Rumble Industries - AdTech


A global technology company that provides advertising solutions for web and mobile based publishers realized that providing every user with the same ad-map was suboptimal, so they reached out to Rumble looking to personalize the user experience of ads for all of their clients. This approach not only had a significant impact on advertising revenue but also a critical improvement on user engagement — both for the client and the digital businesses it serves.

In order for the personalization campaign to succeed, users needed to be served with ads that matched their particular interest in a format that would appeal the most to them. Considering that the goal was to engage users with the ads being displayed, the client selected Click-Through-Rate (CTR) as the Key Performance Indicator (KPI) to be optimized.


Rumble Px offers plenty of Business Tools that companies in the ad-tech industry can take advantage of; in this case the client selected Content and Ad-Map Optionality – which allowed them to personalize the layout of pages, as well as the location and type of ads. This essentially meant serving different ad positioning and types based on optimal user responsiveness and click-through.

While the Rumble Px Platform has the ability to automatically determine the optimal segmentation mix at each moment during the campaign, the client decided to instead take advantage of the Custom Segments feature, which allowed them to create customized segments through several user data streams, such as contextual, demographic, behavioral, engagement attributes, content preferences, and promotion maps.

We built 3 versions of the ad-map, and for a short period of time, ran the three versions on all user segments (selected by the customer). This generated data for the algorithm to learn which ad-map is best suited for each user segment and in which context (hour of day, day of week).

We then trained a simple decision tree to determine the optimal ad-map version, for each case (segment + hour of day + day of week).




This is a snapshot of the training process using AzureMl

Once we had the trained decision tree, we could generate predictions for each case.

In the production environment we left 10% of the users as a control group with all ad-map versions. The control group not only served as a performance benchmark, but also allowed us to detect if the optimization group had a specific instance of lower performance (lower CTR than in the control group), we would be able to react by retraining the model based on that data.





In a very short time after the launch of the campaign using Content and Ad-Map Optionality to personalize ad-maps, CTR went from an average 0.57% to 0.71%. This positive increase in engagement proves how the combination of the right metrics and powerful data tools can help companies maximize digital revenue.




The graph shows the difference in CTR between users enjoying a personalized experience versus the control group. By implementing the Content and Ad-Map Optionality business tool, the client saw a rise of +25% in click-through-rates within the first week.