Meta’s Algorithm Update and the Decline of Interest-Based Targeting in 2024

by Brennan Murphey

The advertising landscape has experienced significant changes over the years, with digital platforms like Meta (formerly Facebook) leading the charge in advanced, targeted advertising. Meta’s ads platform has long been celebrated for its precision, with interest-based targeting being one of its standout features.

Brands could rely on a wealth of data to pinpoint their target audiences based on their interests, behaviors, and demographic information, making it an invaluable tool for digital marketers.

However, recent changes to Meta’s ads algorithm in 2024 have significantly altered how interest-based targeting functions, leading many marketers to question the effectiveness of this once-powerful tool.

These changes primarily revolve around Meta’s shift towards what it calls the "interest graph," a more holistic approach that attempts to map users' broader behaviors and social connections rather than targeting isolated interests. While the concept may sound promising, the real-world impact of these changes has made interest-based targeting less effective for advertisers in 2024.

This article will dive into the most recent updates in Meta’s algorithm, explain how they affect interest-based targeting, and explore what it means for marketers in the evolving landscape of digital advertising.

What Is the Interest Graph?

To understand the recent changes, it's essential to grasp what Meta means by the "interest graph." Traditionally, platforms like Meta and Google have relied heavily on user data from engagement, likes, clicks, and shares to segment audiences based on their expressed interests. For instance, if a user frequently engaged with content about fitness, they would fall into a "fitness" interest group, allowing advertisers to target them with ads related to health and wellness.

In 2024, Meta's algorithm has shifted away from this narrow interest-based segmentation towards a broader, interconnected web of data known as the interest graph. Instead of looking at individual interests in isolation, Meta now attempts to understand users based on their overall activity patterns, social connections, and interactions within the broader Meta ecosystem (which includes Facebook, Instagram, WhatsApp, and Messenger). This means that users are no longer simply segmented into specific interest groups, but rather, their interests are analyzed in the context of their entire social network and behavior patterns.

Meta's interest graph uses advanced machine learning models to infer a user's preferences not just from what they directly interact with, but from who they engage with and what those people are interested in. For example, if someone in your friend group is particularly engaged with fashion-related content, Meta’s algorithm might infer that you could also have an affinity for fashion, even if you don’t explicitly express interest in it.

Why Interest-Based Targeting is Becoming Less Effective

While the interest graph approach aims to create a more nuanced and holistic understanding of users, the changes in how Meta defines and segments user interests have introduced a number of challenges for advertisers who previously relied on precise interest-based targeting.

Here are a few key reasons why interest-based targeting is becoming less effective in 2024:

1. Diluted Audience Segmentation

One of the biggest problems with the shift towards the interest graph is that it dilutes traditional audience segmentation. Previously, advertisers could select very specific interest categories (e.g., "outdoor running," "vegan cooking," "comic book fans") that reflected a clear, self-declared interest from users. The new interest graph, however, paints with broader strokes. Since users' interests are inferred through social connections and patterns rather than direct engagement, the segmentation is less precise.

As a result, advertisers might find their ads being served to people who are less aligned with the original intent of the campaign. For instance, a company targeting vegan consumers may find that its ads are also being shown to individuals who aren't vegan and are only passively connected to vegan-related content, resulting in lower engagement and conversion rates.

While the targeting tool in the Meta ads manager looks similar, it doesn't work the same way. Choosing an interest doesn't target the same people it used to because all interest-based targeting is now based on the new algorithm model of the interest graph, even if you use the "original audience options" instead of Advantage+ targeting.

2. Over-reliance on Social Influence

Another aspect of the interest graph is its over-reliance on social influence, which can be problematic for advertisers. While Meta’s assumption is that users with strong social ties share similar interests, this isn’t always the case. Just because someone’s friends are interested in a particular topic doesn’t mean they will be. This "guilt by association" approach has led to a mismatch in ad targeting, where users are being shown ads based on the interests of their social network rather than their own individual preferences.

This issue becomes particularly problematic in niche industries or highly specialized markets where social overlap is minimal. A user might engage deeply with niche topics that their social connections have no interest in, leading to missed opportunities for advertisers targeting those specific segments.

3. Increased Reliance on Behavioral Data

Meta’s interest graph also places more weight on behavioral data—like scrolling patterns, time spent on content, and indirect engagement—rather than on explicit interactions like likes, shares, or comments. While this shift aims to create a fuller picture of a user’s preferences, it has also led to less predictable ad placements. Behavioral data can be highly contextual and can vary depending on mood, time of day, or random scrolling habits.

By leaning heavily on inferred behaviors, advertisers risk serving ads to audiences that are not actually interested in the product or service, leading to inefficiencies in ad spend.

4. The Decline of Granular Control for Advertisers

One of the strengths of Meta's ad platform was its granularity, allowing marketers to slice and dice their target audiences with precision. In 2024, however, this level of control is waning.

With the interest graph doing much of the segmentation behind the scenes, advertisers are finding that their manual targeting efforts just aren't working reliably.

This loss of control has frustrated advertisers who once relied on Meta’s detailed audience segmentation to drive highly relevant ad campaigns to certain audiences and run split tests.

Some advertisers are even finding that split testing different audiences based on interests produces little difference between the interests, reduces the effectiveness of the overall campaign and increases ad costs.

Not too long ago, split testing interests was an essential part of ad testing strategy. In 2024, this older approach can be futile or even counterproductive because recent user behavior now has a much greater influence on who your ads are actually shown to than the interests you target.

Because of Meta's increased reliance on user behavior, machine learning and feedback loops to determine who ads are shown to, it has been shown that sometimes launching two identical campaigns to the exact same target audience produces wildly different results. The users who interact with your ads early on in a campaign and how they engage with the ads can greatly influence who else sees the ads later on.

5. Privacy Concerns and Data Limitations

Another major factor contributing to the decline of interest-based targeting is the increasing focus on privacy and data regulation. In recent years, Meta has faced mounting pressure to limit the amount of data it collects from users, leading to a reduction in the amount of granular data available for advertisers. As privacy laws like GDPR in Europe and CCPA in California continue to tighten, Meta has had to walk a fine line between offering valuable targeting options for advertisers and protecting user privacy.

With these restrictions, Meta’s algorithm now has fewer signals to rely on, making interest-based targeting less accurate than it was in the past. By shifting to the interest graph, Meta is trying to work around these limitations by inferring interests from broader, less invasive data sources, but the result is a more generalized—and often less effective—targeting approach.

The Impact on Advertisers in 2024

The recent changes in Meta’s ad algorithm have left many advertisers scrambling to adjust their strategies. Campaigns that previously thrived on precise interest-based targeting are now underperforming, with many advertisers reporting lower engagement rates, higher costs per acquisition, and decreased overall campaign effectiveness. In response, marketers are having to rethink how they advertise on Meta’s platforms.

Some brands are turning towards broader targeting strategies, leaning more heavily on lookalike audiences or focusing on high-level demographic and geographic data. Others are diversifying their ad spend across different platforms, like Google, TikTok, and programmatic advertising, where they still have more granular control over targeting options.

Additionally, many advertisers are placing a renewed emphasis on first-party data and building their own customer databases to circumvent the limitations of Meta’s new interest graph. By collecting data directly from their audience through email lists, loyalty programs, and CRM systems, brands can regain some of the precision lost in Meta’s new ad ecosystem.

One thing is certain - your ad copy now has a much bigger effect on the effectiveness of your ad because of how Meta's AI analyzes your copy, tracks who interacts with your ad and uses machine learning to expose your ad to different audience segments.

If your ad copy isn't congruent with other elements of your marketing or you're not careful
to turn off certain "optimizations" that Meta recommends by default, you can wind up getting clicks from the wrong people, confusing your market and winding up in situations where your campaigns barely perform.

Also, if your ad or landing page doesn't include enough text or the right text for Meta's AI to analyze, you can also wind up with lackluster results because despite your best efforts to target your ads at a certain audience, you're unwittingly making it harder for Meta's AI to gather enough data to show your ad to the right people.

Conclusion: The Need for Adaptation

The changes to Meta’s ads algorithm in 2024 mark a significant shift in how digital marketers need to approach their ad campaigns. As interest-based targeting becomes less effective, advertisers must adapt to the new reality of Meta’s interest graph and the broader focus on privacy and data regulation.

While the loss of precision in audience targeting poses challenges, it also presents an opportunity for marketers to rethink their strategies, explore new platforms, and invest in building stronger relationships with their customers through first-party data.

It also forces advertisers to hone their ad copy and write it not only for users, but for Meta's AI algorithm because the copy now primarily determines who Meta initially shows the ad to, not your audience selections.

Navigating this evolving landscape requires flexibility, creativity, and a willingness to experiment with new approaches. For brands that adapt, there is still immense potential to succeed in digital advertising, even as the rules of the game continue to change.