Content Mining with Topic-Modeling
Overview of this campaign
When are readers on the page, what do they read? What are they willing to subscribe for? And what do they keep coming back for? We wanted to answer these questions to get to know our readers better, to find out what is important to them and what they need. And to find out what each of us can do to achieve greater reach, more subscriptions or to strengthen subscribers' loyalty to our articles and products.
We developed our own analysis to identify topics which work well or less well, whose output we should increase or review, in the areas of reach, paid content and retention through full-text topic modeling of our articles. The methodology is based on various text clustering techniques and finds finer topic groups than internal ressorts, sections or tags allow. We then evaluate the performance of these topic groups in the areas of reach, paid and retention in comparison to the costs, meaning the amount and lengths of published articles on the topic. In this way, we can identify topics for which we generate a lot of content but which are not well received by our readers, or topics for which we have little content but which are of great interest to our users.
The results were used to draw up concrete instructions for adapting the content portfolio to the objectives. We analyzed each of our local newsrooms individually, visualized the results and gave a presentation to the employees on site. We started a daily internal digital newsletter that summarizes which articles from the previous day worked particularly well in reach and in subscriptions to give employees a better idea of which topics worked well, and we have conducted one-day paid content training sessions across the Group.
Results for this campaign
The Funke Media Group consists of several regional newspaper titles distributed throughout Germany. Each regional brand has its own regional readership with its own interests. Topics that work well in one region don't necessarily work well in the other. Until now, knowing which issues work well in each region has been subjective and perceived truth. Good KPIs only stood out for outlier articles and were otherwise perceived in total as daily or weekly results. Our analysis has brought the interpretation of the KPIs to the level of topics and was thus able to discover topics that never reach the daily top ranking, but which bring constant traffic overall and may be the reason for readers to come back to our product independent of high-reach news situations.
Our analysis helped identifying topics that work well for each title individually. It also has been shown that there are only a few topics that work well in terms of reach as well as conversions and retention. More often, the optimal portfolio for these areas is different and varies from brand to brand. Thus, depending on which of these three areas is to be strengthened, the portfolio should look different.
Using a standardized analysis to individually analyze the digital content portfolio of each newspaper title made it possible to draw up concrete instructions and tips for adapting the portfolio personalized for each individual title as well as separated by target area (more reach, more subscriptions, better retention).
The visualizations and on site presentations helped communicating these instructions and tips to the newsroom and in combination with the daily newsletter and the paid content trainings, we believe it has had a big impact on how newsrooms now evaluate topics before deciding on how to cover it.