For Publishers: Achieve goals with this Recommendation Engine

📢 Tatvic is now a global partner of Usercentrics, world’s leading consent management platform

For Publishers: Achieve goals of increasing your revenue and reader engagement with this Recommendation Engine

Last Updated:


[vc_row][vc_column][vc_column_text]According to a study by Deloitte, every publisher organization that is thriving to grow in the digital space aims at three major goals:

  • Improve Reader Engagement
  • Increase in Direct paying Relationships with Readers
  • Drive advertiser revenue

Irrespective of being different, these are not independent of each other. The latter two highly depend on how much effort the organization invests in improving Reader Engagement. Reader Engagement is usually measured by the time spent by the user on the website, the number of interactions that the user has on the website and lastly, the number of consecutive articles read by the user.

Recommendation Engine for Publishers by Tatvic’ Cloud for Marketing team, is a tool that enables the publishers to serve real-time personalized content recommendations to the readers. These personalized recommendations improve the metrics that are of key importance to the publishers i.e. Click Through Rates of recommended articles, increased number of page visits per session, increased average session duration etc.

Recommendation Engine can be developed by using multiple techniques as listed below:

Content-Based Filtering Technique

Content-Based Technique majorly depends on the content attributed in order to generate the recommendations. This is the most widely used technique when publications and news are to be recommended to the users. On the basis of attributes of content that the user was previously engaged with, recommendations are provided to that user. Certain publishers capture user ratings for different pieces of content. This can act as a mechanism to decide on what type of content is preferred by the user.

Collaborative Filtering Technique

Collaborative filtering does not depend on the content attributes to generate recommendations. It uses the database of user preferences for different items. Collaborative Filtering uses user profiles to recommend the content pieces to the users whose interests and preferences match with a particular profile.

Despite never having accessed certain content, a user gets recommendations for that since those content pieces were already positively rated by other users who constitute a similar profile. Recommendations that are generated using this technique can either be prediction or recommendation.

Hybrid Filtering Technique

Hybrid filtering technique is a combination of Content based and Collaborative filtering techniques that can optimize the benefits and avoid their limitations. Tatvic’ Recommendation engine is built on this technique. 

Hybrid technique is based on the idea that when two different techniques are used together, it will enhance the accuracy and effectiveness of the recommendations as compared to the single technique. Using multiple filtering techniques can suppress the weaknesses of an individual technique in a combined model. Hybrid filtering technique uses various methods to increase the relevance of the recommendations.

It can either use Content Based and Collaborative filtering sequentially by feeding the result of one technique into another which will process that information to generate recommendations, or these techniques can be used simultaneously and combined result can be used to generate recommendations and lastly a system that unifies both the techniques can be created.

How does Content Recommendation Engine benefit the Publishers:


It has been observed during various studies that the articles that feature as recommended, tend to have more CTR as compared to the other articles. Publishers can get maximum revenue for the advertisement inventories that are placed in those articles.

Reader Loyalty

Content Recommendation Engine displays content options on the basis of historical interactions that the reader had with the publisher website. Hence when he/she revisits the website to discover the preferred content shown to them, it creates a sense of customer loyalty.


When the publishers already have user data from the previous visits, it facilitates them to modify the UI/UX of the landing page for the customer and personalize it with the content on the lines of what he/she was engaged with during the previous visit. This makes the navigation experience pleasant for the user and avoids unnecessary clutter.

Content Discovery

In an industry like publishing, where numerous content is generated on a daily basis, it is highly probable that a reader might miss out on something that he/she would like to read. With our Content Recommendation engine, such type of content is pushed to the front when the reader is on the publisher website and this adds to the customer experience.

Concluding Thoughts

You can increase readership and deepen engagement with data-driven content recommendations that evolve with your reader’s behaviour. With Cloud for Marketing solutions like these, digital publishers can now dynamically adapt and personalise their recommendations to each user’s preferences, content affinity and real-time intent. Learn how Jagran Increased User Engagement in m-Web by 92% using Tatvic’ Content Recommendation Engine

Is this something that interests you? Contact us today![/vc_column_text][/vc_column][/vc_row]

Durga Chouhan

Durga Chouhan

Durga is a GAIQ certified Data Analytics consultant and works as a Customer Success Manager at Tatvic Analytics. She has trained clients on Google Analytics and is passionate about using data interpretation and analysis to create business solutions.

Sign up for
our monthly newsletter

This field is for validation purposes and should be left unchanged.

Other Blog

Scroll to Top

Leverage Tatvic's comprehensive approach to 

Enquire Now

This field is for validation purposes and should be left unchanged.

Leverage Tatvic's comprehensive approach to 

Contact Us

This field is for validation purposes and should be left unchanged.