Why News Publishers Are Still Getting Personalization Wrong

Unprecedented access to data has introduced a whole new set of problems (and opportunities) for news publishers competing for digital attention

News publishing is experiencing a fundamental shift on all fronts – every aspect of its creation, distribution, and consumption is being redefined for a digital existence. While retail was among the first industries to respond to radical digital disruption, traditional news media has taken many more years to evolve. The shift to digital means a painful, inevitable changing of the guard few news organizations are willing to share publicly. It also means taking a closer look at declining year-over-year advertising revenue, accounting for nearly 70% of a news organization’s financial lifeblood. To effectively monetize, news publishers are making personalization and optimization a core component of news publishing.

Digital First Heralds Big Data (and Bigger Business)

Selling papers is all but dead. In a digital marketplace, news media is in the business of selling page views. This means every publisher must embrace technological, data-driven methodologies once reserved for technology companies. If the balance between quality editorial and commercial interests used to present a challenge, unprecedented access to data has introduced a whole new set of problems (and opportunities) for news publishers competing for digital attention.

Consider the meteoric rise of the Huffington Post, acquired by AOL for $315 million in 2011. That publication’s continued success is a direct result of the digital marketplace it has embraced since launch. Its continued success is attributed directly to best practices and methodologies known to technologists, but unfamiliar to many publishers at best (or scorned at worst).

Thanks to integrating a data-driven approach (e.g. allowing article headline testing to influence the editorial process and craft the most attractive content), the Huffington Post was ranked by Newswhip as the top shared Facebook publisher this past September. Of the top 10 Facebook publishers, three are primarily algorithm-driven (Buzzfeed, Upworthy, and Playbuzz). These figures are echoed in The New York Times’ own illustration of how the publisher is losing digital readership.


New York Times Leaked Innovation Report

It’s the Algorithm, Dummy

To stay relevant, news publishers are approaching personalization and optimization as a strategic process:

  • The New York Times’ leaked Innovation Report quickly went viral on March 24th. The discussion of personalization (p 37) is especially interesting. Boris Chen, the Times’ data scientist, is quoted saying “It’s possible we’re using the entirely wrong algorithm,” followed by an appeal to include editors in the process.
  • The following Monday, Cory Haik, senior editor for digital news at The Washington Post, announced Social Flow, the Post’s real-time social context layer. This is part of a series of initiatives to digitally upgrade the publisher’s business model. Jeff Bezos, who bought the Post for $250 million last year, is one of the pioneers of digital personalization as Amazon CEO. Last week, the publication announced the launch of its Kindle app. Mr. Bezos is credited as the app’s “most frequent beta tester”.
  • This past July, Fox News reported a 20% increase in mobile traffic as a direct result of optimization efforts. As Chief Digital Officer Jeff Misenti explained, “We want to make sure that FoxNews.com performs well for our users regardless of how or where they might be accessing the site.”

The above examples illustrate the imperative to bring quality journalism into the 21st century. That transformation is just beginning to take shape. While some publishers are overhauling the entire business, others are free to start on a smaller scale. Many are testing front page elements recommending personalized content, identifying which articles or native advertising elements perform more effectively. Testing different headlines or images for the same article offers an increased level of granularity. By analyzing how different audiences consume content, publishers can make editorial decisions that are supported by the data. As new stories break, the right blend of automation and real-time personalization can actually empower editors to increase readership without compromising journalistic excellence.

As data analysis becomes critical to the success of news publishers, editors will find new, creative ways to integrate the wisdom of digital readers into the editorial process. While machine algorithms can never replace an experienced human editor, they can (and do) make an editor’s job easier. Successful publishers will learn to strike the right balance between editorial excellence and automation. This means integrating the right technologies that enable publishers to listen to readers – offering them precisely what they want, at the right time, and on their favorite device.