How publishers are making news more personal

With the help of new tools, publishers are finding new and better ways to personalize content for audiences that increasingly are looking for information that's most relevant to them.

New technology such as the idio platform can help publishers learn about their audience in order to automatically recommend better news for them. And personalized aggregation tools, which have existed on the Web for years, are now becoming more prominent on new digital platforms with applications such as Flipboard.

Matching news to personality

News organizations such as The Los Angeles Times continue to experiment with personalization, using customization as both a content and advertising initiative. The Times worked with VisualDNA to create a virtual quiz called Newsmatch, which helps readers create a personalized homepage based on their interests and preferences. Newsmatch guides readers through a series of questions around favorite things, most important issues, and life goals, as well as demographic information. The technology generates a personality profile along with articles and topics that are deemed to fit the profile.

The VisualDNA service profiles users into 13 primary "life categories," which cover more than 120 audience groups representing different interests, tastes, needs and intent. Those groups can be used in any combination in conjunction with other demographic filters to target users. 

In addition to helping readers find new content, the quiz allows The LA Times to learn about readers in ways that can inform both editorial and marketing, said Sean Gallagher, the Times' managing editor of online.

Gallagher said he could look at analytics in Omniture to see aggregate data, but Newsmatch will provide better profiles of the regular users, or "Times junkies" (he estimates about 5% of audience uses it at any given time). "It helps me develop a more well-rounded pictures of the typical user," he said.

For instance, he can spot spillovers of interest that he wouldn't necessarily correlate, such as people who like news about both food and sports.

The personalization technology can also inform targeted advertising, as one of the questions asks the user to rate brand preferences.

"Having Newsmatch is just another tool in the arsenal," said Gallagher. "It's one that gives us qualitative instead of blunt-edge quantitative."

The LA Times plans to roll out a more entertainment-focused version of the service soon, he said.

Avoiding too much personalization

One of the challenges of personalization is finding the right balance between giving users what they want and letting them stumble upon news the old-fashioned way. In some ways, online news and over-customization has reduced the serendipity readers experience when flipping through a newspaper. On the other hand, personalization can foster serendipity by suggesting new topics or articles that might fit a user's profile (much like the music service Pandora can introduce you to new music).

Loosely customizing news for users doesn't necessarily take away the option to find it; it simply makes the process less overwhelming. Gallagher emphasized that customization shouldn't put too much work in the users' hands. People will only personalize to a certain extent, Gallagher said.

This is why initiatives such as the New York Times' My Times and a similar project from the LA Times did not succeed. "The 'My LA Times' stuff asks readers to do much," said Gallagher. He believes personalization will evolve to be more of an "Amazon model" of recommendation, based on data. "You can't ask users to sit and fill out a questionnaire," he said. "It's going to be an ongoing relationship where the site interacts with the user." 

Data gets personal

Such data-driven personalization is now possible using semantic technology, which can build recommendations for future stories based on the context of an article. For example, the U.K.-based idio built a "personalized publishing platform," a semantic solution that includes a personalization element. (Other semantic tools, such as the open-source OpenCalais, could also offer these capabilities; idio has just already integrated the function into its platform.)

Like other semantic platforms, idio can extract information from the content, but it also brings in information about the user from social media or other sources to personalize the content. For instance, the platform can leverage a user's Facebook login to customize recommendations. As an example, a music magazine they created called idiomag utilizes preferences on Pandora to recommend content.

In addition to using social media, publishers could also use a user's profile on their site to build up data about the user and have a matching algorithm to choose the best content, explained Andrew Davies, commercial director at idio. Publishers in different verticals can use different methods: a B2B publication could use a LinkedIn login, or a sports publication could ask for a team name.

Davies said they saw an opportunity to build a technology product around personalization. "Because of sites like Pandora, people expect personalization to work," he said.

idio also works with publishers for other personalization initiatives, such as e-mail alerts and browser extensions. Right now the company is working with several big media companies (including The Independent, The Guardian and NBC) to launch a browser extension that delivers stories based on a user's interest.

Davies sees personalization playing a role more through being a "guided journey" on a site rather than forcing an exact number of articles on someone. Only providing users a small selection without giving them more choice "actually might reduce people's browsing time rather than increase it," he said. 

The idio approach is to follow users through a site and suggest a range of further options where they could go forward.

Personalization, he said, will be even more important going forward in the mobile channels, where users "have a lot less space."

Personalized aggregation platforms

Outside of the publisher's own website, personalized news aggregation/curation has been prevalent for a while, via sites like Google News (which has made a big push toward personalization lately), browser extensions like Readness, social media like Twitter, and now on iPad apps such as Pulse and Flipboard.

Aggregation of news on other platforms doesn't necessarily counteract the internal personalization efforts publishers are making.

In fact, there could be more integration of personalized aggregators and news sites. For instance, last month The Washington Post acquired the personalized news aggregator iCurrent, a service that aggregates news based on a user's interest. (A spokesperson declined to comment at this time about their plans for the aggregator.)

The Washington Post move represents a trend of media companies investing in and deploying aggregation and curation services. The New York Times has invested in Daylife and the New York Daily News has invested in LOUD3R. 

The next step for collaboration between publishers and aggregators could be in the mobile space. The popular new iPad app Flipboard aims to better present the content that people are sharing on social networks. In a July interview, Flipboard CEO Mike McCue said that the company plans to work with publishers to help them better optimize content for Flipboard and the iPad.

Like it or not, when media companies publish content, it goes into the personalized stream.

“That’s a very disruptive force for publishers," said McCue. "But if they can figure out a way to leverage that power and figure out a way to make their content get into those social streams, they can open up new audiences and could be able to have their content be more relevant for readers.”