• by David Berkowitz , September 5, 2006

You think tagging's a game? A plaything? Something people do for fun?

So does Google. In other words, you should take this game very, very seriously.

Last week, in the column "Tag, You're It," I wrote, "When tags come together in communities and then diverge into specialized functions, the organisms they create can alter how we access information."

The Googlers must have been reading.

Okay, it's not just now that Google's getting into tagging. Google Base, one can argue, is entirely built on tags. But Google has this way of not sticking to the nomenclature du jour. In Google Base, you can add labels and attributes, not "tags." Tom-ay-to, tom-ah-to.

Last week, Google released a new game that is all about tagging, but it doesn't say the word "tag" anywhere. The game is Google Image Labeler (thanks to Google Blogoscoped for the link), a simple time killer that reminds me of the types of programs used in psychology experiments at my alma mater.

Google randomly pairs two players together and then shows them an image from its image search database; offensive content is screened out. Each player must rattle off descriptors for the image, and when they each write the same one, they score 100 points and a new image appears. Players can agree to pass if they're stumped. After 90 seconds, the clock stops.

The points don't matter much, except for personal gratification. An average game (for me) would score around 500 points; a great one would score 1,000, which means jointly agreeing on nine images in 90 seconds. The highest scorers notched over 1,500 points.

As of now, the labels from the game aren't appearing anywhere. This is an experiment. But just as was the case with Binghamton University's psychology department, experiments are designed with a purpose. Not every one leads to actionable results, and this one has its inherent challenges. Will enough users participate to provide scale? Will the results be skewed toward linguistic preferences of early adopters, who are more likely to hear about this game? Will points alone be enough of a reward for people to keep going? The game has its limits.

It can also be frustrating. I'm convinced that some of my fellow labelers were illiterate, non-native speakers of English, blind, or had severe mental difficulties. But that's part of the fun of the game, and it's why search marketing can be so frustrating, and so rewarding. The other person, if she or she is short of temper, may well think all of those things about me, and yet both of us may be reasonably decent spellers and English speakers. We just can't agree on the same terms. I'm writing "outdoors" and he's writing "outside." I'm writing "night," he's writing "dark." I'm writing "smoke," he's writing "fire." If we had time to have a conversation, we'd have come to reasonably similar descriptions of an image, but there was no time for that here, which is exactly how it is with search. Through search marketing, we get to learn about how people refer to our industries, companies, products, and brands, so that marketers and consumers can get on the same page.

As for the Image Labeler, it's appearing in many other guises, though Google's the first to outright call it a game. People are already playing it on their own in countless other ways thanks to the Web 2.0 applications that incorporate tagging. Yahoo's photo sharing site Flickr, for instance, released a geotagging feature last week in which users could drag and drop their photos anywhere on a map to tag the images with the location where they were photographed. Just 24 hours after the geotagging feature launched, more than 1.2 million photos were geotagged.

If you build it, they will tag. Whether you're a builder, a tagger, or neither just yet, major companies are making big bets on tagging making information more accessible, so your winning strategy should be to play along.