January 25th, 2009

Why local-news aggregation is useful information, not information overload


My post on the Washington state linking project focused on the awesome innovation involved and on the benefits of collaborative linking in general. But the project also shows why this kind of news aggregation is so useful for a local audience.

The biggest danger with news aggregation is that instead of acting as a filter, it can sometimes add to readers’ information overload. I read Andrew Sullivan’s blog as much for his links as for his original posts, but some days his link-blogging is just too prolific for me. (As Howard Owens put it: “To all the bloggers in my RSS reader: You post too frequently. Stop it. Let me catch up, for a change.”)

Commenter Matthias Spielkamp worried that the Washington link project might have had this effect: “I don’t have the time to read through 25 different stories to get a picture of the situation.”

But we shouldn’t mistake a long list of links for confusing overload just because it looks like overload from afar. The closer readers are to a story or event, the more they want to know about it and the less overloaded they’ll feel.

Take a look at the headlines on the current Kitsap Sun flood widget: “W. Wash. flood clean-up information”; “Maple Valley firefighters rescue residents from flooded homes”; “Flooding, for the most part, misses Federal Way”; “At least 500 Snohomish County homes flooded, officials say”; “Flooding halts Whidbey sports action.”

Each of those is a unique story that’s going to be very useful to someone living in that area. None of them repeats the others. Chances are, someone who lives in Maple Valley has friends or relatives in Federal Way or Whidbey. What looks like two dozen versions of the same story to Matthias or me is vital information for people who live in the flooding area.

Jack Lail found the same level of interest for Knoxnews.com’s link roundups on the Tennessee Vols. To me, the roundups look like overload. But they get tons of page views, Jack noted, “because fans are passionate and can’t get enough information on their teams and games.”

It’s kind of counterintuitive, given the emerging consensus that Matthias noted in another comment about the importance of clean news sites that don’t overload the audience. What good human-powered local aggregation does is feed readers’ deep interest in such topics without overloading them with 50 versions of the same AP story.

Comments (3 Responses so far)

  1. Josh–

    I understand your point. But I’d say this, and it applies to a link as much as any other piece of content:

    Editing means making hard decisions on your reader’s behalf.

    Just because a “related links” widget can be of any length doesn’t mean that lower-value content should be included.

    In my years as a writer/editor, I’ve found that a firm length limit forces better choices. I haven’t done this particular exercise, but I’m guessing a list of 5 well-selected links delivers more reader value than 10 links.

    You raise the excellent question of whether hyperlocal content should be included to expand the service to specific groups. I’d still argue that, regardless of the answer, 5 links still beat 10 in most cases.

    A reader’s attention is the most precious asset on the web. Editors ignore this it at their risk.


  2. Craig — I agree generally. After all, the main benefit of human-powered aggregation is that it’s theoretically vetted and curated links, vs. an algorithmic link dump. You risk losing the benefits if it becomes a human-powered link dump.

    What gives me pause is two things. First, however many total links are in a given roundup, you can always have a smaller number in a widget at any one time to avoid at least the appearance of overload.

    More importantly, the few data points we have suggest more is sometimes better: the Washington state snowstorm link roundups in December (on the homepage for 3-4 days, but most viewed story on Sound Publishing’s site for the whole month); Jack Lail’s Vols roundups (twice as many views as the next-most-viewed story). The Washington crew didn’t have numbers at the time, but I believe the waflood roundups were also very popular.

    What these suggest is that for some local topics or stories, the interest level has moved from a general “I’d like to be familiar with this topic/story, so give me a handful of good links” to a more niche-interest “I want to know everything there is to know about this topic/story, so have at it.”

    I don’t think every local subject or event warrants the latter level of linkage. Though on the other hand any subject, no matter how obscure, has its devotees and niche followers.

    So there’s a lot of potential and flexibility: you can have a widget with five links within a story or on a topic main page, but then have a secondary page with 20 links on it for the people who are really interested.

  3. Just wanted to chime in and say great site!

    And oddly enough, I found your site by searching “interesting twitter” or something along those lines in Google and it sent me to your post from December 2007! So as much as you dislike Twitter (i’m still trying to figure out what the damn thing is and if it’s worth it), it just brought you another reader!

    I like what you’re doing. Keep up the good work.

Add Your Comment


Receive new posts by email