THE UPSIDE DOWN: Search Me –when did Google Become So Shallow?

As Facebook and other global tech giants come under scrutiny over the way they amass our data, John Mitchinson reveals how Google has less information.

In the early days of QI, when there were just a handful of us chasing down the killer fact, we used to love ‘going out back’. This is a surfing term for paddling out to the calm waters behind the breaking waves.

For the early QI researchers, this meant the wild hinterland of broken links and eccentric sites you ended up with if you followed a search as far as you could take it, sometimes hundreds of pages into Google’s rankings. Much of it was junk, but not all. Every now and then you’d find an obscure site that unlocked a subject or answered a question.

It was fun, like rummaging in a second-hand bookstore.

For old time’s sake, I try a QI-type query: ‘why don’t the Yazidi eat lettuce?’

In 2002, when we set up QI’s first office, there was no Facebook or Twitter, and Wikipedia had only just launched. Google themselves were less than five years old. They were smart, apparently ethical (‘Don’t be evil’) and their search engine was a game changer.

Instead of ranking results on the number of times a search term appeared on a page, Google’s algorithm counted the quality of links to a page—it was these ‘backlinks’ that revealed a page’s relevance. They made portal sites like Yahoo!, Lycos, and AOL.com look slow and venal. We loved them, playing happily for hours inside their cleanly designed, quick-to-load pages.

But searching has evolved and the internet has turned out to be more perplexing than the warm sea of facts we swam in then.

Now we search all the time, mostly on smartphones. There are 73,000 online searches per second, 92.6% of them on Google, generating $490,000 per second in advertising revenue. Like the store selling picks to prospectors, Google has grown richer than the sites it guides us to.

Thanks, Google version 2019 – bigger, richer, shallower.

Search isn’t about disinterested serendipity: it’s about anticipating what people want, basing it on the preferences previously revealed by search history, location and any other data gathered (Google ad trackers are hidden on 75% of the websites we visit).

As a result, a whole secondary industry has grown up, SEO – search engine optimisation – now worth $80 billion per annum (the same size as beauty or fitness), essentially helping businesses game Google’s algorithm for profit.

Sergey Brin warned us of the direction of travel in 2013: his original vision was ‘that information would come to you as you need it. You wouldn’t have to search at all.’ Nope, we’ll just suck it out of you and sell it on to the highest bidder.

For old time’s sake, I try a QI-type query: ‘why don’t the Yazidi eat lettuce?’ I get a mere 7 pages: 46 results visible out of a total 20,600. Possibly more ‘accurate’ than 15 years ago—there at least three possible answers to this question—but thin. I click on the Shopping tab (that didn’t exist in 2002). The top result is a book: Skinny Chicks Don’t Eat Salads.

Thanks, Google version 2019 – bigger, richer, shallower.