How Google Plans to Stay Ahead in Search

October 10, 2009 12:50 PM

"We ran over 5,000 experiments last year. Probably 10 experiments for every successful launch. We launch on the order of 100 to 120 a quarter.We have dozens of people working just on the measurement part. We have statisticians who know how to analyze data, we have engineers to build the tools. We have at least five or 10 tools where I can go and see here are five bad things that happened." Udi Manber, Google’s vice-president of technology, explains the business of running a search department. "It takes a very, very good engineer about two years to really understand search." From a surprisingly candid series of articles detailing the business of Google, "Can Google Stay on Top of the Web?"

CEO Eric Schmidt on the economy, “If you go back to last year, we began to see slowdown in U.K. quite early. We initially thought it was an error in our system. We made adjustments, and it wasn’t until September and October, when it became clear that it was a global collapse. From our perspective, the low point was somewhere in the spring…somewhere in May or June.”

posted by geoff. (21 comments total)

Google could probably make good money selling sales data to companies like Goldman Sachs. Probably, they already are.
posted by delmoi at 1:52 PM on October 10

Funny, I was just talking about this yesterday with some of my colleagues who are afraid Google will soon rule the world. But it won't. Standard Oil, Ford, IBM, Microsoft--they were all going to rule the world, right? But now? Creative destruction. May take a few years (decades?), but it'll happen. Google's in a hotel room in Saigon getting soft; Charlie's in the bush getting stronger every day.
posted by MarshallPoe at 2:34 PM on October 10 [2 favorites]

Standard Oil, Ford, IBM, Microsoft--they were all going to rule the world, right? But now? Creative destruction.

If by "Creative destruction" you mean "The Federal government squashing them for trying to take over the world and extract economic rents from everyone" Because that's what happened. Even though the government's case against MS got dropped, it still changed the culture over there. Why did the government stop Microsoft from taking over Intuit? Because wallstreet was worried that they were going to horn in on their turf (or so I've heard people speculate)

The reason these companies haven't been able to take over the world is because they have actively been prevented from taking over the world by the government, not some free market mythologizing.
posted by delmoi at 2:41 PM on October 10 [11 favorites]

IBM's problems had far more to do with the marketplace than the government (although apparently just the other day they had another anti-trust probe launched by the DOJ). I don't see any evidence that the mis-steps IBM made wrt desktop computing and PCs were driven by government interference.

Microsoft (who is not exactly struggling) certainly had legal issues, but missing the online services / search / etc market was a strategic, not a legal, mistake. (I worked at MS for quite a while, and you are right that the anti-trust stuff had a big impact on the company, but I'm not at all convinced thats what caused the competitive mistakes).
posted by wildcrdj at 2:47 PM on October 10

IBM's problems had far more to do with the marketplace than the government (although apparently just the other day they had another anti-trust probe launched by the DOJ). I don't see any evidence that the mis-steps IBM made wrt desktop computing and PCs were driven by government interference.

It's not simply a case of not doing "X" because the government says you can't do "X", it's that the constant government supervision, which Microsoft now has creates a cautious culture where people don't even bother with X or Y if they think the government might stop them. (And this is, of course, a good thing for everyone) IBM could have required an exclusive license from Microsoft for DOS.

Imagine this. Compaq and a couple other companies come out with their PC clones, and the market starts to take off. It's too late to get an exclusive deal with Microsoft for DOS so what does IBM do? Well, what they tried to do was come up with innovative PCs (and they did try the whole proprietary microchannel bus thing) but what if instead they had simply used the profits from their mainframe division to subsidize their PC business until their competitors died off, while adding more and more proprietary technology? But that didn't happen.

Why didn't it happen? Because it would have been an obvious anti-trust violation. IBM, had they been "unchanged" by antitrust fears could have crushed the nascent PC industry.

(by the way, that's exactly what microsoft did to break into the video game industry, but there's a difference between getting into an industry, and crushing everyone)
posted by delmoi at 3:11 PM on October 10 [2 favorites]

Also what Sony did, of course, wrt videogames. The "original" videogame companies (those who made their money primarily off of games without being an offshoot of an existing company) were killed off by a combination of Sony & MS, with the obvious exception of the resurgent Nintendo. But, as you say, at least videogames are still a competitive industry, with 3 players who change around positions in various generations (which is pretty much how its always been, although the players have changed over the years as well).

But from my experience at MS a lot of the web stuff was missed just because... well, it was dismissed. Not because of anti-trust or other concerns. Most people genuinely seemed to think it would be an unprofitable business. Also, of course, there was considerable concern about killing the cash cow (Windows) by moving things to the more platform-independent web (which turned out to be pretty accurate I think --- Mac's resurgence is partially due to the fact that people don't rely much on installed software to do what most people use their PCs for now --- web browsing and communication (email/im/twitter/etc)). Not till long after Google became dominant did that idea really shift (out of necessity).
posted by wildcrdj at 4:04 PM on October 10

wildcrdj - Do you not see though how Microsoft not only dismissed the web, but deliberately attempted to kill / control it by going after Netscape and then leaving IE to rot after they won the browser wars?

Name a single other company that could have done what Microsoft did with Internet Explorer. It's taken practically until present day for Mozilla/Apple/Google to turn the tide against Microsoft and open the web back up

You say their ignorance of the internet/web was a strategic mistake and that it was "genuinely" missed, but you seem to simultaneously acknowledge there was a lot of self-interest in marginalizing it, because as Microsoft understood over a decade ago, an open application platform is anathema to their monopolistic business model
posted by crayz at 4:30 PM on October 10 [4 favorites]

Automation won't keep Google on top of the web, attention to detail will. But that means lots of very engaged staff, and from what I hear about cutbacks in the free-goodies department, that era is coming to an end. There may be thousands of tests but Groups search was still fundamentally broken for a long time, right up until Wired news exposed just how broken it was.

It's often the same story: managers think tech can replace engaged workers, the workers get pissed off, the company loses its edge and the decline begins. This is how Google will start making the mistakes that will allow an upstart in.
posted by fightorflight at 5:33 PM on October 10

wildcrdj - Do you not see though how Microsoft not only dismissed the web, but deliberately attempted to kill / control it by going after Netscape and then leaving IE to rot after they won the browser wars?

That's a pretty good point. The web would probably be a much better applications platform today if MS had stayed out of it. I mean, we'd probably all be able to run Java applets, for one thing.

And what was the main goal of the Microsoft antitrust case? To uncouple windows from IE. It might not have been all that successful, it was the fact that firefox was so good, and IE was basically left to languish that people moved on.

People who think that the market is really all self-correcting and that someone will always come up and eat a larger companies lunch are viewing corporate history rose tinted lenses. Without restraint, a huge company would always be able to crush an upstart.

Also, Google has already brushed up against the antitrust division of the government, when they tried to take over yahoo's ad business when Yahoo broke off the Microsoft deal.
posted by delmoi at 5:41 PM on October 10

Microsoft (who is not exactly struggling) certainly had legal issues, but missing the online services / search / etc market was a strategic, not a legal, mistake

David Banks' Breaking Windows: How Bill Gates fumbled the future of Microsoft is an excellent read on this subject. Though the title is misleading as the book gives ample time to the cultural problems within the company, as well as Gates' mistakes. A company's DNA can definitely cause it to miss opportunities.
posted by Blazecock Pileon at 5:49 PM on October 10

IBM's problems had far more to do with the marketplace than the government (although apparently just the other day they had another anti-trust probe launched by the DOJ).

Maybe it is so long ago that it is out of your mindset, but in the 60's IBM basically ruled the entire computer industry. Then there was a large DOJ anti-trust case against IBM, which IBM lost.
posted by eye of newt at 6:10 PM on October 10

Delmoi has a point, and I certainly agree that a pro-competitive legal framework is an essential part of making markets work effectively. The gov goes after monopolists, and that's good. But aside from out-and-out breakups (Standard Oil), my impression--which is based on no research and little knowledge, I admit--is that competitors usually destroy market-dominant companies, and they usually do it via innovation, and often innovation coming from an unexpected place. I remember when Japanese cars were a joke, GUI interfaces were for idiots, the Internet was truly useless. The folks at GM, IBM, and Microsoft continued to think those things long after they weren't true, and they suffered. The anti-trust prosecutions helped (though as far as I know GM was never sued), but competitors beat those company, and they did it by innovations. So I'm confident that Google will eventually get out-Googled.

The importance of both anti-trust legislation and a highly competitive marketplace is of course shown by countries that had neither and therefore were not very innovative--e.g., the Soviet Union (or really any Stalinist state).
posted by MarshallPoe at 6:46 PM on October 10

The gov goes after monopolists, and that's good. But aside from out-and-out breakups (Standard Oil), my impression--which is based on no research and little knowledge, I admit--is that competitors usually destroy market-dominant companies, and they usually do it via innovation, and often innovation coming from an unexpected place.

Two of your examples, IBM and Microsoft are still around, and GM wasn't killed by Japan, they survived that in the 80s/90s.

The point is, without anti-trust controls big companies can just wait for other people innovate, then either crush them or buy them. What if IBM had bought up Microsoft, for example? Or if Microsoft just bought Netscape? They certainly could have afforded to do it around the time AOL did. Why do you think they didn't?
posted by delmoi at 6:56 PM on October 10

This is why IBM is under anti-trust fire: in the mainframe world they're the only game in town, because they killed all their competitors and they buy up anybody new or innovative who appears. Hey're also jacking prices through the sky for all their remaining customers.
posted by fightorflight at 9:19 PM on October 10

right up until Wired news exposed just how broken it was.

Link?
posted by afu at 12:42 AM on October 11

To be fair it was initially slashdot that exposed the brokenness of Google Groups' search. Wired article and followup.
posted by Rhomboid at 1:51 AM on October 11

I don't see any evidence that the mis-steps IBM made wrt desktop computing and PCs were driven by government interference.

The only thing that tells me is how profoundly ignorant you are of the history of computing. The government's ongoing anttrust investigations of IBM is the main reason that IBM were prepared to outsource the development of a desktop OS to third parties, instead of doing it inhouse, the only reason EMC exist, and probably the same for HDS, amongst other issues.

Maybe it is so long ago that it is out of your mindset, but in the 60's IBM basically ruled the entire computer industry. Then there was a large DOJ anti-trust case against IBM, which IBM lost.

And then Reagon threw it out eventually.

I find it amusing/horrifying that Larry Ellison openly says he wants Sunacle to be IBM like it was in the 60s. Maybe great to hear if you're an Oracle shareholder, but I can't think of a comparison more likely to have existing Oracle and Sun customers checking they have a viable platform exist strategy just in case.
posted by rodgerd at 3:11 PM on October 11

(Speaking of Google being broken, I wonder how it is that the truly crapulent Expert's Exchange manage to game them so succesfully?)
posted by rodgerd at 3:12 PM on October 11

I haven’t checked, but do they run AdSense?
posted by Tobu at 5:45 PM on October 11

Expert's Exchange is actually gaming you, rodgerd. If you scroll down (way, way, way down) the answers are there as long as you got the link from a Google search.
posted by jewzilla at 8:06 PM on October 11

To be fair it was initially slashdot that exposed the brokenness of Google Groups' search. Wired article and followup.

This was interesting, and thinking about it again, it seems to me to be a Six Sigma type of criticism. That's when, regardless how good and useful your primary work may be, you are held to account for the shortfalls of something you've devoted less attention to. Thus your work is characterized by the worst of it. That's the exact opposite of reason, as far as I can tell. But it's a useful tool if you want to hold people in check, and make them worry about everything. In this case, given what Google has brought to everyone, I find such criticism to be entirely inappropriate. Undoubtedly Google compared the amount of users of Groupswith that of their other products, and devoted their resources accordingly.

To correctly perceive the quality of Google's work, I would compare them to other companies who have reached a point that they feel their work is finished and have quit devoting any resources to the quality of their primary product, in an effort to minimize costs and diesel on collecting revenues long enough for their partners to retire to a comfortable private life.
posted by nervousfritz at 8:42 AM on October 13