Logo De Tijd

Aiming for higher conversion through a personalized online newspaper.


In 2010 Mediafin, publisher of De Tijd, installed a paywall on tijd.be. This paywall invites reader to register after reading the first -free- article. After registration (name and email address) four more free articles can be read. If the user want to read even more content, a payed subscription is necessary.

The paywall uses cookies (a small text file saved in the browser) to determine the readers’ current status (not registered – registered – payed subscription) and how many articles are still free to read. This means that every user can read five free articles in a 30 day window, after which period the counter is automatically reset.

Starting point:

In September 2015 Mediafin, Adhese and iMinds, with the financial support of the innovation subsidy programme of Flanders Innovation & Entrepreneurship (VLAIO) and Innoviris, started the Adaptive Paywall System project. The aim of this project was to try to achieve more conversions both from non-registered to registered and from registered to paying users on the online channels of De Tijd, by providing personalised content matched to the interests and the profile of every single user. The hypothesis was that this  approach would also result in more impressions, more pageviews and a lower bounce rate.


A close cooperation between all partners involved in the project was crucial from the start. A field study and interviews, both technical and qualitative, formed the base for further workshops and testing.


The technical part of the project posed three challenges:

  1. The exchange, processing and analysation of the historical data Mediafin has amassed since 2010. This data contains information about the users and usage of all online channels of De Tijd. After analysing the data additional data measuring points were defined and tracked by Adhese.
  2. Real time data exchange between Mediafin and Adhese.
  3. Setting up the Mediafin and Adhese platforms up to allow for easy interaction, in both directions:
    1. Adhese can make changes to both layout and content on the Mediafin platform
    2. Mediafin can create campaigns, adjust and optimise targeting, and view reports on the Adhese platform


Target groups and goals

The qualitative part of the project consisted of researching and defining several audience target groups. These groups would then be served with tailored content, thus increasing the number of page views and conversions.The research started with:

  • Several interviews with stakeholders within Mediafin (editors, online editors, marketing department, advertising department)
  • Workshops and brainstorms with those stakeholders

Starting from these interviews, workshops and brainstorms a number of target groups were defined. Those groups were then compared to the data cluster research Adhese had performed on the historical and real time data.

Several adjustments, based on the data, were made to the target groups proposed by Mediafin. This clearly underlines the importance of both data collection and analysis when taking strategic business decisions.

Following the same methodology several strategic targets (for example ‘convince readers who traditionally only consume content during the weekend to also read during the week’ or ‘try to make users who only visit the site on desktop during business hours to also visit on mobile outside business hours’) were defined.

Prioritising and defining scenarios

After defining the target groups and goals, a prioritised list was made. Based on this list brainstorms were organised resulting in several possible scenarios.

Proof of concept

A first proof of concept took place in parallel: Adhese took control of several elements contained in the Mediafin online channels.

This proof of concept turned out successful and was followed by a series of A/B-tests. Those tests tried to measure the impact of layout and style of the website on online conversion.

Both the proof of concept and the following A/B-tests confirmed that Adhese is perfectly capable of controlling and changing elements in the online infrastructure of Mediafin in a real time and live environment.

First scenario

Adhese took control of the newsteaser on Tijd.be.
Adhese took control of the newsteaser on Tijd.be.

In the first real test the ‘newsteaser’ (a navigational object located at the top of article pages) was directly controlled by Adhese.

Visitors were randomly assigned to one of four target groups. Each group saw a different selection of articles in the newsteaser:

  • Group 1: the ‘normal’ newsteaser, as defined by Mediafin (a combination of recent articles and manually picked articles by editors)
  • Group 2: the most read articles from the same category as the main article from the past 12 hours
  • Group 3: the most read articles from a different category as the main article from the past 12 hours
  • Group 4: the most read articles from all categories from the past 12 hours

Articles which had already been read by the visitor were omitted for all four groups.

First results

The results of the first test immediately show the importance of tailoring content to the right audience. Over a three week period the second group (who saw articles from the same category as the main article) scored significantly higher in terms of click throughs (an increase in click-through rate from 5,3% to 12,9% or 144% more clicks) when compared to group 1 (the original Mediafin setup).

Tailored content in the newsteaser provided 144% more clicks.
Tailored content in the newsteaser provided 144% more clicks.

Despite the significance of this first result, this wasn’t necessary the primal aim of the project. More important for both Mediafin and Adhese is that we’ve shown to be able to intervene in online channels in a real time and live environment. This implies that more and more complicated test can easily be conducted on very short notice and without major implementations or technical interventions.

A personalised paywall

This brought us to the actual aim of the project: adjusting the paywall for every indivual visitor of Tijd.be across all online channels. By doing this we hope to trigger every visitor at just the right point in his or her customer journey, thus increasing overall conversion from anonymous to registered and from registered to paying customers.

Several tests are currently being set up: in these tests Adhese controls when and how the paywall is shown to every individual visitor. By offering more (or less) free-to-read content Mediafin and Adhese hope to convince even more non-registered, non-paying customers of the added value of registering and in the end taking a payed subscription to De Tijd.

Follow up:

The technological innovations achieved in this project did not go unnoticed. We are proud to announce that Google has decided to provide us with substantial funding to take this project to the next level. Mediafin and Adhese will be joined by NGDATA in this follow up of the project.



Research and development: a dynamic paywall for Belgian newspaper De Tijd.
Tagged on: