Where news websites get it wrong

Pardon the interruption of usual foreign policy chatter.

Major news organizations, including several thought generally to be Internet savvy, continue to miss basic lessons about news inherent to good Internet-based news reporting. In no particular order:

• Link to unfiltered, raw data sources. News articles are synthesis, summary and analysis of raw data sources. It is rare to see a news article that actually links to a reporter's notes, a transcript, or organization press release. In the Information Age, there is no reason not to publish all available data on a subject, even if it is a mess of jumbled notes on paper napkins, covered with mustard and mayo from a sloppy reporter's lunch (scan it in). Transparency is impossible without giving the public direct, unfiltered access to raw information.

• Link to third party websites. Fearing the loss of revenue, news publishers avoid linking to third party websites for fear of driving traffic away from their site. This fear is sometimes justified, but generally not: The "hit and run" pattern, where Web surfers open many links to many different websites, is highly common, probably the dominant usage pattern for news websites.

  The smart web publisher will choose the best link for the given text, regardless of whether or not you control the link destination. Web surfers will come to trust a publisher who supplies high quality links, and return to that publisher again and again.

• Add links for context, key words and background. Readers should always be able to click on a link in a news article, to obtain more information about what they just read.
• Your site is not better than the Internet. The Washington Post is an example of this awful practice. Major personages and themes are made hyperlinks in each Washington Post article. But instead of each link going to a useful page providing additional information, each link simply performs a site-local search. For example, a hyperlink "Joe Lieberman" will produce a list of Lieberman-related WaPo articles. It is far more useful to the reader to click on a link labelled "Joe Lieberman" and receive wikipedia-like background information on Joe Lieberman.
• Maintain a factbase. As events unfold, readers should be able to click on a single link to receive a "backgrounder" that includes events listed in chronological order. Generally, this is wikipedia, but larger organizations may wish to maintain their own "strategic briefings" on various subjects.
• Link directly to a website, even if partisan or controversial. However, clearly warn readers of bias or offensiveness first. It is inexcuseable to write a story about Internet-related news, and not link directly to the subjects at hand. Example: This CNN story which discusses a Youtube video, but fails to link to it. You still have the standard journalism ethics questions to deal with, of course: does linking to a website potentially encourage bad behavior? One must avoid balance the public's right to know with giving terrorists and extremists a platform.
• Multi-page articles are incredibly annoying. The web is not a platform limited by physical page size limits. Web browsers and modern computers are easily capable of opening novels in a single web page. Web users know that you are forcing us to click for advertising and statistics gathering reasons, and we don't like it one bit. Add more advertisements to long web pages ("multiple ad units" in Google AdSense parlance).
• Build a community by enabling comments. However, this unfortunately requires either a digg-like crowd moderation system, or moderators (paid or unpaid), once comment traffic reaches the level necessary to build a community. As a bonus, consider how to highlight comments that you and the community find to be interesting, informative, or simply noteworthy due to the author (i.e. a celebrity comments).
• Inferior search. If your website search feature performs less capably than Google's search feature for publishers, you have a problem. Another major problem with news website search engines is the lack of indexing fresh content. A reader reads a story on Obama, then does a search for Obama, and the returned results are stale (hours or days old).

A key technical problem is that third party linking often falls into the realm of editorial policy. The news wires such as AP and Reuters are probably reluctant to provide links embedded in their wire stories for this reason.