Printed Matters

On the heels of my post of a couple of days ago outlining a new business model for news, based on Jeff Jarvis’ concepts of “doing what you do best and linking to the rest” and reverse syndication, I now present a list of best practices. Mindy McAdams has a list of best practices from an editorial, or curation, point of view. My list is more about the mechanics…how the links should actually be done, from the points of view of SEO and of encouraging a new link economy.

Suppose you are an online editor tasked with providing a package of, say, some of your own material, plus links to the best coverage on other sites.

1. Link to the original. Whenever possible, link first to the story that broke the news. If it’s copy from CP or AP (etc.), link to the story on their site only. Don’t link to the same story on some other site that happens to be a subscriber to the newswire. This entails you must look for the original source. Sometimes that will take work. Don’t shy from it. And remember, what appears first in Google News et. al. is not necessarily the original source.
2. Then link to articles that advance the story. Never link to duplicate material. Only link to stuff that contains new information or advances the story in some way. You’ll have to read the copy to determine if there’s something new in there, something first coming to light or being connected in that story and nowhere else. Do not link to rewrites of the original story, unless they contain something new. Especially never link to exact duplicates of stories you have already linked to.
3. Only link to the canonical URL of the story. Do not link to the front page of a site just because the story appears there. Find and use the unique, permanent URL for the story. Do not link to a page that redirects you to another page. Don’t blindly copy/paste a link; follow the link to see where you wind up. Don’t link to the “Print this story” page for the story. If there are variables in the URL, strip as much of them away as possible before linking to it. Especially important, strip away anything that relates to your own browser or session. Check, of course, that the URL still works and presents the story you’re trying to present.
4. Use JavaScript to insert your referer ID. In a reverse syndication scheme, you will share in the revenue derived from sending traffic to another site. But that can only work if the other site knows it was you that sent the traffic. To accomplish this, normally you will put some kind of referer ID in the URL so that the receiving site knows where the visitor came from. But that conflicts with #3 above: only link to the canonical URL. With correct use of JavaScript, you can do both. Googlebot will only see the canonical URL and distribute PageRank on that basis, while JavaScript-enabled browsers will carry the referer identification with them when they go.
5. Don’t link to pages behind a pay wall. Never. We want to discourage this anti-web behaviour. Following such links is pure frustration for users and will make them not trust the packages you’re presenting. This rule applies to stories currently behind a pay wall, as well as those that will go behind a pay wall once the story stales, a la The Globe and Mail.
6. Don’t be afraid to change the headline. In the link you are putting together, the link text should tell the user something about why they should follow that link. Avoid the temptation to simply use the headline on the story you are linking to – it may not tell the reader sifting through your package why they should go there.
7. Write your own snippet. Use the snippet to tell the user how this story differs from all the other stories you have linked to. How does this article advance the story? This augments Mindy’s point #3.

Trackback URI | Comments RSS