targeting.” 

But today, we’re seeing a paradigm shift on the largest platform, Meta - where the ads once followed the targeting, the targeting now follows the ads. 

Due to the massive increase in content volume on Meta, choosing the right content for the right person at the right time - and in the right placement - has become a growing challenge. Meta wants users to have an experience that feels relevant to their specific interests and behavior. Processing millions of pieces of content in real time has made it increasingly difficult to keep both users happy with their feed and advertisers satisfied with their performance. 

While “interest targeting” was a good start, Meta had to bet on a new set of tools to handle both the increased volume of ads and the growing number of users. 

Relevance is key - but how do you maintain relevance when the volume of data exceeds the threshold of what the system can handle?

HOW META MAINTAINS RELEVANCE: THE ANDROMEDA UPDATE 

Meta has always been at the frontier of advertising on social media - and the latest update, Andromeda, is a testament to that. 

Although the media and marketing communities are currently flooded with discussions around this “new” update, Meta actually announced Andromeda back in late December 2024 under the headline: 

“[…] the next-gen personalized ads retrieval engine.” 

Meta describes Andromeda as: 

“Meta’s proprietary machine learning system designed for retrieval in ad recommendation focused on delivering a step - function improvement in value to our advertisers and people.” 

In essence, this means Meta is now much better at pairing the right ads with the right users at the right time. It does so both faster and more effectively - ultimately increasing relevance, generating stronger business outcomes for advertisers, and improving the platform experience for users. 

Previously, Meta could only apply limited personalization when deciding which ads to show a user. 

Now, the company has developed system infrastructure capable of scaling model capacity 10,000 ×, enhancing personalization and delivering an 8% increase in ad quality. 

Over the past year, Advantage+ features have been rolled out aggressively within Meta’s Ads Manager, nudging advertisers to lean into the Andromeda logic. 

As of now, Andromeda has been fully rolled out - meaning advertisers must now adopt a new way of advertising.

WHAT IS THE “NEW WAY” OF ADVERTISING? BIG CHANGE MEANS BIG OPPORTUNITY 

To understand the implications - and opportunities - that Andromeda unlocks, we first need to understand how the algorithm fundamentally delivers our ads. 

When setting up a campaign, we choose an objective such as leads, sales, traffic, or reach. 

Once the ads are launched, the algorithm predicts a small portion of the target audience most likely to convert in line with the chosen objective. It then expands delivery portion by portion, from the most to the least likely to convert. 

This means that although we might set up campaigns with specific targeting, we may only reach a few “audience pockets” if the algorithm identifies only a small subset of users who would find our ads engaging. 

Meta aims to deliver on the objectives we set as advertisers - but also to maintain a relevant experience for users. If our ads only resonate with a small portion of the audience, we risk over - saturating a tiny corner and limiting our ability to scale. We also risk being penalised in the auction, paying more to reach less relevant audience pockets. 

With Andromeda, Meta now evaluates ads differently - calling for a more strategic and methodical approach to creative diversity to ensure our ads surface across more audience pockets. Ads that look alike are now treated as one asset under Andromeda, which limits reach within the desired audience if creative diversity is not increased. 

For a long time, we’ve talked about creative volume - iterating ads to find winners to bet on. But with Andromeda, advertisers now need to take bigger swings when producing creative, ensuring broader diversity to fully capitalise on this new system.

HOW ADVERTISERS SHOULD ADOPT CREATIVE DIVERSITY: FROM CONTENT BEING PART OF THE FUNNEL TO BEING THE FUNNEL 

The “funnel” has always been integral to marketing. But with user journeys now more complex than ever - due to cross-media behavior and privacy initiatives like iOS14 - forcing a linear funnel onto digital advertising has become a challenge. We can no longer follow every step of the journey with precision and must therefore think differently when producing and distributing content on social media. 

However, Meta still knows a lot about its users - thanks to the “fingerprints” they’ve left historically and continue to leave in real time - creating an opportunity for advertisers to leverage this through content. 

Most of us have experienced a scenario where we visit a website - maybe we’ve been looking at shoes - and then our feed is suddenly flooded with ads from shoe brands. 

Over time, many users have also come to feel that Meta “listens” and shows ads for products they haven’t even searched for yet. 

While there are many opinions and theories about this, one thing is certain: the algorithm has become far better at understanding user behavior, patterns, and interests - and at using that information to increase ad relevance. 

With Andromeda, Meta not only knows that you like shoes as a category. As Meta puts it themselves: 

“Imagine having a personal concierge who knows your tastes so well that they don’t just understand that you covet shoes, but that you like to wear red flip-flops at the beach.” 

This is the power of Andromeda - but also the opportunity for advertisers to lean into this new paradigm shift in creative targeting. 

We need to integrate broader concepts, angles, and formats into our advertising to ensure that we reach as many sub-segments of our audience as possible. It’s no longer enough to change the color of the font - or the font itself. We need to make bigger creative swings to stay ahead. 

Another key consideration is the awareness level of our audience: are they problem-aware, or solution-aware? As Meta encourages advertisers to consolidate campaign structures and let the algorithm take the lead, we’re moving away from the classical, meticulously mapped-out funnels - and toward letting the system distribute our ads in the most effective way possible. 

THE CORE LEVERS OF CREATIVE EXCELLENCE - BALANCING BRAND-EQUITY LED INITIATIVES AND PERFORMANCE 

Although advertisers might fear the Andromeda update, we believe this change was bound to happen - and that it forces advertisers to take a step back to the roots of marketing: communication. 

When we think about it, we are now "forced" to address how we communicate and to whom. 

This calls for dusting off the marketing theories we were taught in school to ensure our digital marketing initiatives remain successful. We need to put more thought into how we produce content - how we forge both resonance and relevance. 

Performance marketing has long been the main focus for advertisers - optimizing for margins - which has led many brands to neglect or erode their brand equity. 

A recent study from WARC.com shows that performance heavily relies on brand equity. They refer to this as the "multiplier effect." Brands should not think in terms of performance vs. branding, but rather performance x branding as an equation. 

This circles back to the section about creatives being the funnel. Not only do we need to tailor our communication to our audiences and their subsets - we also need to ensure we compose ads and content that speak to the different awareness levels of our audience. 

In marketing, the 95:5 rule is one of the fundamental principles: it describes how only 5% of your total addressable market is "in - market" at any given time, leaving 95% not yet ready to engage or buy. Essentially, this means advertisers need to address both sides to drive sustainable growth. 

We must resonate with the 95% not yet ready to buy (brand-building initiatives), so that when they are ready - representing the 5% - our content feels more relevant and becomes the preferred choice. 

In other words: 

1. Resonance connects (branding) 

2. Relevance converts (performance) 

This is not only a fundamental part of marketing, but now also a criterion for running successful campaigns on Meta, as the platform becomes increasingly intelligent in matching users with the right content, at the right time, in the right place. 

FINAL NOTE ON THIS PARADIGM SHIFT: NO NEED TO PANIC 

In the marketing world, we tend to panic when shifts occur - however, with this new update from Meta, there’s no need.

Andromeda has been subtly rolled out over the past year and has already been shaping how things work for most advertisers. It’s not new - but it’s important to consider how, as advertisers, we can best utilize it.

One of the key challenges for most brands has always been producing enough relevant content for digital campaigns on social media.

The important thing to keep in mind is that every advertiser now needs to adjust and adopt a new mindset when it comes to producing and executing social media campaigns.

Also, while this primarily applies to Meta, TikTok usually follows suit, and its algorithm has long been effective at matching users with content that resonates and feels relevant to them.

LinkedIn, however, is different. Its key advantage lies in its unique ability to target by job attributes. Classic input targeting remains highly relevant on this platform.

But regardless of each platform’s maturity, building content thoughtfully - with your audience in mind - has always been an edge for advertisers when done right. That’s not new, and it won’t change.

What this new paradigm does, however, is force advertisers to bring that thinking to the forefront.

Those who lean into this early will have an incredible edge.

And even more so, advertisers who integrate human creativity with technology and AI to increase content production efficiency will be miles ahead of the competition.