Why signals are set to outpace predefined audiences in 2025—and how to stay ahead

There have been several great leaps forward in advertising over the past decade, but one of the most profound has been the way brands connect with audiences. Today, campaigns often rely on predictive, predefined segments, where brands outline their ideal audiences based on demographics or interests and match these profiles to written audience definitions within a demand-side platform (DSP). While this approach provides an intuitive approach to selecting the right audiences, it’s inherently flawed—it’s manual, generic and prone to missing critical customer subsegments. More importantly, it’s static, and thereby unable to adapt and reach audiences at the precise moments when ads would be most impactful and relevant. However, there’s another promising approach that can complement segmented targeting by layering additional insights, enhancing campaign performance and delivering even more impactful results: a blended strategy that combines behavioral and contextual signals. Artificial intelligence supercharged both targeting decision making and advertisers’ ability to dynamically understand every ad opportunity and reach audiences in real time. Today, targeting is less about predefined segments and more about integrating and balancing every available signal: behaviors, contexts, interests and even intent. It’s no longer just “Does this audience description match my target audience?” but “What signals indicate this ad opportunity will lead to a desired action?” With AI, marketers set key target parameters—such as desired outcomes, budget and brand safety guidelines—and then rely on DSPs to autonomously optimize the mix of creatives, targeting strategies and supply sources. This move toward automation with human insight is gaining momentum with ad tech providers across the industry, as it significantly reduces the need for manual intervention. The power of blended targeting Advertisers have a range of targeting options for reaching their desired audience, including behavioral, contextual, geo, device-specific and more. Among the most commonly used strategies, behavioral targeting enables advertisers to select audiences based on standard definitions that leverage past behaviors to infer consumer intent. Contextual targeting, on the other hand, uses diverse taxonomies or specific keywords to match ads with real-time user interests. After implementing these targeting approaches, advertisers typically use ROAS as a performance indicator for optimization and further media investment. The tendency is to then allocate more resources to the perceived winner and scale back on the less effective method. However, an Amazon Ads analysis revealed that this winner-take-all strategy often falls short of achieving optimal results. Instead, a combined approach that leverages the strengths of both methods can yield significantly better outcomes, as each technique highlights different audience dimensions and taps into what they are most likely to respond to effectively. According to Amazon’s 2024 in-house research that analyzed all active campaigns across a diverse range of U.S. advertisers and industries, advertisers that combined multiple targeting strategies within a single campaign were able to fully leverage Amazon Ads’ comprehensive insights into shopping behaviors and relevant ad moments. The results were striking when compared with campaigns using only a single targeting strategy: Advertisers who implemented both behavioral audiences and contextual targeting ad lines within the same campaigns saw CPMs decrease by up to 19.4%, while ROAS improved by an impressive 38.5%. Combined targeting in action These results highlight the advantages of a combined targeting approach over single-strategy campaigns. By integrating behavioral audiences (which leverage past actions to infer intent) with contextual targeting (which uses real-time content alignment), brands can adopt a more effective strategy for delivering their message to the right audiences at the right time. Here’s an example. Imagine a DSP using AI to help a luggage brand drive conversions and sales. For consumers who recently searched for vacation deals and browsed flights (behavioral signals), the DSP might show an ad for luggage while they shop for groceries. This will subtly remind them of the upcoming trip. Later, when these consumers browse tourist activities online (contextual signals), the system could deliver another ad highlighting the luggage’s travel-friendly features. An extra impression at that moment may seal the deal, capturing an opportunity that might otherwise have been missed. The combination of these two relevant moments creates a seamless, impactful journey that can ultimately help drive the purchase. This balanced, real-time blend of signals and targeting strategies ensures advertisers can capitalize on every ad opportunity, delivering smarter and more impactful campaigns. The future is relevancy at scale As we look ahead, the future of targeting is not just about better technology—it is about rethinking how advertisers connect with their desired customers. By balancing precision with adaptability and blending AI automation with human insight, brands can create campaigns that feel natural, timely and genuinely relevant. This shift challenges advertisers to move beyond traditional methods, leveraging every signal available to deliver value. The winners in this new era won’t just be those who master targeting strategies but those who build meaningful, lasting relationships with their customers—one real-time, personalized connection at a time.

Why signals are set to outpace predefined audiences in 2025—and how to stay ahead

Featured Stories