Gyan Gupta, CEO at DB Digital, shared with media executives at Big Data for Media Week the challenges facing DB Digital as a site without a paywall.
Despite not charging its readers, Dainik Bhaskar Group continues to prosper as India’s second-largest digital group, with a staggering 80 million unique visitors each day (2.6 billion monthly pageviews). It’s also crowned as the world’s fourth-largest newspaper.
Moving forward, Gupta explained that he continues to push DB Digital further with three main objectives:
- Increase user engagement.
- Help editors create more effective content.
- Show users relevant content.
These objectives led DB Digital to question how it would achieve each. In addition, three key questions face DB Digital and its development:
- How can the site tell the editors in real time how a story is performing?
- What are the best identifiers for learning the best performing internal and external distributing channels?
- How can DB Digital make an interface simple and easy to use?
Despite these clear objectives and questions, Gupta explained there were a multitude of challenges along the way:
- Weak meta-structure: Local editors were creating weak-meta descriptions. To change this, Gupta explains that more logical, independent meta description was needed.
- Vernacular language: The lack of available corpus for vernacular language such as Hindi meant engagement was hindered in some regions. A Hindi Corpus was introduced to boost engagement and retain readers and that was successful.
- Fallouts of being a consumer news site in India: Gupta noted, “In the absence of a pay model we were largely dealing with anonymous users and there is no data around that.” Consequently, the limited data points on users meant that data from elsewhere was needed, such as from click streams and reading more closely into reading patterns.
Taking into account these obstacles, Gupta then showcased DB Digital four main recommended logics for boosting engagement.
- Targeted key words as indicators of engagement and identifying higher relevancy keywords in their content.
- Use “Associate Mining,” which identifies stories that are doing well and appropriated score them on performance for future analysis.
- Focus on the user and their “Interest Stories.” Gupta said, “We take into account the reader user history, to identify their patters of engagement, what they have read before.”
- Similar to user interest stores, Gupta explains the “Story Sitch” logic that “puts people together who have shared similar patterns. This allows us to display them the most frequent stories within their pattern.”
Gupta finished his talk by focusing on DB Digital’s use of the innovative real time decision-making tool, Wisdom. It is an exciting tool that enables editors to focus on the top-20 performing stories. Focusing on these stories is key because “the quality of content here is where the brand reflects and weare taking a similar approach for video,” he says.
Showing a picture of LaLaLand with a twist, he explains, “We have completely crossed over into the Big Data Land.”