Personalized push message integration
based on big data analysis

The major Dutch media brand owned by Sanoma Corporation, has been using Pushwize for mobile push notifications since 2013. During this joint journey, benefited several times from the flexibility of Pushwize solution, for instance in relation with projects aiming at personalized user experience.


In 2018, approached Pushwize with two completely new ideas that aimed to utilize’s custom big data system as a source of notification sending. Previously, the big data system ran completely independently from Pushwize, collecting all user analytics from the mobile applications and websites. The team was busy creating precise customer segments and insights, resulting in hypotheses and learnings for new marketing initiatives as well.

Two of these initiatives were sent to Pushwize, for which aimed to discover technological feasibility. Firstly, data scientists found that users may actively use the applications but do not neccessarily subscribe to notifications for each category or topic they read. Secondly, users tend to spend the majority of their time on few locations, typically at home and at their workplaces. However, notification sending was only in practice using the last active location of users. 

Pushwize was requested to create custom integrations that enable notification triggering directly from this database.


The integration of the systems was started by storing the notification identifier (device secret) of each user in the big data system of Storing the IDs was important for creating the missing link between the systems and for finding a feasible parameter for notification addressing. 

Pushwize had already enabled message addressing based on the device secrets, but only used the feature in rare cases, for instance when debugging issues using few test devices. Including thousands, or tens of thousands of device secrets in one API call was not feasible for business purposes.

Thanks to the close cooperation between and Pushwize, a custom development was done in the latter system as well, providing space for putting the new business cases into practice.


Pushwize developed a new API for with the purpose of creating device lists with unique identifiers. Before push notification sending, device secrets are uploaded to the lists to complete the targeted audience segmentation and overcome the problem of dealing with a mass amount of device secrets

As a result, when forwarding the notification requests towards Pushwize, refers only to the unique list identifiers, enabling the technological feasibility of both use cases. readers may now also receive notifications in topics to which they did not neccesarily subscribe, but based on their reader behavior may most likely to appreciate. For instance, a person who closely follows Max Verstappen news, is most likely open to follow generic Formula-1 articles as well. On the other hand, media editors can now decide whether regional news are relevant for people living in the area, working in the area or both. 

To ensure sustainable operation, Pushwize created another complementing API to help to validate whether the stored device secrets are still considered active in Pushwize.

Read more case studies

Push notification system development from scretch for entire media portfolio custom database integration case study

Szifon interactive iOS notifications in action


Contact our experts if you have questions here.