At Data Never Lies, we work with clients from very different industries — from finance and healthcare to e-commerce and creative technologies.
But some projects stand out even for us. One of the most unusual cases we’ve ever worked on involved a virtual modelling agency that creates AI-generated digital models for films, advertising, and brand campaigns.
Their request sounded simple, but it turned into a fascinating exploration of human behaviour through data.
The Challenge: Understanding Attraction Through Data
The agency wanted to know how different audience segments react to different types of virtual models — which visuals attract attention, which drive engagement, and how preferences vary by gender, age, and geography.
This required a combination of predictive analytics, behavioral data modelling, and machine learning algorithms to identify consistent psychological and aesthetic patterns in audience reactions.
The Approach: From Raw Data to Behavioural Insights
Our team built a data pipeline that collected and transformed large volumes of behavioural data — user interactions, engagement metrics, and emotional response indicators.
Then, using business intelligence dashboards and predictive models, we analysed which features of virtual models resonated with which audience profiles.
We combined several analytical techniques:
- Audience segmentation — clustering users by behaviour, demographics, and engagement style.
- Feature correlation analysis — detecting which visual traits (facial symmetry, expressions, colours, poses) drive higher interaction rates.
- Machine learning models — predicting how different audience groups would respond to new AI-generated models before they even appeared in campaigns.
The Insights: What Makes Virtual Beauty Work
The results were both surprising and insightful.
Across almost every demographic group, audience preferences converged toward a few stable archetypes. In other words, what people find “attractive” tends to follow similar patterns — whether it’s real or virtual.
For the client, this meant not just better creative decisions, but also a data-driven framework for testing new virtual characters before launch.
For us, it became another example of how data science and human psychology can work together to reveal something timeless — the universal patterns behind aesthetic choice.
Beyond the Case: Why It Matters
This project showed how business intelligence and AI-powered analytics can be applied even in creative industries where intuition traditionally dominates.
By combining predictive analytics, data visualization, and behavioural modelling, companies can:
- optimise creative testing and audience targeting;
- measure engagement with precision;
- forecast reactions before campaigns go live;
- and build scalable decision systems based on actual behavioural data.
At Data Never Lies, we help companies turn complex datasets into actionable insights — whether it’s financial forecasting, customer analytics, or understanding what makes people click, choose, and connect.
If your company works with digital products, audiences, or creative campaigns and wants to bring AI-driven decision intelligence into the process — let’s talk.