"The Search" by John Battelle
John Battelle, former publishing editor of WIRED, spent two years and interviewed some 300 people to write his book about search engines and Google. It's a nice book that can be read in about four hours.
The Search by John Battelle. Published September, 2005, Portfolio Press. 310 pages, notes, index. $26. Review by Andreas Ramos, andreas @ andreas.com
Executive Summary (for Larry and Sergey)
An outsider wrote a book for outsiders. Mostly harmless.
More Stuff (for Everyone Else)
The Search is a book that could have been much better. Battelle had access to everyone. He could have written about search technologies, the effect of search engines on user habits, how companies are using Adwords to advertise, and so on.
But Battelle comes out of WIRED, and what we get is a 300-page WIRED article, with the old wild-eyed hyperventilating claims. A few will amuse you: "…all-out war for the market of the future", "search will be the application that finally catalyzes the fabled convergence of television and personal computer", "Search will rewire the relationship between ourselves and government", "the gold from which search companies spin their fabled profits", "…marketing leads are the crack cocaine of business", and "…fraught not only with staggering technological obstacles, … but with nearly paralyzing social responsibility".
Like, whatever. There's plenty more of these in the book, but just pick up an issue of WIRED and just put in Google wherever you see Napster, cold fusion, or Segway.
Everyone has forgotten AltaVista, but we should remember search isn't an easy business. AltaVista was the leading search technology in 1996 and was sold for $2.3 billion in 1999. After the dotcrash, Overture bought the wreckage for only $140m, a 94% loss in value. Good search technology doesn't guarantee anything.
Page and Brin had a working version of Google in the 90s and by 1998, they went around to the main companies, such as Yahoo, Infoseek, etc, but those weren't interested. Battelle makes several good points here: those companies didn't care because they already had a primitive search tool on their sites and that was "good enough". More importantly, Yahoo etc. were portals which wanted to keep users within their websites, using their preferred (i.e., paying) partners for airline tickets, hotel rooms, car rentals, etc. They were monetizing the links within their sites. A search tool that showed better links elsewhere would only lead the users away from the site. And what's the use of that? Definitely a bad search tool. Page noticed these portals offered horoscopes instead of good search tools.
Portals were based on marketing ideas that now seem ridiculous: sticky sites (the users stay at the site) and eyeballs (the number of users.) So the portals added free email, personal webpages, games, and more, to keep the users at the site and then show banner ads at them. If you're over 28, you'll wonder: what's ridiculous about that? All the dotcoms were based on sticky eyeballs. If you're under 28, you'll say, so what? What good is 60 million visits per day if the site doesn't make any money?
And yes, I really did manage the website at a 16-million-clicks-per-day dotcom and all I got was this stupid T-shirt.
By 1998, Bill Gross, who started IdeaLabs and blew up several billion dollars along the way, came up with GoTo.com, a search tool where the listings were based on payment. Companies paid to be listed. I remember when this came out. We all felt this was somehow wrong. Okay, it's like the Yellow Pages and the more you pay, the bigger the ad. But still. Results in a search engine, based on payment? Yuck.
GoTo.com then became Overture and they offered the service to Google, but Page & Brin, a pair of purist monks, felt that commercialization of the ranking would result in poor links at the top. Big companies would buy the top spots, regardless of quality. And Google was all about delivering the best results. So GoTo was turned down.
In 2003, Yahoo bought Overture for $1.6 billion, and later, renamed it to something that nobody can remember, and worse yet, it's a name that not even Yahoo uses. At SES in San Jose, I went over to Yahoo's stand and asked if I could talk with the person in charge of their PPC. "Oh, you mean Overture."
In early 2000, a few months after rejecting GoTo, Google rolled out Adwords, which was based on Overture's idea. Google made a few changes: the paid listings would not be mixed with the search results. Thus the ads are at the side of the page. In early 2002, Google added a clever twist to Adwords: the ad's Click-Through-Rate (CTR) became a factor in the GAW ranking algorithm. Overture (or whatever it's called) doesn't use CTR; advertisers with large budgets simply power-buy the top positions. Smaller companies may be better, but they can't compete. Overture wins. The top companies win. The best companies and the consumers lose.
Battelle writes very little about Adwords or Adsense, which is a pity, since these are the bedrock upon which the $90B valuation is based. These tools will bring in around $5B in revenues for 2005, and that has doubled every year. Frankly, I don't see a limit anytime soon.
Google has many other tools, and many of these are significant. But the book doesn't say much about Blogger, Gmail, Google Maps, Google Earth, IM and VoIP chat and so on. The book doesn't talk about the deep changes in the search market, nor Microsoft's upcoming embedding of search in the OS. It'll be the Netscape Browser Wars all over again. The last two years had dramatic growth for Google; the next two years will be quite different. But that's not in the book.
I work with both Adwords and Adsense. Writing about PPC, Rick Bruner (Director of Research at double-click) says "Search advertising is at once starkly simple, bafflingly complex, and highly effective." Adwords appears to be simple, yet it is like playing chess on a 6-dimension board. The rules have implications that are far-reaching, both in the impact on the advertiser's own business process and the consumer's habits and actions. Adwords is radically different from anything on the market. I've also learned quite a bit about Adsense that I've never seen printed in any book or magazine, and many people would be quite interested in that. Investors would love to understand why Google works (and why it makes so much damned money) and its potential, but they won't find that out in the book. Battelle didn't learn much about these tools, asides from the usual (and fairly trivial) issues of clickfraud and trademarks. But this book isn't really about Google or search; it's point-of-view (POV) journalism on a popular topic. POV is good for journalism and porn, but it doesn't work with technical subjects. As I said, it's an outsider's book for outsiders.
Running a Company Called Google
There's a bit in the book about various issues at Google. At Stanford, Page and Brin had to scrounge for parts and they learned to be frugal. This has been a lucky habit; in the late 90s, Google's marketing director wanted to spend $5-10m on branding and marketing. Up to then, Google hadn't spent a penny on marketing. But the idea was killed and the marketing director was no longer at Google. If Google had spent the $10m on branding, it would have used up half of its $20m reserves and when the dotboom crashed in 2000, Google would have crashed as well.
Google is also cautious about hiring. The hiring-spiral phenomenon has wrecked many companies. To avoid competition and possibly losing your job to your employee, managers won't hire someone who is better than the manager. Each level hires lesser-quality persons in the lower level. An A-person hires a B-person, who then hires a C-person, who hires a D-person, and pretty soon, you have the telephone phone company. Page and Brin decided to prevent this by using round-robin interviewing. Candidates were interviewed by everyone and the decision was by group vote. This meant a grueling interview process.
By late 2001, Page & Brin felt the company was getting too bureaucratic, so in a moment of Khmer Rouge enthusiasm, they fired all of the managers. Who needs managers anyway? That reminds me of a New Yorker cartoon: two slackers are walking down the street and one says to the other "I hate my manager. He's always telling me what to do."
Battelle makes several errors. One that stands out; he writes "It's so common in Silicon Valley that it's got a name: entrepreneur's syndrome." It's very common in Silicon Valley, yes, and it has a name, yes, but that's not it. It's called "founder's syndrome". The founder of many dotcoms is often a bright engineer who can build a new product, but he is usually not capable of running a company. These founders try to stay in control of the company, and the board and investors usually have to fire them. We can all think of a few CEOs who got thrown out of their own companies, and we can also think of a few more who should be thrown out. Battelle didn't know the right name for this, but then again, he's an outsider. His reviewers didn't notice this either. Next time, get someone who knows the industry to review the book.
The book pretty much runs out of content about halfway through. The rest is just vague stuff.
The notes are pretty funny. When Battelle asked for interviews, Larry Page demanded the right to edit the book and add rebuttals in footnotes on every page. Battelle should have agreed; this would have been an amusing book with Page's running rebuttals. What's really funny is the conclusion to that note: Page felt that journalism was extremely flawed and he wanted to improve it. The book shows that Page was right.