work
travel
contact
I’m not sure if it’s a ‘skill’, a ‘knack’, or maybe an ‘art’ (by this point), but I can say that my procrastination abilities have become quite well-honed at B-school.
Case in point: we just wrapped up our MCP project. The MCP is a full-time effort (and then some); there’s no classes in April because of it. While every group scrambled to finish on time, my group’s project was especially back-loaded; lots of analysis couldn’t even get started until last week.
Goes to figure, then, that I’d decide last week was also the perfect time to ‘tag’ my entire iPhoto library with comments and keywords. Doing so is the digital equivalent of rummaging through a shoebox full of old photos and writing helpful notes on the back of each picture, explaining who’s who, etc. It’s exactly the sort of ridiculous undertaking that nobody ever bothers with — unless, of course, there’s other, more important work that needs doing.
But it brings me around to this: one thing you hear, working in the web industry, is that “Metadata Is Expensive”. In other words, “scribbling notes on old photos is time-consuming monkey-work”. And I can say, from recent experience, that that’s true, regardless of whether the photographs are digital or physical. (Metadata, incidentally, is defined as ‘data about data’; on the web, metadata helps classify, describe, or organize web pages.)
You hear this ‘metadata is expensive’ maxim especially in regards to search. Google, for example, gives a cold shoulder to metadata — it reads only the regular, visible words on a web page, and ignores any behind-the-scenes attempts to categorize a website. There’s a bunch of valid reasons for this, namely:
[A] Google patented software techniques which make their current setup pretty awesome [B] Early web history showed people will cheat and write deceptive metadata to lure an audience.
[C] The notion that ‘metadata is expensive’ to create. It just isn’t worth the time.
Obviously, there’s a bit of a vicious cycle with that final point: I can now imagine plenty of people crafting nice, careful metadata code for their websites, if they thought Google might actually use it.
The main reason I’m increasingly confident in the above statement comes from my own recent behavior vis-a-vis iPhoto: apart from the procrastination element, I did have some good reason to slather metadata across my entire iPhoto library. The soon-to-be-released Mac OS X Tiger will supposedly allow me to search against it. Given how I value those pictures greatly, it struck me as being worth the effort to organize my photos. However ‘expensive’ it was. (Just a few hours’ work, really.)
But Google isn’t going to start acknowledging metadata, I think, largely because of reason [A], above. They’re on top of the search-engine world right now, and won’t benefit from rocking the boat. As for [B], I think the web today is capable of solutions that weren’t on the radar in the ‘90’s. And regards [C], well, like I said: ‘expensive’ is relative. People will gladly bear the cost of metadata on things that they personally value, and that extends off the desktop onto the web. See Flickr.com.
So. I’m increasingly of the opinion that if Google doesn’t do metadata, somebody else will. In fact, it seems like one of the obvious avenues for second-tier players like Yahoo, MSN, and Jeeves to gain some competitive advantage in the search space.
And I want to see what that strategy looks like, if it happens. Especially if it provides me new opportunities to put off doing real work.
