20

Sep

By Dennis R. Mortensen
Defining News Aggregator and News Originator traffic sources in Analytics

Understanding Article Traffic Sources as a news media publisher is an absolute necessity for a number reasons. One of the most obvious reasons, is the eminent need in applying different optimization techniques for every unique traffic source segment. Optimizing for increased Search or increased Social Media performance is surely different – to the extent where you might actually damage another traffic source segment if using the incorrect methodology.

I personally tend to work with 5-8 Article Traffic Source segments when discussing data with news media publishers – sources like Search, Social Media, but also News Aggregators and News Originators. Some of these segments are de-facto defined by todays Web Analytics solutions, but News Aggregators and News Originators are rarely if ever included in that mix. I am most sure both you and I, have an opinion about how to classify referring URLs – at a level where referring traffic from Google News becomes News Aggregator traffic, and that traffic from The New York Times ends up in the News Originator segment.

The apparent black and white classification is achievable without much debate, but it is those web properties on the fringes that need a proper rule set for classification, where one can prove whether it is one or the other. Further to this, a proper rule set provides for automation, which is unquestionably desirable.

I’ve gone with the below set of definitions and unless you tell me I am wrong, the next commentary you see from me, on either of the two segments, will likely to be based on the below.

News Aggregators

Visits from people who arrive from Websites, where more than one-third of their Article Excerpts are outside content (collected or consolidated headlines – manually or automatically. This includes the Associated Press and other wire services). These sites include Google News, Drudge Report, Digg, and Techmeme.

News Originators

Visits from people who arrive from Websites, where two-thirds of their Articles Excerpts are original content. These sites includes The New York Times, The New York Daily News, and Wall Street Journal.

Some Analysts choose to have a classification in between the two, which they name something along the lines of News Commentators, which I decided not to use. If anything, one can think of the News Commentators segment, as aggressive News Aggregators, but news aggregators never the less.

For me there two exciting questions in this; First, do we agree on the definitions? Secondly, which specific segments do you work with, if in news media or similar?

Cheers :-)
/ Dennis (@dennismortensen)

  • Jen

    Interesting as always! I think these definitions make sense, but I am mostly commenting because when I see “News Aggregators” I think of RSS readers. Are those largely not a factor anymore? Readers and other push content/syndication? I’ve always pushed for any syndicated content to have campaign tagging of some sort, but I am not sure how practical that is when syndication is so near the heart of what you are doing online.

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hey Jen,

    Most publishers own and control their RSS & Email segment and optimize the channel to the best of their ability. However, the News Aggregator lottery* is, for me at least, a very different thing. It is ultra short term, comes in bursts only and very difficult to monetize, but so large in volume that it cannot be forgotten. ALSO it is a channel where you cannot demand it tagged (you rarely know when you get picked up.. aka win) and have to rely almost entirely on referring domain or URL.

    *Getting pushed to the front page of Yahoo, Digg or the Drudge Report etc.

    BUT I can see that I need to present the RSS & Email segment in conjunction with News Aggregators, assuring people get my split. Good input. This is exactly what I am looking for.

    Cheers
    d. :-)

  • http://www.allinonestats.com Thomas Bosilevac

    I always get excited to see articles outside the normal topics. While I may not have enough domain experience to disagree with your commonsense definitions, I think, much like all segmentation, it should be up to the website owner. If I see a strong “persona” difference between users coming from sites such as DIGG (complete aggregator, while others my argue social media) and an editorial blog, then by gosh segment them differently!

    Unlike CPC, Organic, Display and other medium segmentation, various website verticals and taxonomies are still in the air.

    As a solution, when bringing on a new client, I normally grab the last 6 months of referring URLS and start to categorize them. This allows me to segment a bit more granularly than most out of box medium reports. Perhaps an upcoming function in Yahoo Analytics?

  • http://visualrevenue.com/blog Dennis R. Mortensen

    >>As a solution, when bringing on a new client, I normally grab the last 6 months of referring URLS and start to categorize them.

    Ha. This is exactly what I do as well, BUT as an Analyst, I would rather have, as you suggest, as set of proper predefined segments, right out of the box.

    Great input.

    Cheers
    d/ :-)

  • http://carsonhsmith.com Carson

    I include a segment isolating portal traffic (yahoo, aol, msn), but wonder if it would make more sense to call that aggregation… I do think editors at my company consider portals as quite different from Google News — they can pitch stories to them whereas Google is like a robot.

    As for as generating the reports, I spend a lot of time creating my own referring source segments in Omniture Discover. I’m not as consistent about definitions as I should be, and I agree it would be great to have more standards. Maybe we should all get together sometime and hash this out? :)

    A side note: One issue I’ve encountered with Yahoo’s portals (not search) is about 2/3ds of the referrals get classified as direct by both Omniture and Google. At first I thought it had something to do with volume, but similar pickups by MSN don’t result in extra direct traffic. Anyone else noticed this?

  • http://visualrevenue.com/blog Dennis R. Mortensen

    Hi Carson,

    >>I spend a lot of time creating my own referring source segments
    It takes time managing a 800 row long lookup table, where the classification of some URLs change and new entrants arrive if not daily, then weekly. I feel your pain mate! :-)

    >>I’m not as consistent about definitions as I should be
    >>Maybe we should all get together sometime and hash this out?

    I am personally an advocate of standards and believe that a few folks within News Media and Analytics can create a set of de-facto standards, which are likely to get picked up by the WAA or similar association for standardization.

    FYI:
    As a Director of the WAA I am actively supporting the following Special Interest Group, that might interest you:
    http://www.webanalyticsassociation.org/default.asp?page=sig_media

    cheers
    d.

  • http://carsonhsmith.com Carson

    Dennis, the interest group looks great. Do you know if there have there been any efforts to organize a group in DC? This town is not exactly overflowing with media companies as NY is, so perhaps not…