How Tags Happened at Technorati
It’s been six months since we added Tags to Technorati (where I’m Senior Designer), and as it turns out, it was a pretty big deal. So before we get too far away from it, here’s the story of how it came about. From my perspective, anyway.
Firstly and most importantly, Technorati did not invent tagging. We were inspired by the tags that Flickr users were using to describe their photos, and the tags Delicious users were using to describe their bookmarks, and the many tagging adventures that came before them. We thought bloggers should have something similar – an open standard for adding tags to their posts. If there was such a thing, we could display all kinds of different kinds of content on the same page – photos, links, and posts – grouped by tag.
Secondly, it’s important to note that many people at Technorati worked on various tagging solutions at different points. So credit goes to the company as a whole. We’re a small company now and were even smaller six months ago. Just about everyone had a hand in our tags implementation.
For me, it all started with New Year’s resolutions. In Fray, we’ve always had a New Year’s resolutions story, and it was always a big hit with posters. In December 2004, I was in my second month at Technorati, and I had an idea: Why not encourage people to post their resolutions to their own blogs, and then use the power of Technorati to gather them all together on one page?
Over Technorati’s winter break, Tantek Çelik, Jason DeFillippo, Bradley Allen and I met at Crepes on Cole and banged out the Resolutions 2005 page with help from Kevin Marks and Aaron Bannert who were there via IM. The page was set up to show any post that contained a link to it – in other words, if you linked to that page, then your post appeared on that page.
The page went up on December 29 and we encouraged people to post their resolutions and include a link to that page. And they did! Hundreds of posts came in. It was great. But the system we’d devised had one critical flaw.
There were two kinds of posts that linked to our resolutions page. The first was what we’d wanted – people posting their resolutions and linking to our page for more. But the second was different – it was just people saying “look at all those resolutions over there.” It was not a participation in the theme – it was just a pointer.
What we needed was a simple way to tell one kind of a link from the other. Tantek mentioned the “rel” standard for hrefs that he used in his XFN work. Basically, the rel attribute was a way to describe the relationship implied in a link. With XFN, I could say that Tantek is a friend of mine by putting “rel=friend” in a link to his site. I suggested we just do the same thing here, using “rel=tag” to allow a blogger to say “with this link, I intend to tag my post as being about the subject I’m linking to.”
The best part about this technique was we could read the tag from the location in the href. So if someone wanted to tag their post “iPod” they could link to any URL that ended in that text, whether it was our tag page (technorati.com/tag/iPod) or the product page at Apple (apple.com/ipod) or the Wikipedia entry (wikipedia.org/wiki/Ipod). All would result in the post getting tagged as being about iPod.
We were making the taggers do a little bit of work to be included, but it made sense to ask the people who wanted to participate to do the work, instead of the people who just wanted to make a pointer.
In the first week of January 2005, Technorati founder David Sifry and coder Kevin Marks sat down and kicked out a beta version in a weekend. Dave wrote a service that grabbed the feeds from other tag providers, Kevin coded up a spider that would crawl blogs looking for those rel tags. Kevin also added an awareness of categories in RSS and Atom to the spider, so people could use those, too. I designed some templates to encourage fun browsing.
Tagging in Technorati was released on January 14, 2005. And we knew at the time that any search service could read the rel=tag standard. We wanted them to! The success of tags would be good for us, good for bloggers, and good for the web in general.
Since then it’s been one of our most beloved features, and not just because it’s a browsing experience as I wrote back in January. It’s because tags are carefully created visible metadata that, for the most part, you can trust. When a blogger says their post, photo, or link is about iPod, you can generally believe it.
Together we’re creating a web that’s both more organized and more human. A web where the content creators are in control of how their words are categorized, not some academic in an ivory tower. A web where the difference between a reader and a writer gets blurrier every day.
And I’m so happy I could play some small part in helping it along.