We recently met Gyan Gupta, the former chief executive officer of DB digital, who in the past year has gone on to build his own start-up called Chamko. Gupta was associated with Dainik Bhaskar’s DB Digital for more than 8 years when we met him last year during the INMA study tour. At the time, he shared insights to the software development the company was doing using artificial intelligence (AI) for its news websites. Although his current project is not about news media, our idea was to query him about the newspaper industry’s digital transition from a data scientist’s point of view.
Asked about the future of Indian newspapers, Gupta says, “Newspapers per se will not go. A few companies will survive but I cannot say which these will be. I cannot also say in which form they will survive. It could be in digital form or in both print and digital. It is a bit difficult to say. However, I don’t think that they can keep giving digital news free of cost to the reader.
“The publishers need to think about how to make users pay for news on digital platforms. They need to find out – what is the kind of content that a user would consume? And what kind of content would the user pay for? Then what percentage of the users will actually pay? Finally, what kind of value would advertisers get from digital readers or users? It’s almost obvious that the content has to be compelling and habit-forming, almost addictive – but not all content qualifies. For instance, readers are willing to take digital subscriptions for the New York Times but not for USA Today.”
At the same time, Gupta says that publishers need to understand that digital advertisers can get a lot of value from the demographics and user data that they can slice, dice and deliver. The problem here is for the news media to be able to first use these demographics and data to give users what they want to consume. Next, they need to provide the relevant analytics for advertisers to realize value from the various aspects of the demographics.
Looking at the larger picture
Gupta believes that the Indian news media needs to look beyond their own day to day publication issues or simply local competition. Publishers should set aside their individual differences and get together. The Indian Digital News Publishers Association (DNPA), which was started last year by about 10 founding news media members, was an attempt in that direction. “You have to look at the larger picture, down the years into the future. There are so many issues such as the quality of journalism or fake news. It is a bit difficult because earlier the competition was with local players and now the competition is with global players. Technology is a different theater and it is a huge differentiator. Just look at the cost of optimizing one video player. How do you optimize it and spend time and money on that. The association should think in that direction. On the other hand, lots of things are possible in Europe because the publishers are able to come together and discuss these things.”
AI engine captures reader interests
Gupta says it is not impossible to build technology that will work for Indian news media, even in the face of large global players. “At DB Digital, we were able to build an engine that was able to point users to what they wanted. It was an intuitive relevance engine depending on what you were reading and your interests were. To that degree it was able to capture your interests. In a way it was able to pinpoint to the reader.
“However, too much personalization becomes a bias. One of the challenges with AI is that it has an inherent bias. The whole advantage of the printed newspaper was discovery, and readers love surprises as they turn from page to page. I think that is a factor that has to be kept in mind – a balance is needed. Otherwise personalization could be dangerously inward and self-limiting.
“Moreover, to take these developments further, you need to keep building on the network. AI is something that can be used to build tools to deliver stories that will hook the reader. And they will only pay you if they are addicted! The idea is to first get the reader addicted to the content and then bring in the personalization or customization aspects.”
According to Gupta, technology development is not inexpensive and it requires serious long-term investment. “The cost is high. First the technical personnel and the data scientists are relatively expensive. Then the hardware, software, network and server costs are high because a lot of computing is needed even to find one article. Data costs are themselves fairly high.”
Control over content
We discussed the increasing number of technology players who are combining media and entertainment content and also providing the distribution channels. Increasingly content is also owned by the network service providers. Gupta says that the issue here is not simply content but that the global technology giants control the distribution. “When distribution becomes more important than the content, it starts dictating the content and when this is controlled by the technology players, it becomes a dangerous game. Leave aside the business model; in the past the publisher has always had control and with the new technology this has now slipped away to the technology providers and aggregators. Unfortunately, the network has become more important. The Indian publishers should get together to more forcefully present their interests as far as the terms of engagement.”