’s Ben Humphry Explains How Predictive Modelling and Data Can Boost Brand Campigns

Ben is one of Europe’s leading data companies specialising in digital target group marketing and brand engagement measurement. The company, which is owned by Deutsche Post DHL, say this provides sellers with higher yields, and buyers with high quality audiences – regardless of whether they’re using direct sales or programmatic buying. VAN caught up with Ben Humphrey, who is Managing Director, UK and Ireland, at, to find out how data is being used for brand advertising.

What’s the role of data in brand marketing?  What kind of uplift does it deliver for brands?

Brand owners have always focused specific brands on specific audiences – and the role of planning is to ensure the message reaches as many of that audience as possible. Data allows planners to perform that role more accurately than simply selecting media based on site and/or channel.  For a campaign run for L’Oréal to reach men aged 20-39 in the portfolio of four leading German publishers, research company Nurago compared our data-driven targeting with more traditional use of channels.

A comparative analysis then showed, that our data-driven solution was more successful in reaching the target audience – reaching up to 168% more than the channel-based alternative.

One of the things offers is predictive modelling? How does it work?

The technology works by taking a data source – often first party data, and extending it by matching patterns in user behaviours. This offers several advantages over first party data.

Firstly, it offers full reach in accordance to European data protection; all users have a pseudonymous profile, regardless of whether or not they have registered.

Secondly, it allows greater depth of knowledge – typically, registration data may have some basic demographic information or inferred information based on a postcode, whereas our models predict demographic, product interest and purchase intent.  We can model audiences based on bespoke campaign specific information and brand sentiment.

Finally, a prediction is dynamic, allowing the users’ profile to change as their interests or circumstances change. If two users use the same device, the prediction will adapt in real-time, whereas the first party data remains the same.

How are you coping with different interpretations of privacy laws in various European markets? 

We’ve ensured our solution is compliant with EuroPriSe, the European Privacy Seal, which surpasses the requirements of European Law. It is the only European seal by an official data protection authority and the targeting product is right now the only solution with a valid certification. One of the great advantages of a predictive model is that all our predictions are pseudonymous, meaning that they are not linked with an individual. We guarantee the highest degree of data protection, because neither nor its clients have access at any time to users’ personal data.

Subscribe to Weekly VAN Newsletter

Ad TechDataEuropeProgrammatic