Why Google doesn’t want better search results

There’s been a lot of chatter regarding the web’s need for a better search. That makes sense, since the current system developed by Google is ten years old (if you count from when AdWords was launched). Of course, there have been tweaks, but the basic concept has stayed the same.

But as Vivek Wadhwa pointed out in his article a while back, the reason why Google doesn’t improve search isn’t lack of technology, scaling problems or difficulty in building new algorithms. No. It doesn’t improve it’s search results because better results would decrease revenue.

That’s right. AdSense accounts for around %30 of Google’s revenue. And a considerable part of that 30% is generated by content farms. You know, the kind of publisher which doesn’t care about how crappy content is, as long as it drives traffic, keywords and comes in gigabytes per day. Like DemandMedia.

The situation is really simple: Google sends organic traffic to some content farm. The content farm displays ads, including via AdSense. If the user clicks on a Google ad, the content farm makes 10 cents. Google makes 5.

Cutting out content farms from search results (either algorithmically or by crowd-sourcing this process) would come at a cost for Google. It would negatively impact revenues. And don’t imagine a catastrophe, but losing up to 5% of the revenues because of a cleaning out search results could start a chain reaction for the company. Let’s not forget what happened when the company missed the analysts estimated earnings by 3 cents per share (at this year’s first quarterly earning conference call). On that morning, Google stock took a nose-dive of over 8%.

How about making search social?

How about integrating more social signals with the PageRank algorithm?

How about delivering more information and less noise in search results?

How about filtering out content farms?

Well, they could. But it wouldn’t pay off on the short term.

And if you’re a corporation, the price of your stock tends to be more important than user experience, web democracy and other liberal fantasies. Of course, on the long run, this kind of decision-making slowly and invisibly erodes the value of the company

Google’s challenge right now isn’t to figure out how to improve search. Or how to make it social. That’s a technical problems which is either solved or easy to solve considering the talent they hire. For that matter, even Blekko, a search company with under $25 million in funding and with less traffic than the most popular site in Romania (around 200K per month), has some better ideas on how to make search a lot better.

The problem isn’t how to make it, but how to milk it.

The challenge for Google is to figure out a way of monetizing this social-next-generation search. Find a new business model is something startups do (or try to do). And once you go all corporate and IPO-ish, rolling back to the startup cocoon isn’t quite that easy.

Having products like Buzz or Wave flunk the market test is an affordable risk – after all, they were just a few million in development costs. But when it comes to search – the bread and butter of Google, providing 98%+ of their revenue – they can’t afford to gamble by “fixing it” just because some upset users say they should.

Should I give them better search results?

What do you think? Does Google actually want better search results?

VisageCloud – Face recognition meets Big Data

Bogdan Written by:

2 Comments

  1. May 9, 2011
    Reply

    Spot on. First time hitting you blog, you've just gained a slot in my RSS reader.

    • May 9, 2011
      Reply

      I appreciate it, Catalin. Hope my future posts will also be of interest 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *