How Nicholas Carr Uses Google's Scale Argument Against Tim O'Reilly
Have you read this CNET interview with Google Chief Economist Hal Varian?
Varian, who has worked on the design of advertising auctions, econometrics, finance, corporate strategy and public policy at Google since 2002, says that scale, or the amount of data for search and search advertising, at a certain point ceases to matter. Instead, the search algorithms are Google's special sauce.
This bit of minutiae isn't so interesting until you consider that Microsoft has said its 10-year search and search advertising pact with Yahoo, in which Microsoft Bing will power Yahoo's search on the back end, will give the company great scale to compete with Google in search.
Varian says "we're very skeptical about the scale argument." I take that to mean Varian has been part of discussions with other Google executives, possibly co-founders Larry Page and Sergey Brin, as well as Eric Schmidt and Marissa Mayer, about the scale issue. Varian said:
On this data issue, people keep talking about how more data gives you a bigger advantage. But when you look at data, there's a small statistical point that the accuracy with which you can measure things as they go up is the square root of the sample size. So there's a kind of natural diminishing returns to scale just because of statistics: You have to have four times as big a sample to get twice as good an estimate.
He then goes into some statistics before concluding:
So in all of this stuff, the scale arguments are pretty bogus in our view because it's not the quantity or quality of the ingredients that make a difference, it's the recipes. We think we're where we are today because we've got better recipes and we have better recipes because we spent 10 years working on search improving the performance of the algorithm. Maybe I'm pushing this metaphor farther than it should go, but I also think we have a better kitchen. We've put a lot of effort into building a really powerful infrastructure at Google, the development environment at Google is very good.
In other words, the Microhoo deal is not a big deal to Google, which believes it will continue to out-innovate Microsoft and Yahoo. Did we really expect Google to believe anything else?
What's more interesting to me about this is how Nicholas "The Big Switch" Carr uses what I'll call Varian's scale-doesn't-matter argument to bludgeon Tim O'Reilly's points about scale in the Web 2.0 world. It's a wonderful pissing contest between two of the finer minds in high-tech academia. Carr wrote yesterday:
I was reminded, in particular, of one of Tim O'Reilly's fundamental beliefs about the business implications of Web 2.0: that a company's scale of data aggregation is crucial to its competitive success. As he recently wrote: "Understanding the dynamics of increasing returns on the web is the essence of what I called Web 2.0. Ultimately, on the network, applications win if they get better the more people use them.
I had previously taken issue with O'Reilly's argument that Google's search business is characterized by a strong network effect, which I think is wrong. But Varian's argument goes much further than that. He's saying that the assumption of an increasing returns dynamic in data collection - what O'Reilly calls "the essence" of Web 2.0 - is "pretty bogus." The benefit from aggregating data is actually subject to decreasing returns, thanks to the laws of statistics.
This Carr versus O'Reilly battle is too infrequent to be a pull-up-the-chair-and-grab-the-popcorn event, but it's entertaining even in its irregularity.
Anyway, is Varian right that the glut of search data has diminishing returns and that search algorithm superiority rules the day? If so, then Ballmer is wrong.
That isn't a first in itself, but it would mean that Microsoft still hasn't grasped what it takes to win in search. Of course, Bing's engineers, including former Yahoo Qi Lu and Sean Suchter, may have some things to say in Ballmer's defense.
If those guys had to make a what-Ballmer-really-meant defense, what would it be? Or is there no defense for the scale argument?
Read more about this on TechMeme here.