Publishers taking content aggregation more seriously

On our editorial desk, we're starting to have serious conversations about content aggregation. That's an alien concept to most editors who are steeped in the notion that if they didn't write about it themselves, the story doesn't exist.

Of course, they can provide tremendous value developing a list of sources, monitoring those sources through modern techniques such as RSS readers, and curating or aggregating the best of those sources and presenting them to their readers.

Against the backdrop of a debate raging in content and search circles about original versus aggregated content, Hearst Entertainment has launched an interesting experiment that's 100% aggregation. The site is called LMK (for "Let me know") and it consists of 100% automated content from other, non-Hearst Web sites organized around specific topic areas.

Hearst does employ human editors to curate the initial set of sources, but there is no ongoing day-to-day human involvement. Hearst executives said in an interview that the site "includes very good but not great pages for 2.3 million people, places and things."

I took a look myself. A couple of interesting things. First, the home page barely has any links. It's pretty hard to get into the site from the home page. Which says it's a search engine play through and through.

I clicked on "Barack Obama" and got a plain-looking page:

The news articles consisted mainly of Reuters and AP articles (due to their frequency of posting vs. other news organizations) plus one BBC and WashingtonPost.com articles. So if you're looking for diversity of sources, it seems more-frequent-publishing news outlets will crowd others out.

There was a collection of images, too. When I clicked on a few of the images, they opened up in a new window, but mostly I encountered "page not found" errors.

What was interesting was the automated "Bio" section on the right...Lots of "Preceded by" and "Succeeded by", with vital stats such as birthplace, childrens' names, birth date, etc. Many data points were inexplicably duplicated, likely an artifact of uber-automation. In general, it was a bit too automated, and thinly populated, but an interesting experiment.

Then I went back to the home page and clicked on College Football, which was one of the first topics that human editors started curating. Now THIS section was really interesting:

A night-and-day difference. First, it was much more visually engaging. But they had more than just articles (themselves separated into news versus blogs). Scores were treated as a separate content type, visually represented on several scoreboard-style widgets throughout the page, based on several upcoming matches. Articles pertaining to those matches automatically appeared in those boxes.

Photos were also treated much more intelligently here. They appeared in an AJAX-y, lightbox effect:

The underlying photo widget automatically advances every couple of seconds. When you click on the image, you get the above effect. And if you don't close out, that image automatically dissolves into the next image in the sequence, turning it into a highly engaging automated slideshow. It would have been nice to include a caption or a link to an article to know what we're looking at.

At the top of the page, there was a lot of topic-specific information architecture, such as links to individual teams, players and coaches. The information architecture changed to specifically match players and coaches.

While the hypercritical blogosphere will surely pounce on their execution, especially on the compromises that inevitably occur when substituting automation for ongoing human content origination, I give Hearst credit for trying to solve the right problem. That is, how to create topic-specific pages that are presumabl SEO-friendly (though with a complete absence of original content, this could undercut their strategy quite a bit, bit will let SEO experts correct me if I'm wrong).

The thought that went into the design and the information architecture is highly original for a media OR an Internet company, even if the content itself is not. In other words, if I were a college sports fan (which I'm not, as anyone knows me), I would be much more inclined to bookmark some of these pages, versus Google News.

By the way, Hearst plans to monetize by selling ads on these Web pages -- pages which consist of links to other people's content -- sound familiar? I don't blame them in the least. If it's good enough for Google, and it's good enough for Hearst, it should be good enough for you and me.