Why news personalization has flopped with users, media

Photo by samplediz

Theories on the Internet’s effect on news consumption generally follow one of two themes. The first is that it frees users to find the news of interest and consequence for them — personalized news — without the constrictions of time, editor-gatekeepers and journalistic agendas. The second is that the Internet is a mirror reflecting itself, that active (aggressive) users flock to news and views that please them, actually gaining less perspective, while the majority of users continue, as they always did, to be uninterested in most news.

Both presume, with varying degrees of credulity, to be the answer to the pesky revenue drain that legacy media have been unable to address.

The second theory is really the dark side of the first theory, though both are based on a liberated mass of news consumers, finally free of print and the constraints of legacy media. In the cheerier first scenario, these users have either chafed to determine and channel their interests or are newly empowered and inspired to do so — and the new solution both empowers humans intellectually while largely removing them from the gathering process.

But, to be cynical, are we not overestimating most people’s ambition, particularly as being self-informed often amounts to a non-paying civic duty? Are we not also overestimating people’s available time, as technology after technology saves users time from the previous way of doing things, only to have them redirect that time into the new thing?

Over at the Guardian, there is some reflection on a new study that looks at news personalization, the “My Page” efforts and the perhaps surprising conclusion that “that readers are reluctant to take on the role of editorial selection, and still enjoy serendipitous discovery.” I’ve found this study fascinating, as much of its areas of exploration and terminology were on my mind as I started this post (solely off the Guardian writeup).
Why don’t users want to define their interests? Why do they like surprises? The answer to the former is a matter of technology limitations and the growing pains of gathering enough data to properly refine and market news-personalization services, particularly while adjusting data for each new wave of content-delivery platforms. To answer the second question requires acknowledging human curiosity, the repetitiveness of national and print media, and a desire to communicate with others the “important” and the intriguing.

But there’s a trend that transcends technology, as the study author notes (PDF):

Explicit systems, according to Gauch (2007), are held back by the time required to use them, by inaccurate reporting of interests, and by profiles remaining static despite users’ interests often changing over time.

The perhaps utopian of personalized news, free of gatekeepers’ censorship (and their costs), was also implicitly one of little effort on the part of the end-user. To an extent, this isn’t unreasonable: The Internet, especially bent to one’s will (here, we mean preferences and interests), should save time — and deliver a trove of data to advertisers. And just because neither has happened, except in fits and starts, doesn’t mean it can’t or won’t.

But the time and effort required to tailor our Web presence, let along our news interests, cannot be taken lightly. Personalization is still the future whether users personalize themselves or are passive — there are fewer avenues or opportunities to market to anything but a one-on-one audience, and personalization done right offers a modern replacement for the advertising targeting that print was long able to promise. To put it simply, as the study says:

[T]here are commercial advantages to customization: it can provide rich data on audience interests and demographics, enabling more precise targeting of advertising.”

I’m of the mind that “news personalization” has, perhaps unintentionally, been a continuation of legacy media’s inability to adapt to changing conditions — a refusal, in fact. To me, it rarely works in practice to deliver what readers demand, which studies show over and over does include important news, properly presented.

Instead, “news personalization” in legacy media has often been a sleight of hand: A different presentation to make users believe they’re selecting the same content that’s always been produced, or hyperlocal or “user” sites that crank out content on the backs of low-paid or unpaid citizen-journalists. The efforts made are like most others in the past 15 years: To make the same pitch to readers in a different way, in the hopes that this one convinces/fools them. This is not to say there haven’t been some real successes, and the study highlights some of them. But, particularly within smaller media outlets, there has been more rebranding than revitalizing.

A real change in content production, delivery and targeting is what news personalization truly requires, particularly as the dominant platforms — mobile, Facebook, Twitter, the iPad — are both individual-focused and appealing to collectives.

The solution is not algorithms alone, nor is it turning back the clock. I don’t even have a full-on solution. What I do propose is something oft-discussed, infrequently practiced and requiring great effort to tap its equally great potential: Interest-based targeting backed by curation.

More to come on this.