IPM: Serving accurate news using data scripts | Presentations and Cases





Journalism is about getting it first and getting it right. The news flashes are obsolete in no time, challenging marketers to continuously adapt their messaging at the speed of light. Information is part of the most volatile product that we have to master. 

Users spend 31 minutes each day consuming news digitally, and Google Search is their number one choice. Yet, it has been postulated that search engines have a negative impact on the diversity of news that people access. This is attributed to the algorithmic filtering used by tech giants. 

An in-depth analysis using Google Analytics and Google Ads performance data of the news websites, La Libre and La DH, owned by Groupe IPM, confirmed the assumption. Only 29% of the most recent articles were accessed in our advertising campaigns. This is in complete contradiction with the business objective of IPM. Their goal is to bring the newest information to their readers. Within the world of journalism qualitative traffic is not enough. Generating added value to our readers is about bringing the most recent news first. 

These insights made it clear that IPM was in need of an innovative solution that could bring the latest and most accurate news to all different readers. Semetis brought the solution of algorithmic manipulation using multi-source data points in real-time.




Step 1 - The set-up: From an advertising point of view, targeting is crucial. In the case of IPM, it is clear that the target groups are very dependent on what the news is about. Our personae are evolving constantly along with the information. The common denominator is to target these curious readers that actively search for the most recent information. 

A second complexity is that geolocalisation targeting is tricky and cannot be a criteria. At the first sight, it sounds logical that a region would be interested in local information. But in reality, they are also interested in national,  international or even local information from other regions. 

In a nutshell, our only targeting criteria from a business point of view is to target readers searching for the most recent news and absolutely avoid bringing them outdated information. 


Next to targeting, Messaging is decisive. Readers have high expectations and once they are hunting for precise information we cannot attract them with a generic message such as “ Discover the news on La Libre”. You need to get the message in order to entice the reader to read your article. Competition in the search engine results list is fierce. IPM needs to stand out by delivering the adequate message to the right audience. To overcome this hurdle, we investigated several options available on the market, among which Google News. A product developed to bring the newest information high-up in the search results ranking. Unfortunately, the algorithm behind this product is not available to advertisers. Given the multitude of dimensions to be taken into account to serve the proper message to the different audiences, automation was our way of moving forward. Automation seems extremely appealing but it needs to make sense from a business point of view.


To leverage the best of both worlds into our data mechanism approach, our strategy was based on 2 axes - automation and control:  

  1. Automation to respond to a massive volume of searches that differ every day. We decided to leverage the Dynamic Search Ads format from Google. This solution enables you to target searched keywords on Google based on the content available on your website. Data is required to automatically adapt the advertising message to the keywords being bought. Sadly there is a “but”.This solution cannot satisfy our expectations, as it takes the full content of your website into account, this includes articles from 5 years ago. This is where our script enters the game. 
  2. Control a very accurate response. Data is required to automatically serve the most recent news to our readers. This control is required to guarantee to stay in line with IPM objectives. 


Semetis wrote a customized script to implement this dual strategy using multi-source data points

  1. The news article's publication date guarantees the recency of the information. The data was pulled from the dataLayer. The rule was set to a maximum of 48 hours old information. 
  2. The website sitemap that is summarizing the website structure from an SEO point of view. This data point is necessary to target all available articles within the IPM website ecosystem
  3. The popularity of the news topic to take into account the interest of readers. The data is pulled from Google Trends to measure the hotness of a topic and the relative volume. 

Next,  automated ad texts were created using the content of the recent website articles and customized for every search. This automated and innovative script was our path to success. 


Step 2 - The optimization: Our usage of data is not stopping here. A data infrastructure was implemented and linked to our advertising and analytics platforms to measure our results. A new data point was created being “publish date”.  This enabled the media buying teams to evaluate the recentness of the content consumed through our advertising and optimize the campaigns directly. 


Secondly, we used data to evaluate and improve our script. Daily we were monitoring the landing pages of our advertising campaign to validate the relevance of automation. Initially, the script was only taking into account the articles published within the last 48 hours. After 2 weeks we realized that the volume of requests covered was too low. We decided to enhance our definition of recent articles and took a slightly longer time lapse 




The results were incredible. We succeeded in bringing the proportion of recent article reads from 29% to 81%. More concretely this means that 81% of our automated ads were directed to articles that have been published within the last 2 months. 


Extra bonus:

  • The cost per click of our campaigns decreased by 38% allowing us to target more qualitative search requests with the same budgets
  • The consideration of newspaper subscriptions increased by 622%. This could be measured through button clicks from the article to the newspaper subscription e-commerce. This is an ultimate proof that bringing qualitative content to the fingertips of readers, convinces them to become a customer. Together with IPM, we succeeded in turning the audience of readers into action-able media audiences that are interested in becoming subscribers. 


IPM Group's usage of data allowed them to be first and right, something that readers highly appreciate.

publication author margaux marien
Margaux Marien

| LinkedinThis email address is being protected from spambots. You need JavaScript enabled to view it.


Get in touch

Semetis | Boulevard Saint-Lazare 4, 1210 Brussels - Belgium


Connect with us

Cookie Policy

This website uses cookies that are necessary to its functioning and required to achieve the purposes illustrated in the privacy policy. By accepting this OR scrolling this page OR continuing to browse, you agree to our privacy policy.