Soft attributes like psychology, allegiances, values, and networks are the frontier of audience intelligence and marketing research.

They are proving to be far more effective and valuable in our clients campaigns than traditional hard attributes such as demographics and socioeconomics. But, to use the effectively, to extract them accurately, and to obtain them at scale, you need to engage Machine Learning and AI based systems to overlay the data ONTOP of the traditional Hard data points.

There is no ‘off the shelf’ tool to do this. We have enabled this approach through the Data Kinetics Audience Intelligence platform ive been working on for the last coupe of years. Its not perfect, its needs ‘humans in the loop’ to check the AI outputs and tweak, but its proving highly valuable to our global brand clients.

Usually, research and insight agencies generate audiences for brands by focusing on "hard" data points—age, gender, income, and occasionally some "life stage fluff." This includes superficial commentary about how people live and what might be important to them based on their life stage. But while this approach provides a baseline understanding of a target audience, it lacks depth, nuance, and, most importantly, context. It is often too generic and fails to capture the complexity of human behavior and motivation.

In contrast, creative, cultural, and target market nuances are often driven by soft attributes—values, psychology, allegiances, and beliefs. These factors offer a more detailed understanding of an individual’s desires, fears, and motivations. For example, knowing what drives someone to buy a new product—whether it’s peer pressure, family influence, brand loyalty, or cultural identity—gives marketers a more accurate way to predict and influence consumer behavior. Soft attributes reveal not only what consumers want but also why they want it, and this insight is critical to generating a more meaningful response from the audience.

The Data Kinetics tool I have helped design identify’s patterns in Audiences based on these attributes and then groups them as "tribes." https://liveandbreathe.com/data-kinetics

The AI can detect these tribes based on predefined sets of soft attributes and then we know the most effective messaging—whether that’s through words, imagery, or storytelling—that is likely to resonate with that particular group.

These messages are pre-tested to ensure they align with what is known to work. The focus is on understanding the psychological and emotional underpinnings of the audience, rather than simply segmenting them by age or income bracket.

Of course, AI’s ability to understand people based on these soft attributes is still evolving. While the technology is quite good at recognizing individual traits once it has ingested a person’s language and behavior (both online and in the real world), it is far from perfect. True human behavior extends beyond digital activity.

It involves complex dynamics like mindset, relationships, and emotional triggers—what gives someone the confidence to try a new product, or what fears hold them back. We are starting to understand how to identify and extract these attributes from digital behaviour as swell. Understanding for example if peer validation plays a stronger role in purchasing decisions as opposed to family influence, or perhaps the allure of other brands? These are the questions that soft attribute analysis seeks to answer.

To enhance AI’s accuracy, we use a hybrid system that combines machine learning with human expertise. As I mentioned in a recent interview with Bloomberg, while it’s inevitable that AI will make mistakes, combining human judgment with machine recommendations helps minimize errors. Domain experts are vital in this process—they check the data and refine the meta-tribes that the AI identifies, ensuring that the results are grounded in real-world understanding. This combination of human oversight and AI analysis creates a more balanced and reliable approach to audience segmentation and marketing strategy.

Even in its current form, soft attribute-led audiences are proving to be far more valuable than those based on hard data alone. By focusing on psychological and emotional drivers, we can more effectively match opportunities or address challenges within a category. We gain a clearer understanding of what makes an audience tick, allowing us to develop creative solutions that truly resonate. For example, are individuals driven by a fear of rejection, a desire for community, or a need to stand out? Or are they more likely to seek approval and conformity within their social group? These insights help us craft marketing campaigns that tap into core human emotions, rather than relying on broad assumptions. This gives clear direction to creatives and content base don hard data.

Ultimately, it is this deeper understanding of soft attributes that unlock the motivations and desires that truly drive consumer behavior, enabling marketers to create more personalized, impactful content. Hard attributes may still have a role to play, but they are no longer sufficient on their own. The future of audience intelligence lies in the subtle, complex world of soft attributes, where the true drivers of human behavior can be found.

The next stage of for us is to be able to respond to these at scale and personalize them, but that needs full in integration with generative AI content creation. We are working on that but the tools to date are not quite there yet to deliver consistent ‘brand safe’ outputs - but its close . But that’s for my next newsletter!