(Rich Gordon)

Over the past year, I've been arguing that news Web sites should strive to build digital networks. On the Web, there are two, overlapping types of networks: content (Web pages connected with hyperlinks) and interpersonal (online communities and social networks). I've contended that network-oriented thinking will produce larger, more loyal audiences than trying to build a "destination" Web site.

For a Web publisher, digital networks are effective for two reasons. First, they help online users discover relevant content on your site. If your content is widely distributed on, or linked from, other Web sites, people are more likely to find it. And second, once a user is looking at content on your site, you have an opportunity to direct them to other relevant content beyond whatever it was originally captured their attention.

But while these approaches should help publishers build traffic, digital network building doesn't fully address one key fact about Internet usage. While online users visit many different sites, each one generally has a small number of (to borrow a term from the cell-phone world) "faves." These are the sites that you visit regularly - a default home page, a bookmark, a link from an email newsletter you read regularly, a URL you type in because you know it by heart.

We know from television research that people don't take full advantage of their content choices. For instance, Nielsen Media Research reported last year that the average U.S. household had 104 cable channels - but tuned in just 15 of them for more than 10 minutes a week.

The disparity is even greater online. There are millions of Web sites, but traffic is heavily concentrated on just a few. The graphic below, courtesy of Matthew Hindman of Arizona State University, shows nicely how news and information on the Web is dominated by a few heavily trafficked sites.



It would be interesting to know how much Web site usage is generated by people clicking on links, and how much by people who make a habit out of visiting certain sites. If definitive data exists on that point, I haven't found it - but a recent academic paper provides some food for thought. Mark Meiss and colleagues at Indiana University collected anonymous usage data from 100,000 users on the university's network. They measured almost 1 billion "HTTP requests" by human users (excluding search-engine spiders) over about seven months. (Generally, an HTTP request is either for a page or for a graphic or image embedded in that page.)

The Indiana researchers found that more than half (54 percent) of all HTTP requests are NOT the result of clicks from another ("referrer") page. It's not clear if the same percentage would apply to page views, but it's worth pointing out that once someone is viewing a given Web site, clicking on links to other content on the same site will generate "referrer" HTTP requests. So there's reason to believe that if you account for all usage on people's "fave" sites - not just the first page visited - it will make up even more than 54 percent of all Internet traffic.

What are the implications of this research? For one thing, it means that - as my colleague Limor Peer has argued - that even in a medium with millions of channels, it is possible to build a destination Web site. Limor and I had a friendly disagreement on that point last year on this blog, when I argued for building networks, not destinations, on the Web.

The idea of "faves" gets me thinking about the importance of:

1. Habit. In the print world, the decision to subscribe creates the conditions for a habit of reading that publication. The publisher still has to create a winning product, but regular delivery significantly increases the odds that you will open the publication regularly. It's harder to build habit online, but not impossible, as the Indiana research shows.


2. Brand. In the world of marketing, a brand is a cognitive shortcut - meaning that people who understand and value a particular brand don't need to devote much thought to alternatives. Developing the right brand image - one that translates into, "I need to check this site regularly" or "This site helps me manage my life" - is critical to making your site a "fave."


3. Getting inside consumer's heads. Research here at the Media Management Center demonstrates the value of "experiences" as motivators and inhibitors of media usage. The research into online experiences shows that focusing on key experiences can help create a Web site that people are more likely to use.


4. Audience fragmentation. In the mass media age, because technology constraints limited the number of publishers or broadcasters in any market, it was possible to build large audiences with a single product - say, a newspaper or a broadcast news show. On the Web, different people will inevitably have different "fave" sites, varying based on who they are, when and where they access the Internet, what they're interested in, and what problems or needs they have.

With these issues in mind, we can identify some ways of increasing the chances that a Web site will become a "fave":

Have multiple sites with different URL's. Traditional media thinking argues for having one Web address that serves as a gateway to all of your content. This leads, however, to bloated home pages with hundreds or even thousands of links. Many publishers are finding that they can more successfully build audience by having multiple URL's. One example: the Monroe (Mich.) Evening News decided to build an entirely separate site, Monroe Talks, for conversations and blogging - and after just four months, the new site had more traffic than the old one. This indicates that the audience interested in talking about the news may be more important than the audience interested in reading the news. (More on that point below.)

Publish breaking news prominently - and not necessarily just news from your town. A site that satisfies people's desire to be "in the know" about the latest news is more likely to be visited regularly. The conventional wisdom these days is that local Web sites should focus just on local news, but breaking news may be an exception. Even a local news site may be able to build audience by making sure that the most important and/or interesting national or foreign news is included in the breaking-news area on the home page. If it doesn't, people will bookmark some other site for their breaking news "fix." At Vorarlberg Online, the Web portal operated by Vorarlberger Medienhaus in Austria, the editors have a "three minute rule," meaning that breaking news must be posted to the site within three minutes.

Use email newsletters. For most people, the online application used most consistently is email - whether it's desktop software like Outlook, or a Web-based service like Gmail or Hotmail. Email newsletters are the digital equivalent of print subscriptions - they arrive on your digital doorstep regularly and therefore, help build a habit.

Enable communities and conversations. Our research into media experiences finds that many transcend media - for instance, people will use newspapers, magazines and Web sites more if they are deemed credible and trustworthy. But two experiences are truly unique to online. One experience, which is about participation and communication, can be summarized by statements such as “I do quite a bit of socializing on this site” and “I contribute to the conversation on this site.” The other, which represents people's interests in hearing what other people have to say, can be summarized as "This site has a strong sense of community to it" or "This site offers a variety of different perspectives."

Think utility. The Web is not a passive medium, meant for reading or viewing. One of the key attributes that differentiates the Web from other media is that it is a medium for *doing.* One ultra-successful site that isn't always recognized for its utility functions is Facebook - its core audience of young adults checks the site constantly to communicate with and stay current with their friends. As venture capitalist Fred Wilson observed on his blog (one of my "faves," by the way), his daughter "uses Facebook the way I use Outlook."

Think games. Competition is seductive, and not just for heavy users who play console games. This is why two of the most popular Facebook applications are Scrabulous and poker. Games are not as popular on news sites, but on a recent visit to Austria, I found executives at Vorarlberger Medienhaus very interested in discussing strategies to improve the company's Web-based card game, the online version of a popular game played in local pubs. The game already attracts several hundred concurrent users during peak hours, but publisher Eugen Russ wasn't satisfied because a competing game (operated by a local brewery) was getting even more.

Finally, I'm wondering whether Web publishers couldn't do more to encourage users to bookmark their sites, or make them default home pages. You could serve up banner ads to to people who've visited a site more than a few times a week, encouraging people to make the site their default home page. Or you could run a contest offering rewards to users who make the site their default home page, or to people whose visits are not triggered by links on other sites. If your site doesn't address recurring user needs, this isn't going to work. But once you have a site that people do find worth visiting regularly, it would be worth trying to cement that behavior pattern.

Other ideas? Leave a comment below.


By Rich Gordon (richgor-at-northwestern.edu)
Rich Gordon is Associate Professor and Director of Digital Technology in Education at the Medill School of Journalism at Northwestern University.