Roularta Creates New Revenue Generating Product Based on Look-Alike Modeling

About Roularta

Roularta Media Group is a Belgian multimedia group with over 1,308 employees that produces a consolidated turnover of circa 300 million euro. As a true multimedia group, Roularta’s family of brands spans newspapers, magazines, local media, digital products and services, as well as events, tradeshows, and books. Through Trends Business Information, Roularta also houses all financial and marketing data for every Belgian company.

Challenges

As a large, multimedia company operating in Europe, Roularta is rightly concerned about privacy when it comes to targeting on their own sites, and the products they build for partners. Needing to balance consent and privacy across a multitude of brands while safeguarding customer data and delivering value to their advertising partners, Roularta needed a solution that would help them centrally manage consent. Additionally, they need to ensure targeted advertisements were large enough to be valuable to advertising partners, sometimes, without overlap between consented individuals and those with matching interests. To solve for both, Roularta turned to BlueConic to help manage consent, their first-party data, and create models based on these data to continue to provide value to their partners.

Solutions

Privacy and Consent Solutions

Roularta works with BlueConic to include consent in their BlueConic profiles and synchs it with data from their consent management platform, Didomi. While Didomi powers the customer-facing interface for consent preferences, consent is stored and captured in BlueConic profiles. This allows Roularta to control which profiles get sent to ESPs and advertising platforms based on consent – helping them avoid costly fees and adhere to privacy regulations. If a consumer makes a request, Roularta can also easily and readily pull a cross-brand profile of this specific individual because of BlueConic profiles.

Additionally, working with BlueConic has had meaningful impact on how data gets passed to their partners. Originally, Roularta was working with OneTrust, but switched to Didomi because of its specific focus on publishers, the platform’s support of the IAB Transparency and Consent framework, and its ability to request and store consent cross-domains. Using BlueConic, Roularta was easily able to switch consent management platforms while maintaining consent in profiles they had previously captured through OneTrust, overriding those profiles with Didomi consent preferences once they were available. As a multi-brand publisher, the switch to Didomi took a few weeks, so being able to continue to adhere to consumer preference while making the migration was critical. Roularta also offers a premium product to personalize ads for partners by collecting interests on their site. Only if a consumer has provided consent to both Roularta and their partner will personalized ads show up – ensuring both Roularta, and their partners, are adhering to privacy regulations.

Using BlueConic’s consent management capabilities, Roularta doesn’t have to operate with an all or nothing approach to consent. They can store multiple objectives designed to delineate between objectives for retargeting, personalization, email opt-in, analytics and more in a profile. A person that has only rejected consent for retargeting, for example, means Roularta can still personalize this person’s experience on site to drive lead generation.

Leveraging first-party data for look-alike modeling

Roularta, like most media companies, achieves growth through audience monetization. However, some audiences are too small and specific for advertising partners. While these are highly engaged audiences determined by their behavior onsite as well as other attributes, Roularta needed to expand the audience in order to make it a worthwhile investment for their partners. For example, Roularta pulled an audience that was “interested in cars” that was 37K people, too small to sell as an audience to partners.

Turning to BlueConic, Roularta created realtime look-alike models using AI Workbench to score both known and unknown profiles. Online visitors are included or excluded from look-alike audiences as they browse the website, ensuring content visitors see is always relevant. The look-alike audiences are then sent directly to their ad channels to show specific advertisements. Creating look-alikes based on targeted profile attributes resulted in engagement rates on-par with their known audiences. Roularta can fine-tune look-alike models directly in AI Workbench based on performance data seen in Insights and Dashboards in the BlueConic platform. For example, they can define the threshold of certainty someone is a look-alike and use that to tweak their segments.

Results

Mitigate Consumer Data Risk with Consent Management

For consent management, Roularta centrally houses almost 31.3 million profiles with at least 89% of which have one consent objective attached. With consent in place, they are mitigating data risk where it matters, while still taking advantage of opportunities to use audience data where it has been consented to.

Expanded Target Audiences with Look-Alike audiences

In Roularta’s test of look-alike audiences for automotive segments, they saw a 6x increase of their target audience. Even with this tremendous growth in audience, Roularta saw similar levels of engagement from the look-alike audience and known-interest audiences. They saw their click through rates, their measure of engagement, have a relative difference of only about 10%.

Launches Dedicated Look-Alike Model Product to Boost Revenue

As a result of the success with other partners using look-alike models, Roularta now provides a dedicated product using their first-party data and look-alike modeling. Using BlueConic as a backbone to this new product, Roularta works with their partners to define target audiences, create custom segments, use real-time look-alike modeling, and activate those audiences with connections to Google Ad Manager and other ad platforms. They’ve seen their engagement increase across new segments in the market because of the relevant content they provide based on look-alike modeling.