Predictive Content Recommendation

Business Case:

KPI to be measured

Bookmark this resource Follow

Ask a question

Was this article helpful? 0 out of 0 found this helpful



Predictive Content Recommendation is a THRON feature, included in the Intelligence module, that allows you to:


  • Accelerate conversions with Dynamic Content. Effectively offer specific call-to-action based on the level of interest of the user on specific topics.
  • Personalize the Omnichannel Customer Experience based on content topics viewed. Regardless of the device used, THRON’s Predictive Content Recommendation suggests the most relevant content based on user interests derived from content topics for each individual user (not from agnostic 3rd party information).
  • Set the content prediction engine with your editorial strategy. THRON's PCR provides a powerful console where you can easily set logic and priorities for the Recommendation engine. Leverage the application's machine learning algorithm but always still comply with company strategy and business needs.
  • Jump from Segmentation to Individual Personalization. THRON avoids the need to create Customer Segmentation to offer users a personalized customer experience and consistent call-to-action suggestion. THRON analyses User Behavior based on relevant topics & content usage to predict the best smart content to recommend for every single user.
  • Customize layout in all your touchpoints. Ensure your websites are completely personalized and ‘made to measure’ by integrating a simple HTML5 Widget.


THRON's Predictive Content Recommendation is the gateway that drives customers towards the brand channels that contain content of interest, building a true personalized experience for every single user.

What is revolutionary about our Predictive Content Recommendation is that apart from its ability to aggregate content from any channel, it allows you to personalize communication for any touchpoint. From a website or mobile app to e-commerce, our Predictive Content Recommendation enables you to exponentially raise user engagement.


What’s in it


During the creation of a new Recommendation you will have to specify:


  • The name of the Recommendation
  • The classification to be used to recommend content based on tags
  • The folder from which content to be suggested will be withdrawn


The algorithm used by each Recommendation to suggest content can be enhanced thanks to specific rules which can be entirely configured by the user. Available rules are:


  • Tag boost: you can define a set of tags in the presence of which the content will be suggested in the top positions.
  • Domains: you can limit the view count to specific domains.
  • Recently viewed: you can include among recommended content also that recently visited by the specific contact
  • Content without visits: you can include among recommended content also that which does not have visits. In this case you must specify the URL to which users will be redirected after one of these contents is clicked.
  • Decaying rate: you can configure, for each recommendation, whether the most recent content has to be proposed to users more compared to old one. This allows you to leverage recommendation for content of different nature such as editorial recommendations (which importance fades over time)  and product content recommendations (which never become obsolete). More specifically, you can set:
    • How long does it take before content is considered obsolete
    • How long does it take from the moment when it is considered obsolete to the moment when it has reached its maximum obsolescence
    • Whether or not to exclude from the recommendation contents that have reached their maximum obsolescence


Once the configuration of the single Recommendation is complete, you will be able to include the widgets within your channels by simply copying the embed code provided by the application.


How to include it

Once activated, Predictive Content Recommendation can be included in web pages with any of the following methods.


Via embeddable widget


Available widgets are:

  • Adaptive Content Wall: a wall of content that dynamically adapts to the size of the page .
  • Horizontal Content List: a horizontal list of contents.


For experienced users, more information on how to customize widgets is available here.



If content recommendations have been created, you can use THRON API to take advantage of the predictive algorithm and get the best content to propose to the specific user who is viewing your page in real time.

All information about the APIs to be used can be found here.

Was this article helpful?
0 out of 0 found this helpful

Have any question?

Open a ticket