The Side-Effects of Google Instant

In trying to match the speed of your mind, the search giant just gave its servers up to 20 times more work. That’s not great for its energy bill or its carbon footprint. And all those seconds you’ll save on searching? You’re likely to spend them on Google anyway.



If you haven’t tried Google Instant yet, you’re either not much of a Net surfer or you live outside the U.S. (where it seems Instant is slow to roll out). It’s impressive, intuitive, and a little annoying, almost as if Google is trying to guess what’s going on inside your head. Google’s CEO Eric Schmidt hinted, back at the launch of real time Twitter integration, that Google’s ultimate goal was to achieve search at the speed of the mind–and Instant is pretty much as close as they’ll get for the moment.

Still, Google has one disadvantage when it comes to matching your brain. It relies
on millions of slices of silicon to do its basic search tasks. And Google Instant just makes its server burden much worse.

The idea is that you’ll save many thousands of seconds every year–a couple each time you access Google to look for something. Google serves up countless millions of search matches every second. Multiply that by a few seconds per user, and Google is potentially saving the world billions of dollars worth of work-hours.

But this effort inevitably costs Google. It costs a lot. Its entire search system has to work more swiftly, and serve up dynamically-changing lists of search results as you type into its query box–not only does it have to dig them out of its archive, but it also has to format them and push them to your browser in a time fast enough to be relevant (before you type the next letter is a good measure of this speed.) During the Instant presentation Ben Gomes explained that the server burden has risen by somewhere between 10 and 20 times.

That’s likely to increase Google’s carbon footprint, not to mention its electricity bill. All that server power has to come from somewhere, even while Google uses tricks (like biasing its guesses towards words that’re more likely than others–guessing “snow” is more likely to follow when you type “s” than “syzygy”).

As for all those spare seconds that’ve just been given to you, you may well end up spending them on Google anyway. Why? Two reasons. Firstly, as Google Instant “guesses” search results that match your incomplete query, you’re likely to notice sites popping up that you wouldn’t have seen before (even if they’re only tangentially linked to your intended phrase).Human curiosity being what it is, you may well click on one of them.


Secondly, now that Google delivers things so very swiftly–often beating you to the punch–you do get a genuine feeling of satisfaction from using it. And that may be a gentle prompt for you to try out a few clicks on Google that you may never have tried beforehand–such as a quick image search.

If Instant does cause that kind of halo effect, Google’s server burden may be yet higher than the company suspects. And that’s not even accounting for any new Internet memes based around the real-time search results. Will Instant be the straw that broke the back of Google’s server farms?

To keep up with this news, follow me, Kit Eaton, on Twitter.

About the author

I'm covering the science/tech/generally-exciting-and-innovative beat for Fast Company. Follow me on Twitter, or Google+ and you'll hear tons of interesting stuff, I promise.