Making old content the new black
Why does old content have to fade so quickly? Blogs, zines, news articles all suffer from “out with the old, in with the new” mentality. There is a lot of old content that retains relevancy far beyond its publish date. The problem is, how do we resurrect it?
Simple, stats tracking. Amazon solved this problem a long time ago. They meticulously track peoples browsing and buying habits and offer up recommendations based on trends. Chris Anderson’s article from 2004, The Long Tail will really blow your skirt up. Anderson starts off by explaining how Joe Simpson’s book, Touching the Void, brought little fan fair in 1988 until ten years latter when Amazon noticed the buying trends of a new title, Into Thin Air by Jon Krakauer. Amazon’s recommendation algorithm began serving up Touching the Void as a “buy this book with…” recommendation. Touching the Void ultimately out-sold Into Thin Air.
Why couldn’t the same approach be taken for time insensitive content on sites like A List Apart, NYTimes, CNN, even this blog? How much relevancy should we give old content? Stats have always been fun to look at with tools like Mint and Google Analytics, but when do we start capitalizing on them and using them in clever ways to help direct our readers to the content they so desire? How do we make old content the new black?
Comments
Brooks
Interesting. Any idea how to pull it off? An adjacent idea: could OpenID be used to provide cross-site recommendations?
Adam Spooner http://adamjspooner.com
This is purely hypothetical, but wouldn’t it be feasible to link into one’s Mint database, extract the related/desired info, and display it? Or someone could even write their own algorithm, in their language of choice (Python, Ruby, PHP, what have you), based on page views/relevance in related material and do the same? I agree that there is a lot of “old content” out there that is still quite relevant, and it would make for better browsing if it was easily retrievable.
Nathan Borror http://www.playgroundblues.com
@Brooks - I’m not sure. Thats an interesting thought though. In PIP you could create a new category that handled interests and go from there. The problem lies in getting others to follow.
@Adam - Very feasible indeed. I’m not familiar with Mint but I’m sure this could be done. I’m kind of rethinking the value of having that data on my server because I use Reinvigorate right now and they’re lacking an API.
I think the biggest hurdle is identifying what content is “timeless” and what is not. Maybe its a tag or a category that you apply to each piece of “timeless” content but I’d rather it somehow be determined programatically. Just not sure how.
Karl Peterson http://www.sidearmdesign.com
Nathan, I love the article. One thought on automatically identifying “timeless” content is to check the frequency of new comments on old content and set a priority for each piece of content based on that. It would seem that articles or pictures getting comments on a regular basis over a period of time would be considered “timeless”.
Another idea would be to add hyperlinks to articles or websites that may have spawned the idea for your content. As soon as those links start going dead the article would start loosing it’s importance. This method might work well for writing about technology. For instance, if you were writing about css 2.0 you could link to w3 and as soon as they switch over to css 3.0 the link would die. Kind of cumbersome, but it might work.
Nathan Borror http://www.playgroundblues.com
@Karl - Tracking activity would be a pretty good indicator. I think thats the best way so far. The link checking idea is clever but I’m not so sure how reliable the results would be since a lot of sites are adopting the religion of links should never change.
Johan
Just use a good site index and clever tagging?
stats dont mean anything except web design params. like resolution, …
Nathan Borror http://www.playgroundblues.com
@Johan - I agree with good application of ontologies but I think a lot can be gleaned from content stats. The main thing being how well a piece of content is performing over time.
Johan
From stats I found content positioned on top of page gets visited more, also linked content.