Personalizing news, a pain point or the wrong problem to solve?

There are ever-growing sources of news for people world over. Every day, hundreds of newspapers, news websites, apps, aggregators, social platforms and bots deliver news to people around the world. Either people self-select in choosing what they want to read or people are presented with news based on some sort of personalization options. I am interested in looking more closely at the range of personalization options in the digital medium currently in vogue. Listed below are some of the common methods used to deliver personalization of news for the user

  • Broad categorization is the most common personalization method used asking users to choose multiple categories that are of interest to them. This method is used as a starting point by many solutions. A few variations within broad categorization are: a) Choose a few interest-based categories out of the hundreds shown to you b) select a few influencers or people you like, who may curate or create news content for their followers or c) select brands or publishers you like to read
  • Social platforms personalize based on what friends/family like, read, view, tweet or share. Facebook was doing manual curation up until their recent announcement about algorithms personalizing content for the user (both approaches got bad rap for Facebook recently)
  • Social aggregation to bring out the most shared/liked in your respective social networks. Nuzzel app is an example of such a service
  • Location based news personalization focuses on what is happening around a person’s location and delivers content from local/hyperlocal sources
  • Newsletters that are curated with specific areas of interest are distributed through email lists

I am sure there are a few more options. Personally, I consume a combination of sources: apps, news sites, and newsletters. In the world where instant gratification with entertainment value is the norm, it is tough to engage audiences. The above personalization options are only scraping the surface in terms of creating a highly engaging experience. While many of these personalization options are interesting, personalizing news for the mass-market has various problems. Here are a few:

  • News cycle is too fast and many of the existing solutions use manual curation but this doesn’t quite scale when you reach a global audience
  • Learner algorithms, which can deliver personalized news articles are nascent. Solving the personalization problem using this method for a global audience will take time
  • The quality of the content from publishers is not consistent and the only way to deal with this is signing deals with every publisher out there. Again not easy to scale without a self-service model with all publishers sharing relevant metadata
  • Users can not quite communicate their preferences using categories as there are too many choices. Most of the time people may feel they may miss out on something if they don’t choose a lot of options
  • News consumption patterns differ as user intent changes during the course of a day/week and therefore personalizing news becomes much more difficult
  • Social signals don’t solve personalization either as the user may be interested in content that is dissimilar to their friends or influencers they follow
  • News was always more of a discovery process for people and not about getting personlized content delivered to them

Does this mean personalizing news is a hard problem to solve? Do users need better solutions? I think just focusing on offering customizations to the users will not cut it. While there are problems with personalization, companies can continue to innovate and deliver better experiences. News content is universally consumed across the world and appetite for news will continue to stay high. But users need more than personalization options, they need better news discovery and new distribution channels. Some companies are ideal for delivering personalization/discovery today and a few others are moving in the right direction. Companies that will succeed will have one or combination of the following elements: a) huge distribution b) scalable platform that relies on machine learning c) focus on short-form instantly gratifying moments d) specialize on verticals such as sports, finance, entertainment e) new methods of delivering news content such as automated newsletters f) offer alternate distribution channels. I will discuss these alternatives in more detail in my next post