Why personalization will be the next revolution in the news industry
The shift in news from a one-size-fits-all model to a personalized model will happen fast and will be brutal.
The complete shift to digital has taken about 20 years. Finally, large news organizations have reached the point where their main distribution channel is the smartphone. Even legacy media like The New York Times think in that way. Their CEO, Mark Thompson, said so recently, during a visit to Stanford University. “Most of our readers have never seen a New York Times newspaper,” he said. “The print version is now a sub product assembled by a small group of paper design specialists.”
The next shift won’t take 20 years. It will happen incredibly fast, and it will be brutal. The news platforms that won’t be able to transition quickly from their broadcasting model (one-size-fits-all) to a personalized model (each user gets tailored content), will struggle to keep a viable audience.
This may sound peremptory. But the changes that deep personalization of news platforms will bring are far-reaching. The transition has just begun. After having explored the topic during my first three months as a John S. Knight Journalism Fellow at Stanford, this intuition has become a powerful conviction. Here are some reasons why.
Larger news organizations like The Washington Post publish daily about 1,000 news stories (from staff and wire services). But different organizations I have interviewed admit that even subscribers click on average on fewer than five articles per day. This can be confirmed by the extrapolation of media metrix data.
Unsurprisingly, all prefer keeping this exact number secret. But why is it so low? First, the attention span has completely changed: Users are overloaded with information, social network posts, notifications, emails, etc. Second, the size of mobile screens (70–80 percent of the consumption now) allows space to tease only four or five stories on the landing page. For the readers who open their news app, scrolling down is often too much effort — journalists may accept it or not.
And here it becomes even worse for news organizations: On the handful of stories that are presented on the mobile landing page, only one, maybe two, really interest the “average reader.” Facebook, who competes with news organizations, does much better and takes about 20 percent of the internet time Americans spend on mobile devices. Why? Facebook delivers a highly personalized experience ‒ news apps don’t. News organizations still address their products to an “average” that does not exist. Each reader has different needs.
Let’s wrap up to face the cruel reality: On 1,000 stories produced daily, only five get a real opportunity to reach a broad audience. What about the 995 other stories? Are they all crap? Or do they deserve a better chance? As an investigative journalist who has spent about 20 years trying to write outstanding public-interest stories, I profoundly believe that the best journalism is not good enough. We also need the best technology to distribute it.
How to fix it?
Imagine now a general news platform that can distribute the content that matters to each different reader, at the right moment, in the right form. The smartest news organizations will be able to personalize their news feeds in a such a subtle manner for every single reader, that it can become addictive, too.
News platforms could personalize the importance of each section for a particular user, the delivery time, the headline, the recommendations, etc.
Even the content of the stories themselves could be tailored for each reader. In just 10 weeks, a team I have been working on for a journalism class have developed a prototype of a writing tool that makes it possible for nearly every article. With this new way of crafting stories, journalists could, for example, modify the tone of their article significantly by varying the title and a few sentences. This would allow new strategies, to better engage younger readers without deceiving the base of older readers.
When the users check their phone 80 times per day, the incentives to open the news app would be much bigger. And at the end, it would eventually improve the subscription rate.
The tailored delivery of stories would also allow republishing content on a much broader scale. Today, in an effort to value their content and lower the costs, some news organizations manually publish “evergreen” stories again and again on their homepage. For the readers who already have read this particular piece, this may become very annoying. If the title or the picture has been changed in the meantime, they may feel tricked.
There’s no such risk on a personalized news platform. News organizations would know who has read which story and when. If a reader missed a story two days ago (a story that similar users have read and rated well), that reader would likely be happy to have it top of his list.
Instead of wasting their resources by letting the premium, well-researched stories or crafted opinions be washed out by the next thousand stories, those pieces would get a fair chance to be read.
And if the news organization is able to know where each reader lives and commutes, it could be valuable to deliver local or even hyperlocal content to them. The same goes for special interests.
The fact, for example, that a soccer field in my small village will soon be covered with artificial turf would never have found a space in my local newspaper. It would have been irrelevant to the vast majority of other readers. But as an inhabitant of this village, this really matters to me. And I want to have it on the top of my feed, or even get a personalized push notification. The soccer fan of the neighboring village might be interested, too. Following this logic, news organizations could build long-tail audiences thanks to personalization, distributing a lot of special interest stories with a small audience.
The business side of news organizations would also benefit largely from personalization. It would allow improving the marketing to gain new subscribers or contributors. On the other hand, it would open up new possibilities for advertising.
The technical challenge is high
Such a significant shift will take time and will need serious resources. Only the bigger players will be able to make the journey towards personalization on their own. Smaller newspapers or news websites must partner with each other. Their IT systems must be radically improved, and this will be very costly. They must be able to capture and use at its best the signals of each user: their demographics, their location, their reading history — not only what they clicked on, but what they effectively read and how they rated or shared a particular story. To be effective, this data gathering has to be done across devices and platforms, no matter if mobile, web or tablet.
To accept this, readers must believe this offers them a significant advantage. It’s the same way we accept giving our location to Yelp! or Google Maps, or how we share personal data with Facebook or Instagram, or how we accept that Google search results aren’t the same as for our neighbor. Because we see the immediate benefit.
News organizations, of course, must uphold the highest standards. They must be committed to maintaining their journalistic integrity and to operating transparently. They also need to find the balance between the reader’s personal interests and relevance from a general news perspective.
But who better than trained journalists, and legacy newspapers as institutions, with a strong culture scrutiny but also self-criticism, is in position to do that right?
If they don’t do it, some Silicon Valley kids will do it for them. ◼
Note:
In a future article, I will write about the different fields where news organizations can act to personalize the reader’s experience. The hard and the easy ones.
I will also write about the signals that can be captured to improve the user experience on news platforms. I am especially interested in the impact of the next generation of sensors on devices, which will allow emotion recognition or eye tracking.
I am also interested in the privacy question and the ethical questions around a profound personalization.
Here is my list of signals: reading history, reading deepness (scroll, time spent), ratings, shares, network (friends, family), device ID, network info, readers demographics, surveys, interests on other websites, search queries, GPS, microphone (capturing the presence of noise or not), face recognition, emotion capture, eye tracking.
I am happy to have your point of view and start a conversation.
Please write me at: titus.plattner@stanford.edu
