I’m a big fan of The Guardian, but their auto-linking software needs some tweaking. It’s a classic example of trying to provide that extra value to data on the cheap.
My argument for a while has been that the only lasting way for traditional media to make itself competitive again is not to create more, but to create better.
In one key sense this is about injecting extra value into words: metatagging them, in short, so that other content belonging to the media—or others—adds context.
But this is not easy. Lots of people are trying it, and some are doing interesting stuff with it. But building a library of words that creates automatic links to categories within the one site, as The Guardian is doing, is not it.
Take the example above. It’s in an article written by a woman who has given up sex for a year (neat and easily sellable book idea, or what?). But in the example above, where she’s talking about her lack of love life as a young journalist (tell me about it) she mentions her dating experiences.
The Guardian’s autolinker parses the story at some point and inserts a link for the word ‘dating’ to the paper’s ‘Lifestyle and Dating’ section.
Another example appears lower in the story, where relationships are mentioned, leading inexorably to a link to the section on Relationships:
Now there’s nothing wrong with this story appearing in either of those sections (and it does), but to autolink these words to the section is meaningless. It’s out of context. It lacks context. It’s not contextual. It’ doesn’t add value.
Indeed, it cheapens all the good linking that is going on in The Guardian, because it reduces the reader’s trust in the value of all those links.
If you as the reader start to see links all over the place to places that don’t add value to what you’re reading, pretty soon you’re going to stop seeing those links.
So, Guardian, drop the autolinking bot and spend time thinking up a better way of adding value to your content. Metadata is too valuable, too important, to leave to cheap bots.