Battelle, John “The Search – How Google and its rivals rewrote the rules of business and transformed our culture”, Nicholas Brealey Publishing, ISBN 1-85788-361-6, (2005)
Do you know?
• Vinod Khosla, the well-connected partner at Kleiner Perkins tried to persuade Excite (the company he had invested in) to acquire the technology and its creators’ services. Page set the price for Google at $ 1.6 million; Khosla was to persuade Excite to offer $ 750,000. Rest is history!
• Interestingly, in early 1998 Page submitted his first paper to the SIG (Special Interest Group) on Information Retrieval, but the paper was rejected! (It got published later in conjunction with Stanford digital libraries project)
• Alta Vista had 300,000 visitors on its first day Dec 15, 1995 and 4 billion queries in a year. For DEC Alta Vista really was a means to sell more hardware – an irony! For a short while corporate people saw the potential of Alta Vista. Compaq that bought DEC thought they were going to get richer by taking over Alta Vista. Ever the child of wayward parents – after a series of siphoning off and acquisitions Alta Vista waned away
• GoTo.com, the company founded by Bill Gross can quite legitimately claim to have created the business model that made Google possible.
• Wojcicki the marketing executive’s experience in early days sums up the travails of early days for any company, even the hugely successful Google. “We were rejected by a lot of ad firms at the time because nobody knew who we were. We say “we are from Google” and they’d look at our logo and ask “Is that a children’s cloth company?
• After watching a demo Andy Bechtolscheim, Sun co-founder and an outstanding designer said, “I don’t want to waste time; it will help if I just write a check”. Page and Brin pondered how much to ask for and told him a number. Andy said “I don’t think that is enough; it should be twice that much” and wrote a check for $100,000.
• Page selected Terry Winogard, a pioneer in human-computer interaction as his adviser. Page did not land on the idea of search at the outset; far from it. Despite Stanford alumni getting rich starting companies, Page found Web interesting for its mathematical characteristics. Each computer was a node; each link on a web page was an edge – a classic graph structure. “Computer scientists love graphs” according to Page
• On April 29, 2004 Google declared it would sell $ 2,718,281,828 worth of shares – mathematical equivalent of “e” – in effect declaring “the geeks are in control.”
• Google had to revise the price band from $ 108-135 to $ 85-108. The bad news would never stop it seemed.
• Friday Aug 13, 2004 had enough bad news to support that Friday, the thirteenth is an unlucky day. Al Qaeda had threatened to attack the opening ceremony of Athens, Olympics; yet, Larry & Sergey decided to go ahead with the process of auctioning the stock price. “How can these guys tempt fate so badly?” was the feeling
• In early 2001, Eric Schmidt (Google CEO) received a call from Larry Page; “the call lasted an hour. It was for a reference and from some kid? Just bizarre; I was trying to be helpful” recalls Eric Schmidt. This was followed by an interview during which “they criticized every single technical point I made. I was just floored. It was just really arrogant”. “Six months later I went back and checked; and everything they said was right! That is kind of humbling – beat by two 27-year-olds” recalls Schmidt
• Google runs over off-the-shelf parts – cheap hard drives, cheap memory chips and cheap CPU’s – instead of buying heavy artillery from IBM or Fujitsu. The beauty of the system was that it scaled – the more computers you threw at it, the more robust it became (Page 129)
• As far as the Internet ecosystem is concerned, Google is the weather
Does it interest you? If so read on
Google usage becoming part of everyday activity for millions, if not billions of human beings today, there is naturally a substantial interest in the company and its products from people belonging to all walks of life. It is easy to visualize Google as a search company; it is equally easy to mistake Google as a technology company, while in reality it is a media company; in fact, Google is a technology-led media company. The meteoric rise of Google also leads many to think of Google as “instant success”; thanks to Google’s powerful technology epitomized by its “Page Ranking” algorithm, Google could grow real fast; but in reality, Google’s early version started in 1996; till 2003 Google did not have a viable “business model”; Google’s introduction of “AdWords” and “AdSense” was the real turning point in its commercial success. Its IPO in 2004 had enough turbulence until the last moment; interestingly, Microsoft IPO in 1986 and Infosys IPO in 1993 were equally turbulent. Google founders Larry Page and Sergey Brin have in Erik Schmidt (earlier Sun CTO and Novell CEO) an unusual “driver” who could steer the company “safely”, though the founders do much of “backseat driving”.
The eleven chapters of the book
1. The database of intentions
2. Who, what, where, when and how
3. Search before Google
4. Google is born
5. A billion dollars – a nickel at a time
6. Google 2000-2004
7. The search economy
8. Search, privacy, government and evil
9. Google goes public
10. Google today & tomorrow, and,
11. Perfect search
capture this fascinating story exceptionally well. I practically finished the book in one sitting over the weekend (it was so compelling). I decided to review this book in an unusual way – string together several key sentences of the book with some modifications here and there to make it readable! I do hope the readers enjoy this unusual rendering!
Search has been the focus of interest for decades; Library Science professionals and AI professionals have started work on search much earlier.
Melvil Dewey, credited as the father of modern library, in the late nineteenth century introduced “Dewey decimal system” – a universal classification scheme; though updated many times over the years, Dewey system would be unable to scale to the enormousness of today’s Web (Page 33)
The work of AI pioneer Doug Lenat “Cyc” attempted to conquer AI’s brittleness by coding in hundreds of thousands of commonsense rules and building in models, based on these rules. Cyc alumnus Srinija Srinivasan had run directory-based search at Yahoo from day one. But such brute force method has failed; successful search will see clever applications of algorithm; if any company can claim to have created an intelligent search engine it is Google (page 17)
Cornell mathematician Gerard Salton created the first digital search engine through his SMART system – Salton Magical Automatic Retriever of Text – and pioneered search through his “Text Retrieval conference” over eighties and nineties. (Page 33)
In the Internet world, search has very interesting patterns too
The long tail of search queries has the patterns that average frequency for ranks 1-100 on Google, “trumps” the frequency for ranks 11-110 by thousand fold! The tail is extraordinarily long and the power of search lies in that tail; no matter what the word is, somewhere on the web there is most likely a result that contains it (Page 27-28)
There were many attempts that could have eclipsed Google’s work; none had the “luck” that favored Google founders Larry Page and Sergey Brin. Louis Monier the creator of Alta Vista was a pioneer; his Alta Vista had everything it takes to win, except that DEC management could not think beyond hardware and Alta Vista was sold and re-sold many times leading to a situation of a genius child born and reborn to “wayward parents”!
“I’ve always been interested in big, nasty problems, Search provided one of the nastiest” said Monier (Page 45).
Solving for this scale required multiple crawlers that worked in parallel, building the Web index. Thanks to DEC Alpha’s 64-bit capability, Monier could index the young Web (in 1994) – 10 million documents with billions of words; he even hacked up an interface in two months and everyone who used it loved it, but not DEC management – “what good was a search engine when it came to selling hardware?” (Page 46)
Alta Vista had 300,000 visitors on its first day Dec 15, 1995 and 4 billion queries in a year. For DEC Alta Vista really was a means to sell more hardware – an irony! (Page 50)
For a short while corporate people saw the potential of Alta Vista. Compaq that bought DEC thought they were going to get richer by taking over Alta Vista (page 51).
Ever the child of wayward parents – after a series of siphoning off and acquisitions Alta Vista waned away (page 53)
Yet another visionary who had some luck, but not as much as Google was Bill Gross.
GoTo.com, the company founded by Bill Gross can quite legitimately claim to have created the business model that made Google possible.
In 1985, Gross was working on a major piece of the search problem – a natural language text interface – GNPDevelopment; it worked with Lotus 1-2-3, a limited functionality software tool.
Gross next addressed the “indexing” problem thru Megallen as a way to “search all your files on your hard disk instantly”.
Gross’s core insight, the one that drives the entire search economy, is that the search term, as typed into a search-box by an Internet user, is inherently valuable – it can be “priced”. Search economy found the right business model (Page 106)
Gross met with Page and Brin to suggest the two companies merge, but Page and Brin turned a cold shoulder to Gross’s overture – the reason, Google would never be associated with a company that mixed paid advertising with organic results. Interestingly, several months later, Google introduced AdWords, Google’s answer to Overture! (Page 110-121)
GoTo.com founded in late 1997 was finally sold as Overture for well over a billion dollars, though it could never anticipate Google (Page 102-105)
Stanford University has a knack of incubating Google-like companies.
Page selected Terry Winogard, a pioneer in human-computer interaction as his adviser. Page did not land on the idea of search at the outset; far from it. Despite Stanford alumni getting rich starting companies, Page found Web interesting for its mathematical characteristics. Each computer was a node; each link on a web page was an edge – a classic graph structure. “Computer scientists love graphs” according to Page (Page 68)
At one point BackRub crawler (an early implementation before Google was born) consumed nearly half of Stanford’s entire network bandwidth, an extraordinary fact considering that Stanford was one of the best networked institutions on the planet (Page 78)
Steve Hansen, the computer security officer at Stanford University (now at Google) sent the following mail in February 1997 “I have received numerous complaints regarding excessive and / or unauthorized access. If research is to be done out on the Internet it must be done with much more care and supervision that has been evident with the “BackRub” project (Page 79)
Google was not accepted by all users easily
Many were not pleased with the upstart search engine’s seemingly blind judgment regarding their site. After all, this was the first time anyone had claimed to rank the inherent value of a web site. Page and Brin had clearly hit a nerve with every person who labored over a web site. To many, unleashing a ranking system based on a blood less algorithm felt like a supreme act of arrogance – who are these kids from Stanford, telling the world how we ranked? (Page 80)
Contrary to what many feel it was not hunky-dory all along; Google had rather humble and sometimes humbling early days
Press coverage of Google often glosses over the fact, but the truth is that the company lacked a viable plan for making money until early 2001. “There was genuine concern at the board level” says Ram Shriram, a board member.
Interestingly, in early 1998 Page submitted his first paper to the SIG (Special Interest Group) on Information Retrieval, but the paper was rejected! (It got published later in Conjunction with Stanford digital libraries project) (Page 82)
Near the end of 1999, with more than $500,000 going out every month and less than $20 million in the bank you didn’t need a Stanford PhD to do the math; the company needed a business model that worked (Page 123)
Vinod Khosla, the well-connected partner at Kleiner Perkins tried to persuade Excite (the company he had invested in) to acquire the technology and its creators’ services. Page set the price for Google at $ 1.6 million; Khosla was to persuade Excite to offer $ 750,000. Rest is history! (Page 83)
Rejected but not deterred, Brin and page went back to Stanford and kept working on Google – Stanford.edu (page 84)
Wojcicki, the landlord, was hired for marketing. Her experience in early days sums up the travails of early days for any company, even the hugely successful Google. “We were rejected by a lot of ad firms at the time because nobody knew who we were. We say “we are from Google” and they’d look at our logo and ask “Is that a children’s cloth company? (Page 127)
Tough decisions – “our competitors had huge marketing budgets” but “marketing could have killed the company; we only had twenty million; spending ten million would have out us into half”. Ram Shriram recalls “we were the only company not spending money on marketing; were we the dumbest people in the business? (Page 128)
Google founders did not like the advertising model of DoubleClick, but it was the most successful in 1999. Google thought that DoubleClick was the ocean liner Google swim to, should life saver fail! But NASDAQ crash of 2000 changed everything! DoubleClick stock had plummeted from $ 100 to $ 15. (Page 124)
Had Google not adopted some of the advertising techniques, it would have ended up a small, but nice, high-end company. (Page 125)
After watching a demo Andy Bechtolscheim, Sun co-founder and an outstanding designer said, “I don’t want to waste time; it will help if I just write a check”. Page and Brin pondered how much to ask for and told him a number. Andy said “I don’t think that is enough; it should be twice that much” and wrote a check for $100,000. To celebrate Brin and Page went to Berger King and had breakfast.
Wojcicki rented a spare room in her room in Menlo Park; concerned about her privacy – she was pregnant at the time – insisted that her tenants enter through the garage door; Google got the most shopworn cliché’ in the Silicon Valley – a garage address! (Page 86)
Sequoia and Kleiner, two of the most influential financiers in the valley invested $25 million in early 1999; Google had arrived (Page 90)
The investor’s insisted that the company quickly recruit a new CEO, much as Tim Koogle had replaced Jerry Yang as CEO at Yahoo. Page and Brin chafed at the idea of being told what to do by their new board members (Page 91)
In May 2002 Google announced a deal with AOL with guarantee of tens of millions of dollars of advertisement revenues; if AdWords had a recession, Google’s debt to AOL could have forced the young company out of business (page 145)
The business model evolved over time
Brin and Page were deeply suspicious of blending advertisement and search. Over time, the founders have clearly made peace with their reservations about advertising.
Google essentially copied GoTo.com’s approach but made a part of separating advertisement results from natural search results. In October 2000, Google introduced AdWords. The service was a hit, revenues began flowing in and the mood in Google campus improved (page 126)
Year 2001 saw Google growing big and seek “content” like buying out Usenet with 500 million discrete postings; it was consistent with Google mission statement that evolved later
“To organize the world’s information and make it universally accessible and useful”. (Page 140)
Google incorporated an ad’s popularity – its click-through rate – into its overall ranking. If there were advertisers in accounting area, and Accountant One is willing to pay $ 1.00 per click, Accountant Two $ 1.25 per click and Accountant Three $ 1.50 per click, Accountant Three would be listed first by all search engines. But in Google’s scheme, it is only till Google had time to monitor click-through rates. If Accountant One who paid less was drawing more click-through than Accountant Three, Accountant One would graduate to the top slot, despite his lower bid. (Page 142)
Concurrent with Google News launch, Google re-designed its home page, adopting the familiar tabbed design and by 2002, Google as we know it had taken shape (Page 144)
Google enabled lots of small business to get visibility; for example 2bigfeet.com a small store in Albany, Georgia could 95% sales coming through search engine. However on November 14, 2003 when Google changed its ranking algorithm suddenly, 2bigfeet.com found its order book drying up. Google became a monster that can “make” or “break” fortunes real fast for many small businesses. After all, if you are not in the first two pages of Google screen, your just do not exist! Google did the tuning to shut out spammers but it worked out very differently to end customers! Interestingly after a few months 2big feel.com came back with top ratings! Google giveth, takes and gives back too! (Page 155-157)
By 2003, Google had amassed more than 100,000 advertisers using AdWords, yet its investment in customer service was minimal. Launched in March 2003, Google AdSense program allowed third–party publishers to access Google’s massive network of advertisers; it was instant hit, thousands signed up. It was almost magic; add a few lines of code to your site and in a month or so checks from Google started showing up in the mail! Next to Google’s core infrastructure, AdSense created the second best asset – network of advertisers. (Page 151-152)
Google has upset many other businesses too “Search has ruined the real-estate business” laments a California-based real estate developer. Search has become a new interface of e-commerce (Page 179-180)
As far as the Internet ecosystem is concerned, Google is the weather (Page 183)
The magic of intent-based marketing is it shifts marketing dollars from the unknown to the knowable. Search turns a cost center into a profit center (Page 111)
The core foundation of marketing – $ 100 billion industry driving every business on this planet – is shifting to a new model – shopping will basically become an application of search (page 172)
One of the well kept secrets is Google’s formidable infrastructure and the way it was built over the years
Google garnered impressive word of mouth among its users for one thing – it worked! The service never showed signs of buckling under exponential growth it was experiencing.
Google runs over off-the-shelf parts – cheap hard drives, cheap memory chips and cheap CPU’s – instead of buying heavy artillery from IBM or Fujitsu. The beauty of the system was that it scaled – the more computers you threw at it, the more robust it became (Page 129)
Because of its prodigious and scalable architecture in the first days after 9/11, Google realized that it had an extraordinary asset – the ability to cache any information and show it to anyone on demand. Google has one of the largest data centers in the world and one of the largest collections of band width in the world” (Page 144)
Google is different in many, many ways
Larry Page always wanted to be an investor. But his role model was clear; Thomas Edison and not Nikola Tesla. Page said “I wanted to invent things, but I also wanted to change the world. I wanted to get things into people’s hands because that is what really matters”. (Page 66)
Page was concerned that someone might steel his idea. Page respected and participated in the academic tradition of sharing research through published papers; he was also influenced by the more closed approach of a corporation protecting its intellectual property.
The inventors, Larry and Sergey, faced a classic entrepreneurial dilemma; if they started a company, it could be crushed by larger, richer competitors; on the other hand, if the company took off, the upside would be huge. They took a more conservative route of licensing technology to a major player. (Page 83)
Google was very careful in hiring early employees. “I interviewed every single candidate for the job” recalls Ram Shriram, the early Director. By the end of 2000 it had grown to nearly 150. To say the founders were obsessed about who might join the company was an understatement! (Page 130-131)
In early 2001, Eric Schmidt (Google CEO) received a call from Larry Page; “the call lasted an hour. It was for a reference and from some kid? Just bizarre; I was trying to be helpful” recalls Eric Schmidt. This was followed by an interview during which “they criticized every single technical point I made. I was just floored. It was just really arrogant”. “Six months later I went back and checked; and everything they said was right! That is kind of humbling – beat by two 27-year-olds” recalls Schmidt (Page 132-135)
When finally Schmidt joined Google as CEO, reactions were sharp. ”What the hell in Schmidt thinking? Doesn’t he know that the Internet is over?” Yet, Schmidt’s timing could not have been better; Google had its first profit in summer 2001 and has never had a down quarter. Either Schmidt was a genius or he was very lucky (page 137)
In September 2001, Brin and Page gathered all the engineering managers together and informed them they were out of a job; things would be done differently at Google (Page 141)
“Do no evil” philosophy
Paul Bucheit, a Google engineer came with “Don’t be Evil” on July 19, 2001. That became the most important three words in Google history (Page 138).
Not many were impressed; some thought it was downright arrogant. Jeff Bezos, Amazon founder summed it up “you should not be evil; you shouldn’t have to brag it either” (Page 139)
Google’s foray into china is a first test of “Do no evil” philosophy (Page 205) Google did purge the offending sites from its index. Why did Google blink? (Page 206)
As 2004 dawned Google had become the talk not only of Silicon Valley, but Wall Street as well. 2003 revenue was nearly $ 1 billion and profits around $ 300 million.
Brin and Page struggled with the idea of becoming public. The founders were worried that the company would be forced to a mind-set of short-term thinking (Page 214).
On April 29, 2004 Google declared it would sell $ 2,718,281,828 worth of shares – mathematical equivalent of “e” – in effect declaring “the geeks are in control.” (Page 216)
S1 (the public offering document) starts thus “Google is not a conventional company. We do not intend to become one” (Page 217)
While Page and Brin would jointly hold just 30% of the actual shares of the company, they nevertheless have control over every major decision; because their shares hold ten times the voting power of public shares under “dual class” shareholding structure (common among media companies); and they would adopt an “auction scheme” for shares allotment (Page 218-19)
Google’s penchant for secrecy reached paranoid levels; “I came out feeling that I had visited a fascist state” wrote a valley entrepreneur. Eric Schmidt was seen walking around wearing a T-shirt that read QUIET PERIOD on the front and CANT ANSWER QUESTIONS on the back! (Page 222)
“Why not to bid on Google IPO” said San Jose Mercury News
“Google IPO, asking too much?” asked Business Week (Page 223)
By late July 2004, Google announced that it will list on NASDAQ, it will have Morgan Stanley and credit Suisse First Boston as lead banks and price band would be $ 108 -$135; unusual as most IPO price is in $10-20 range. Its offering suffered numerous setbacks and delays.
Friday Aug 13, 2004 had enough bad news to support that Friday, the thirteenth is an unlucky day. Al Qaeda had threatened to attack the opening ceremony of Athens, Olympics; yet, Larry & Sergey decided to go ahead with the process of auctioning the stock price. “How can these guys tempt fate so badly?” was the feeling
Finally on August 19, 2004, Larry rang the bell on NASDAQ floor.
What happened next, put to rest nearly every doubt? By the end of the day the stock had climbed to $ 100 and by November reached $ 200.
Eric Schmidt summarize “we were on a turbulent flight; as long as we got all the passengers off the airplane and we are safe, I am happy” (Pages 222 to 228)
Google faces its own challenges too.
Google still remains an extraordinary experiment in bottoms up innovation. It will be challenged in its ability to continue to innovate in a focused, market-driven innovation (Page 245)
By mid 2002, tens of thousands of young technology workers were unemployed and no one save Google was hiring. While the rest of the Valley languished, Google prospered. A backlash began to grow built on envy and jealousy.
The case of “American Blinds” can cause enough damage. While it may not sink Google, one should not forget Microsoft trial that has ensured that Microsoft stock price has not risen since the case was filed five years ago (Page 186).
One should not forget “click frauds” (Page 187)
Google faces the problems of privacy invasion; you type in your phone number and your address and map of your house pops up, thanks to “reverse directory” search; this kind of information was always public, but not widely available. (Page 190)
Regardless of your desire to know your new co-worker has a messy divorce, most of us will not waste an afternoon to find it out. The very fact it is so much trouble to find such information has muted that information. But what if it were as easy as typing his name with Google (Page 191).
And at its very heart, privacy is all about trust; you no longer control your private documents and communications and you are in a relationship of trust with your service provider (Page 197)
The founders holed up for an all-night writing session and emerged with what became known as the Tablets – a declaration of what makes Google (Page 231)
Google post-IPO organized into core search, advertising products and “20 percent” (engineers pursue interest beyond their workload) and “10 percent” (really wild ideas) (Page 232)
Criticism galore – “Google is a monarchy with two kings, Larry and Sergey”, wrote Brian Reid, Valley veteran (Page 232)
“Eric is a puppet; Larry and Sergey one whimsical people” (Page 232)
The trouble with Google is that the company rocketed from being unknown to having the status of Apple and Microsoft in five short years (Page 234)
Yahoo & Google
Yahoo and Google are so similar and so different. Both founded by two Stanford PhD candidates (EECS); had silly names, that caught on; funded by top tier VC; wildly successful IPO; multi billion dollar market cap. Yahoo founders are “nice” guys unlike Google founders, Yahoo CEO “runs” the company while Google CEO is a “puppet’; Yahoo got chastened by stock price dropping from $ 500 to less than $ 10. The only thing Google has failed to do is to fail! (Page 236)
At Yahoo it is humans backed by technology, while at Google humans enter only when algorithms fail (Page 240)
The future of search
Forget about computer on every desk (Bill Gates words). Instead the entire world needs to be computerized (Page 249)
As every engineer in the search field loves to tell you, search is at best 5% solved (Page 252)
No one can or should archive the sum total of humankinds’ information. When it comes to making the world searchable, let the world do it (Page 255)
Search scholar Ramesh Jain calls for a “search steering wheel” – a control mechanism for browsing through search (Page 259)
Search has been stuck in “C-prompt” phase, according to Tim Bray of Sun (Page 261)
IBM Web Fountain project has a lot of promise (Page 268-270)
Global-spec’s human editors identified 100,000 specific sites (Page 275)
Brewster Kahle of Internet Archive says “The lesson of the first library of Alexandria is – don’t have just one copy” – sums up the need for distributed computing (Page 278)