Google and the Mission to Map Meaning and Mine Money:
It was 3.15 am Greenwich mean time, hour of the wolf, 30th April 2004 when an email newsfeed finally appeared on Google's Initial Public Offering (IPO), which would take the company into public sale on Wall Street. The Financial Times had posted up links to the IPO filing to the SEC (Security and Exchange Commission) and Larry Page's users' manual . Sixty pages later, as dawn broke over London, it was clear that they had done their maths, and figured the engineering. They were safe and on their way.
At about the same time, the spacecraft Cassini arrived at Saturn , having used the gravity of Venus and Jupiter to power its slingshot trajectory since its launch in 1997. Google's pilots: founders Larry Page and Sergey Brin and Chief Executive Officer (CEO) Eric Schmidt, are undertaking a similar manoeuvre in cyberspace; instead of being sucked into what sf writer William Gibson has called "a black hole... the unthinkable gravitic tug of Big Money" they are using the force of their multi-billion dollar IPO to power the next stage of their jaw-dropping journey into hypertext, context and meaning.
Although Cassini's flight had looked almost flawless, in 2004 fingers were still crossed for the success of its European Space Agency probe Huygens, particularly after the Genesis probe crashed into the Utah desert at 193 mph in September, because the parachute switches were installed backwards. When they discovered that Huygens was using a different radio frequency from Cassini, NASA remedied it by remote engineering over a billion miles, but might not have wanted advice from the world's press at the time.
Google were not so fortunate in their own transit of Venus. When they failed to schedule the publication of an Interview with Playboy outside the SEC-designated 'quiet period', they found, as in the Irish joke, that the eyes of the Skibbereen Eagle and every other news channel on the planet were upon them.
They should have enforced a later publishing date, but the Playboy interview is an American rite-of-passage, a mainstream statement that shows the jocks from high school who really calls the shots. More seriously their July price estimate and volume had to be diverted sharply downward for the August IPO, which jumped 18% on the first day of trading, finally giving Wall Street the bonus it craved, unearned unless you count the work done on the barrage of negative publicity, fuelled by Google's own exaggerated errors, which helped scare off retail investors.
But it is probably fair to say that the Black Hole of an IPO did little more than rattle Google's insulation tiles, the real journey will now follow. After the first day of trading the Wall Street Journal's headline was "Google's Debut is Considered a Success". (Their second headline was: "Google's Wealth Could Bring Woes" - gee, thanks, guys.) The FT's Lex gave what the Trotskyists used to call 'Critical support in struggle':
Wall Street will see the auction as a failure. Execution was poor. Strong vested interests and tough markets made life even harder. However, Google did get listed and its unconventionality shone light on important areas.
James Surowiecki of the New Yorker was even more forthright in an FT Comment article entitled Ignore Wall St's whining - Google's IPO worked:
Wall Street can spin this however it wants. But Google went public without the marketing support of a major investment bank, without handing out favours to well-connected executives and without dictating a price in the manner of Soviet central planners. Because it did, it now has hundreds of millions of dollars that it would not otherwise have had. By any standard, this was one IPO that worked.
Money loves only money, and if Google ever runs out of cash, expect the banks to be gleeful to the point of sadism as they dispatch the management and, almost certainly, break up the company. But, same difference really, even if Google had, like previous dot.coms permitted Wall street to make obscene profits out of its IPO, it would still get the same treatment if it ever mismanages its income flow.
The founders' founding ideals of free, unbiased, fast, public Internet Search would then be repeated publicly, before being ignored behind closed corporate doors, because the short-term profits from Search bias are so big, particularly if, like Google, you have a huge brand based on honesty. That remains a threat.
Yahoo! survived as a company following its IPO, true, but only after not only its founders had been replaced, but the "professional" management demanded by Wall Street had themselves been purged. That's not what Page and Brin intend to let happen, if they can fight to prevent it. So far so good.
This is partly a battle for America's soul; whether the country returns to the hugely ambitious, ingenious engineering, which has been its hallmark, or follows the Enron route to making money by gambling with its own, and other people's, primary assets. Having survived the dot.con boom, Google's management share with the whole IT industry the knowledge that financial speculation in tech shares has done nothing but damage to anybody except the tiny minority of individuals and companies who won the sweepstake of other peoples' investment.
This IPO was in no way an attack on capitalism, but on financial speculation and short-term management objectives which prevent the growth of companies centred, like Google, on products which people want, and want enough to pay for, one way or another, over the long term.
It now looks like Google's management have pulled it off. This is partly because the quality of their products speaks for itself, even if the Google Press Office is on permanent sabbatical, but also because they have tapped into the mother lode, the equivalent of Gould and Curry's Comstock seam in Mark Twain's Nevada.
Google's success depends on many things, but part of it is a critical development of the concept of context. Context is at the heart of Stanford's work on IT theory under the leadership of Professor John McCarthy, as it has been at the centre of the late Jacques Derrida's philosophy.
John Hegarty, chairman of Bartle Bogle Hegarty, uses the concept of "Truth to Product" which suggests that the most striking advertising campaign in the world will turn out to be a total waste of budget if it does not tell an essential truth about the product as it is perceived by the consumer. He argues that context is central to the development of a brand.
Microsoft, rightly or wrongly, is perceived by many to be the evil empire and is therefore attacked by mindless viruses that distract the brand from its global domination. It is within this emotional context of domination that the brand must respond and organise its defence. Context even defines the framework of a virtual product.
It's obvious really, but everything about us relates to context. Where we were born, where we live, how we talk. Our ambitions, desires and fears all relate back to context in some shape or form. It is embedded in the very DNA of our thought process. It can't be jettisoned simply because we can so easily communicate across borders.
Google developed because its founders had found a new and extremely powerful way of cracking the maths of context on the Internet. The operation of a free global service was ultimately paid for, and then some, by using the same core technology to provide a context for advertisers.
Contextual advertising gives a strong material incentive for vast amounts of money to get diverted away from print advertising and onto Internet advertising. This is what will make Google, and some of its competitors, into the flagship industries of the 21st century.
As long as Google keep tapping this seam, the banks cannot touch them. If Google is safe it is because its IPO made no concessions to short-term profit or speculation. If you wanted to back the strategic vision of the founders and their CEO, you could have bid for shares, otherwise you could naff off.
Company structure is a Delaware Delight: the dual formation allowed by the State of Delaware means that the founders and their CEO with B-type shares have ten times the voting power of other shareholders. They have almost complete control and they intend to keep almost complete control, for the foreseeable future, partly by trying to spread the sale of "A" shares as widely as possible through a Dutch auction. And they are paying no dividend.
Google will preserve its long-term goals by effectively remaining a privately-run company, even though its shares are publicly traded. This was spelt out by Larry in the SEC (Security and Exchange Commission) filing.
Since Larry uses first names for himself, Sergey Brin and Eric Schmidt, Google's CEO, in their IPO filing, at the risk of being over-familiar, this book will do the same. This is also the moment to apologize if through laziness I have sometimes attributed too much to Sergey and Larry, and not given enough credit to the perspiration and inspiration of the rest of the Google team. Although, like Microsoft, the founders' leadership remains critical to what Google is and will become, that future depends ultimately on the quality of work they inspire from the collective effort of the whole company.
"If opportunities arise that might cause us to sacrifice short term results but are in the long term interest of our shareholders, we will take those opportunities." Google said. "...As a private company, we have concentrated on the long term. This has served us well. As a public company, we will do the same."
The core message of the SEC filing is that Google's IPO is solely a long-term investment. There will be no quarterly massage of their figures and there are a dozen pages of good reasons why your initial purchase could decline in value in the short term and is subject to significant risk in the long term.
Over time, Google's founders have learnt how to rewrite their operating system from scratch until it works the way they want it to. It was natural for them to do the same thing with their IPO. Back on planet Earth this caused consternation, and the SEC's server was reported to have crashed with the demand to see Google's documentation.
The FT reported it fully, explained industry concerns and then, via its Lex column, came down firmly in favour. They must have been relieved that their world scoop in the Fall of 2003 predicting Google's IPO in April 2004 had come true, albeit on the last possible day. Two days later, after studying Larry's exhaustive summation of all the things that could possibly go wrong (...earthquakes in California and Japan? - check) they were a little less confident.
Meanwhile the Business and Media section of the Sunday Observer in London became quite schizoid. A page one splash complained that UK residents were being discriminated against by not being allowed to buy shares. On page two readers were advised to:
"Avoid (the IPO auction) like the plague" (because it is) "a throwback to the inglorious days of the dot.com bubble" (which) "could end up massively over-pricing the shares or perversely, selling them at a huge discount" with a "mind-blowingly hip prospectus - which could have been written by the re-incarnation of Mother Theresa".
Shares in US IPOs are rarely sold outside North America, because they are primarily subject to state not federal law. So Observer readers faced the Groucho-like paradox of being refused shares in a bubble company run by Mother Theresa in iPod earbuds and bitch jeans. The Observer's usual well-informed sanity had returned by page 3 for a full page analysis of the IPO, culminating in veteran IT-watcher John Naughton taking an engineer's pleasure in their final joke .
"The 'proposed maximum aggregate offering price' is $2,718,281,828...It's e - the base of natural logarithm. With the possible exception of pi, e is the most important constant in mathematics since it appears in every calculation involving limits and derivatives."
Another wag suggested this meant the correct bidding price for Google shares was either $2.71 or $27.18. The more serious implication is that while Google can describe just about all the conceivable future threats to the company, they can only guess its current and ultimate value.
But then, nobody else knows any better. Revenue over three and a bit years of profitable trading in a new product is not going to predict values in another three, or thirty, years time. But Google already had another cunning plan in place: better maths for a healthier future.
At the heart of Google's technology is the experience that "democracy on the web works" because it creates part of a measurable context for web pages. By requiring blind bids, they did their best to create a similar grand consensus of what potential investors and their financial advisors really believe the company's finances and brand identity are worth.
Although take up by retail investors was disappointing, the auction did appear to work, because Google's core indexing technology was able to process the results properly, which is rare. In July 2004, Google opened a site to allocate bidders' IDs and suggested that they expect to sell their stock in mid-August for between $108 and $135 a share, valuing the whole company ultimately at up to $36 billion. Shortly before the IPO this dropped to $85 to $95, and on the first day of trading the shares closed at $100.34. Glitches aside, it is all good. Now they have to create the products that justify their $23.7 billion market valuation, 30% less than Yahoo!. Yahoo! is a huge success story, again, but may still be significantly over-valued in 2004.
If the company was broken up tomorrow, that sum might look ludicrous, but if they stick to their knitting and maintain their priorities, in ten years it could look quite modest. Twenty years ago nobody came close to guessing that Microsoft's revenue in 2003 is "expected to be in the range of $8.9 billion and $9.0 billion" with operating income "in the range of $2.8 billion and $2.9 billion". This is not to imply that Google is 'the next Microsoft'; far from it, the creation of both companies was determined by variables that are now both historic and unique.
To be fair, the Observer's completely contradictory response to Google's limits and derivatives, as manifested by an IPO through 'Dutch auction', exactly reflected what was going on in the global news media. Internet veterans were delighted by another enthralling chapter in the Google saga; bankers hoping for short-term gains (and devil take the pension funds) were appalled. Insider trading was made virtually impossible, banking behemoth Goldman Sachs was shown the door after 'playing politics' by arranging a separate meeting with one of the original investors; the banks were being paid a miserly 3% for their services, a fraction of what they had expected. A number subsequently dropped out, complaining of too much work for too little money: clearly a shocking reversal of the proper order of things. However, the departure of Merrill Lynch, with the biggest retail client list, may have hit Google's retail sales volumes.
Blaspheming against mammon makes powerful enemies, but Google, rather than striking out on their own with a mixture of innocence and arrogance as usually described, are actually following a path they have learnt from the real veterans at Intel, Microsoft, News International and Berkshire Hathaway. Although the founders are world-class developers, not MBAs, they are great learners and their close relationship with Eric Schmidt is based on taking his business experience with Sun and Novell, where he was CEO before moving to Google, very seriously indeed.
News International's parent company, News Corp is moving from Australia in mid-2004 to incorporation in Delaware, for similar reasons to Google, upsetting the Australian stock exchange in the process. Split voting rights for shares is not unusual, but the tiny state of Delaware has a legal framework that is recognised to be amongst the fairest, fastest and most intelligent, almost anywhere in the world. A recent example, reported by Geoffrey Colvin of Fortune suggests: "A Delaware ruling could make directors personally liable for pay suits" as a response to the revelations of the kind of perks that Kozlowski enjoyed at Tyco.
A dual ownership structure lends itself to long-term focus but is clearly also open to corruption and abuse. The Australian version enabled Rupert Murdoch to maintain his corporate and dynastic strategy when News International's resulting debts made his bankers change their pants. Similar Canadian law enabled 'Lord' Conrad Black to siphon off large sums from the company which owned the Chicago Sun, the UK's Daily Telegraph and the Jersualem Post groups, according to depositions by the A-group shareholders. When Black subsequently attempted to sell some of these titles, a Delaware court under Vice-Chancellor Leo Strine forced him to hand them back.
Dutch auction apart, Google's IPO strategy imitates closely that of Microsoft, twenty years ago. Their SEC filing makes it clear that they have listened, and listened good, to Warren Buffet and learned a great deal from Microsoft's strategic history since its IPO. Both share the conviction that the long term value of the company is everything; nobody and nothing is going to interfere with that. Speculation is death.
Like Gates and Ballmer, who led an IPO which was "a Bataan death march" according to their bankers, Google's IPO attempts to prevent litigation and depress speculation by taking the worst possible view of the company's prospects. Sergey even adopted a central Billism, that an IPO is an inevitable but unwanted "distraction" from the long-term goals of the company. This is the exact, dialectical opposite of what the dot.coms did, where hype was transmuted and spun, albeit briefly,into fairy gold.
By also using a blind auction it disperses ownership as widely as possible, making a future hostile take-over much more difficult. Yahoo! already had 2.7 million shares of Class A stock as part of its settlement of the Overture suit on keyword auctions, and sold 2.3 million for $191 million. That was a mistake; if they had hung onto them for a month they would have made 40% more. AOL finally did something right by bringing 1.9 million shares for $22 million to parent Time Warner (at about $12 a share).
Sergey had even said that assimilation might be Google's ultimate destiny, resistance may be futile, but people said he was depressed by the problems of the forthcoming IPO. Takeover is a problem that Microsoft have never had to face.
Rather than simply brooding, Sergey was probably gestating ideas of the best way of coping with the intense local pressure within Google and the SEC rules which had made an IPO inevitable, when that IPO would be more than capable of destroying the company, and statistically was likely to do so. Larry has mentioned a similar mood in Y2k when Sergey was pondering on Google being a highly successful operation which was fashionable enough also to be losing money, and what to do about it. The answer, contextual text advertising, is now worth billions and is definitely responsible for [95]% of Google's current income, well worth a dark mood.
Perhaps the biggest small miracle of all is that the business friendship of CEO Eric Schmidt, Larry Page and Sergey Brin's has been declared to the SEC as a significant asset (another first) despite the G-forces that the IPO must have subjected their relationship to. Silicon Valley is paved with the bones of founders and CEOs who have fallen out with each other. (Steve Wozniak, co-founder of Apple, did not speak directly to Steve Jobs for years after a misunderstanding about some money).
This may be because Larry Page shares the philosophy of Intel, the Dutch, and his late father, Professor Carl Page, that violent controversy is the negative feedback on which friendship, business and science can flourish. When they first interviewed Eric for the job of CEO the founders started arguing with him and have been doing so ever since. Once they agree, if the maths runs and the engineering model actually works, then it flies like Cassini, as they have demonstrated.
If the prospect of an IPO made Sergey depressed, some of Google's oldest friends - people who had been watching the company professionally since it first exploded out of Stanford, were almost equally worried. They felt that if Google became overvalued in a wildly speculative IPO then it would go the same way as Netscape.
We knew that Google, unlike Netscape, have been making huge profits out of contextual advertising: the AdWord/AdSense programs. But if this was a temporary, windfall profit which the likes of Yahoo! and Microsoft would be quick to reclaim, then expectations of growth and profits would be a mirage, and the company itself would do a dot.com. A high valuation of Google means that it has to be capable of producing products in [7] years' time that people will still want to pay for. While Netscape never did get that, Microsoft never forgets it.
Google's valuation problem is that it does not make better washing machines. Like the water supply in Chinatown, its business is perceived as "The future Mr Gittes. The future." In the short term this makes over-valuation of the company almost inevitable as everybody wants a piece of the future. A dot.com-style over-valuation is what concerns Bambi Francisco , of CBS marketwatch. At a valuation of $35 billion, Yahoo!'s current worth, Google:
... would be closing in on Walt Disney's $47 billion valuation and would stand at more than half of Viacom's $67.4 billion value.
If history repeats itself, maybe Google will buy Viacom, Yahoo! will buy Disney and some starry-eyed analysts will begin slapping $400 price targets on both Google's post-IPO shares and Yahoo's.
Yep. We've seen it before. Remember when Amazon.com got the infamous $400 price target back in 1998, or when Yahoo! received a $600 price target back in January 2000?
Or, recall when AOL scooped up Time Warner just when AOL hit its peak valuation? ...Well, we all know what happened after that. Most of us went on sabbaticals, and some hid deep inside caves.
If the founders are ever pictured in the tabloids as having developed a taste for yachts, bimbos and cocaine, then dot.com investment idiocy is probable, but on their current track record? No chance.
All their acquisitions to date have been strategic, extremely cautious, and, apart from Google News Alert, all have become profitable. Much-envied employee benefits aside, they have never wasted a penny and show no sign of ever losing track of what generates the bottom line.
There are good reasons in their backgrounds why this is likely to continue; both founders are only a generation away from hardship. Sergey was born in Moscow and Larry's father, the late Professor Carl Page was the first person in his family to graduate from high school. The only sad note in this story is that Professor Page should have died in 1996 before Larry and Sergey published their breakthrough paper on hypertext search architecture. But by 1996, Larry's brother, Carl, was already on his way to becoming a multi-millionaire by selling his eGroups software to Yahoo!. Over-achievers? You betcha.
The real concern of experienced observers was not that the founders and the CEO would start developing delusions of corporate grandeur and were about to try and buy Sony or the Danish Royal Family, but that their market-share could prove temporary, and their ability to generate a range of profitable products over time was unproven.
Fred Vogelstein of Fortune he was concerned that the company culture was too spontaneous and chaotic to cope with organised competition, but even more worrying, Google's service was impersonal. It appeared to have no permanent user base.
Although the Internet Protocol (IP) number of every searcher, which (temporarily) identifies any machine connected to the Internet, is buried deep in Google's archive, there was nothing more to connect the user with the company. That would mean that if Yahoo! or Microsoft developed a better search technology, as they promised, users could switch from Google without penalty, just as they had switched from Netscape as soon as Explorer became a better browser.
Studying the technology for several years had suggested to me that this view was overly pessimistic. Google's core technology looked highly transferable from Internet search to information management elsewhere, as in its 'Google in a Box' Enterprise Intranet Search servers. There was no actual proof that this was happening - plans and finances were confidential - but it certainly smelt right.
Second guessing Google's plans is like playing chess against a team of grandmasters; you will never have thought through as many of the possible options and combinations as they have. No one, as far as I could discover, saw the strategic significance of Google's apparently very modest acquisitions, whose cost was trivial compared to the billions of dollars worth of established search companies that Yahoo! was intent on buying up. It was also true, but little noticed, that Google also already had a strong user base not amongst its users but from its advertisers, who cannot as a rule switch their budgets easily and quickly.
The question of a floating user base was met on All Fools' Day 2004 with the release of the Beta (trial version) of Gmail, giving users up to a gigabyte of email storage, searchable using Google's technology, and funded by contextual advertising. This blew everybody's hats off. At a stroke Google had moved from an impersonal service to providing one of the most personal services on the planet.
Nobody wants to change their email account and hence address once it is established, and email has proved the stickiest parts of Yahoo! and msn.com (through Hotmail). Email itself dates back to the early eighties, and was heavily used well before the Internet appeared, so the chances of a successful email business maintaining its broad customer base past 2010 is quite feasible.
So much for the problem of an impersonal service without strings or ties for the user. Email is so personal that the requirement for Gmail to search text automatically spooked up a storm on privacy. Liz Figueroa, the Californian state senator, introduced legislation to block indexed email services like Gmail. According to one UK magazine (which cannot be named for legal reasons):Apparently, this violates the assumption that emails are private. We don't know which is worse - interfering in something she appears not to fully understand, or the naivety displayed by that assumption...
Their veteran tail gun proposed a new service - BlackMail - based on the Monty Python TV gameshow. All emails would be searched for incriminating and/or embarrassing details and then, in return for an adequate payment, partners, parents or bosses would not be informed of their contents.
It looked as if for once Google had failed to do their prep, one of their great strengths, and had delivered a crucial new product (albeit in Beta) without having thought through all the legal and security implications - or so I complained to Danny Sullivan's forum on Google's strategy at Search Engine Watch .
Wrong again, because this was to miss the bigger picture. Google's IPO, launched 29 days later, required full disclosure of company activities. Failing to release Gmail, the clearest indicator to date of Google's long-term product viability, would have made them vulnerable to legal actions on the grounds of hiding critical future assets. Ready or not (and the engineering was in much better shape than the public relations) it had to be released before the IPO.
Although the sheer scale of Gmail news coverage, and the size of the resulting privacy storm took Google, and everybody else, by surprise, scanning hundreds of Gmail stories using a Google newsfeed suggested that [65-70]% of coverage was positive.
Google have previously earned a Kim Song Il level [98]% of positive publicity, so adverse publicity must still be painful, particularly when most of it was so darned ignorant, but it also meant that without spending a penny more than the original press release Google had achieved saturation press coverage comparable with the launch of Windows, which had cost Microsoft millions of dollars. Like the golfer Gary Player, the more they practice, the luckier they get.
Being praised by Bill Gates at the World Economic Forum in Davos was probably a bit like Shere Khan admiring your cubs at the Jungle Book 's watering hole. ("Damn - I knew I should not have missed that session", said Sergey when asked about it afterwards). Google had already let it be known that they had refused to sell themselves to Microsoft, preferring to be offered for sale to everybody else either on Internet auction site eBay or in the "Dutch Auction " which subsequently showed up better in the maths. In summer 2003, Google co-founder Sergey Brin had joked, "I think there are a lot of liabilities in acquiring Microsoft."
Industry observers smiled and then shook their heads. 'Ah, remember the last company which humiliated Microsoft? Netscape? Bought by AOL for $9 billion in 1999, and now resting in pieces after final shutdown in late 2003 by owners, poor Time Warner, whose total dot.com losses including AOL's other write-downs, hit another new record of $99 billion. Google's been a great brand but now they have grown too fast and trodden on the tiger's tail. The fickle public will desert them in millions as soon as Microsoft starts to compete seriously, and then it will be tears before bedtime again'.
Well, up to a point. Yahoo!Ős CEO Terry Semel has pointed out that the Search field is not a zero-sum game and there is plenty of room for all three companies to prosper. More, if you include AskJeeves/Teoma or Baidu.com, which a Google-led consortium has bought into after it gained the majority share of the enormous, and exploding, Chinese mainland Search market.
Rather than attempting to crush each other, Google and probably Overture have enabled the diversion of vast funds flowing into classified advertising, worth up to 65% of all current newspaper and magazine income, to be diverted onto the Internet.
Simply put, this technology gives both classified advertiser and customer a better, faster, cheaper deal. That will spell the end of the golden age of printing enabled by digital publishing, although display advertising, which deals with image and aspiration rather than classified's hard product information, does not yet have, and is unlikely to get anytime soon, the quality it would need to migrate from print to the Internet. But it is still going to be a huge shift and Google are in the vanguard with a better indexing technology.
This book will argue that there was something both new and very special about Google's breakthrough technology - context mapping. It has begun to unlock the capabilities of our machines to understand the context, and hence relevance of our documents, a major stumbling block for the whole industry for half a century.
Google represents the engineering of an idea or set of ideas about meaning, and whilst they can protect their engineering from illegal copying, ideas, of their nature, cannot be patented. Their technological breakthrough, unlike their engineering, is no longer exclusive.
Context mapping is a tad technical but it is worth understanding because it will be the beating heart, or so Bill Gates promises, of the computers you will be using in ten years time. It already concerns you every time you look for a new web page or lose a personal letter on your own pc.
Tony Perkins of the AlwaysOn network reported from the Davos Economic Forum in early 2004 that:
Mr. Gates admitted that Microsoft took an approach to search that he now realizes was wrong. "Our strategy was to do a good job on the 80 percent of common queries and ignore the other stuff," he said. He noted that Google was "way better" for people investigating a rare disease, exploring a hobby, or searching for a specific restaurant. "It's the remaining 20 percent that counts...because that's where the quality perception is," he said. "They kicked our butts," Mr. Gates said, as he took personal responsibility for losing out to Google.
But as Gates also predicted that Microsoft search technology would soon outpace that of its rival, so he is really focused on the opportunity. "We will catch them," he said, smiling broadly.
 Beware that Microsoft smile, but unlike Netscape, or even the Apple's desktop, Google employs a new logical method as its core, a different kind of maths application. You cannot really patent an idea, and the idea behind Google will probably inherit the earth, but there is a difference between understanding a fundamental theory, and being able to apply it in the very specific engineering context of four billion HTML web documents or the Windows desktop.
The mathematical engine that powers the Google Search program itself is its core technology, and it is new. It may be possible to reverse engineer the software, or reinvent it, but it cannot then simply be bolted on to established software, like email programs. Finding a solution to this problem is already requiring competitors, particularly Microsoft, to do a lot of real head-busting work to catch up and it is very hard to retrofit onto existing (legacy) software - Microsoft's most precious asset and greatest liability. Sergey refers to this as "integration challenges".
Microsoft CEO Steve Ballmer told CNET that: "We did not commit serious research and development effort (to Internet Search) as soon as we probably ought have... We had a lot on our plate, and we did prioritize, for better or for worse. In a funny way, we made the same prioritization as our No. 1 competitor at the time, Yahoo!. I mean, as bad as I feel, I hope they feel even worse, because they actually had the lead in search, and they didn't invest, and Google came out of nowhere relative to both of us."
During the nineties, Microsoft had had little incentive to intervene in the crowded field of Internet Search. Free search was simply not a profitable arena. That has changed with the success of Google's AdWord. Initially I believed that the more targeted advertising Google pulls in, the greater the pressure on msn.com as an advertising-funded portal. That was wrong.
Partly as a result of Google's activities, Internet advertising has strengthened considerably overall, taking msn.com into its first profit (in February 2004) in nearly ten years of competition with AOL and Yahoo!. Ad sales are driving both growth and profits at msn.com. The $200 million in profits anticipated for the year ending June 30 contrasts with a $500 million loss the year before.
This is the reverse of the Netscape-effect which had forced Microsoft into strategic investments without any model of how to get that money back. But still, it must hurt Bill Gates's indomitable pride that Google Search has made most of Microsoft's search look sick, and the exceptions, on msn.com itself for example, have usually had their search functions bought in from outside Microsoft.
Google may now have given the same energizing shock to the heart of Microsoft that both Apple's desktop and Netscape's browser have delivered in the past. Painful for Microsoft's competitors and executives, no doubt, baby, no doubt, but a fit and fighting Microsoft is darn useful for the United States as a motor of the world's economy. Put bluntly, to stay on top Microsoft now have to adopt Google's core technology, by whatever means necessary, and apply it to Longhorn , the next generation of Windows.
Context mapping is possibly more important than Apple's desktop - the Mac's and then Windows' adoption of the Xerox Menlo PARC's desktop metaphor (Graphical User Interface) - now on almost every pc in the world. Applied maths is an acquired taste, so I will try to explain the general ideas in English rather than algebra, but understanding Google's triumph does require some thinking about the maths and engineering of some of the fundamental theories of how machine intelligence works as well as the blind clash of armies of software hacks in the night.
At the dawn of the 21st century, Google has been a phenomenon, no question, but not an anomaly, as co-founder Larry Page points out. It is one of very few really significant survivals of the dot.com bubble, when it picked up the seed capital that enabled commercial development. "We missed both," says Larry with a certain relish: "We did not launch during the boom and we did not crash during the bust".
This book is mainly concerned with the how and why of the ideas which have made Google a phenomenon, but good ideas do not pay the rent, and it is the engineering of Sergey's "pretty maths" into a user-centric engineering design under Larry Page's leadership which does.
Focus on the user and all else will follow: This user-centric, rather than profit or developer-centred design philosophy is still unusual. Apple and Sony are masters in the field, and the loyalty which their brand inspires has helped them through patches when their technology has faltered. Google will benefit from the same loyalty; although their brand identity is a lot more complex, it can be simplified as the general perception of a company led by a couple of brilliant nerds that you can trust.
Considerable as have been the achievements of Gates, Jobs and Ellison, this is probably not the accolade that they have sought. There is a better parallel here with the founders of Adobe, but luckily for them the closest that Adobe have ever had to confront the general public is in the legions of not-very technical designers who adore Adobe Photoshop. Google's relationship with the public is now much more intense.
The fact is that almost everybody loves Google. Co-founder Sergey Brin knew how and why that had happened when some kids wrote to say that they had used Google to diagnose their father's heart attack - in real time. "An iconic moment" said Brin in an meeting with Danny Sullivan, editor of searchenginewatch.com . No kidding. Equally iconic, if less serious, has to be an appearance with Bart Simpson. 
When I searched Google for this picture, someone on the same thread pointed out that Springfield's main cinema in The Simpsons is called the Googleplex, the name of Google's headquarters in Mountain View (Calif). (So target readers can be Lisa Simpson and the Comic Shop Man - Matt Groening's portrait as an obese nerd obsessed with meticulous detail - and Tom Baker's Dr Who scarf.)
For those who like their consumption more conspicuous, the Sex and the City 'girls' started checking the status of their (phew, final) dates on Google. There were a lot of subsequent puns about the interest of bankers in Google's IPO and potential billions of dollars in commission. In an investment environment still blasted by their own greed and stupidity during the dot.com bubble, the fees the banks were promising themselves from Google's IPO were huge, but Google's founders had other ideas. There was also a tremendous amount of internal pressure for an IPO from original investors and those already vested with stock options, or employees who want to be.
In 1999, the New York Times had reported that "Google is not 'even on the radar screen' of most users and that it will be hard for the company to compete with sites like Yahoo! and Excite for advertising dollars".
So all this happened in the first three years of profitable trading, whilst Excite , and a great many other dot.coms went down the pan. Their first paid product, Adword, was released in 2001, because, as Larry Page tells it, the girls of Palo Alto were not interested in Sergey the dot.com president if Google was not making any money. There was just too much competition from other virtual millionaire males, most of whom found buying a garage in Palo Alto in Y2k was a lot easier than impressing its savvy daughters.
Developing a new technology, which is centred on the needs of the user rather than market-share or profit, without ever neglecting the bottom line, has made Google into a world brand. In 2003, it became the most trusted brand in the world by leapfrogging over Coca-Cola and even Apple's notoriously fanatical fans. By the end of 2003 it had been "profitable since 2001:Q1, employed 700+ people, of whom 60+ have Ph.D.s, was available in 86 languages, had 14 offices, and delivered 150M searches/day", and was valued by Financial Times at $15 billion, in October 2003, when a 2004 IPO was first mooted.
The recipe is easy, but the ingredients are hard to get. Take an unfashionable area, which everybody uses and nobody is satisfied with: here, the messy, unprofitable and increasingly disappointing field of Internet search in 1997. Combine a brilliant mathematical breakthrough with some world-class engineering at a university - Stanford - with massive resources and a relaxed attitude to thesis completion (both co-founders are still 'on leave' from their PhDs). Adopt clear engineering targets and stick to them. Find a new way of generating (advertising) revenue. Then work on optimising it until it grows exponentially into one of the most widely loved, and useful, pieces of software on the planet. Simple really. Co-founder Larry Page listed the reasons "Why Google is still around" very modestly:
"We're lucky. We have a deep technical understanding of what we are doing, which is not true of many companies. Everybody searches. Copying does not cost anything. Distributing another copy costs basically zero because Google surveys the free part of the web."
Google's triple whammy started with a technological breakthrough triggered by the most advanced applied maths and computer thinking at Stanford, Michigan and Maryland universities. They declared a new open world standard for Internet Search on behalf of science and democracy and borrowed enough equipment, money and talent to give the world an enormously powerful, free, Internet search service before developing its technology to generate the Internet's version of classified (contextual) advertising: a 'river of gold', now worth squillions of dollars.
Google's technological supremacy not only helped create the brand, it also gave them 70% + of the global Search market in early 2004, (before the divorce from Yahoo!). This makes Google's benchmark ideals of an unbiased, free public service, not controlled by commercial interests, increasingly important. As the web continues to grow exponentially, it gets more and more technically difficult to achieve that standard both for Google and for its competitors, because the commercial and political incentives to screw up the Open Search advocated by Google and smaller rival AskJeeves(Teoma) increase daily.
But as Biggie Smalls remarked shortly before they shot him: "Jealousy comes with the territory. The more money, the more problems". Ironically, given Google's genesis, The Times's technology columnist, David Rowan has attacked Google's success as a threat to free speech in the Times Online, London.
"Google, the online world's dominant information provider, is now so powerful as to constitute a potential monopoly. If Yahoo! and Microsoft fail to squeeze Google's market share in the search engine wars now being fought, every internet user will be the poorer", according to Rowan.
... "Any company that controls around 80 per cent of web search requests is starting to wield an unhealthy influence on our access to information. If your opinions fall foul of Google, who can stop it from dropping links to your web page? Already the Church of Scientology has used legal threats to have anti-Scientology pages removed from the search index, albeit temporarily. And when Google was negotiating with China to have access restored to the country's web-surfers, there were rumours of compromises, never confirmed, that had blocked sites that might embarrass Beijing."
Political opposition to Open Search is not always for bad reasons. The German government set the legal precedent by demanding that Google (Germany) conform to German federal law and ban the search of neo-Nazi and fascist regalia sites. That seemed acceptable, and Google had complied when the government of China objected to dissent sites appearing in Google and AltaVista searches.
But that note on China is a little odd coming from the Times - a paper owned by News International whose Harper Collins publishing division decided against publishing former Governor-General Chris Patten's account of the return of Hong Kong to China because it might have been offensive to the Chinese government. James Murdoch, chief executive of BskyB and son of founder Rupert, spelt it out when he said :
"...Hong Kong's democracy advocates should accept the reality of life under a strong-willed "absolutist" government. Speaking at the Miliken Institute's annual business conference in Beverly Hills, Murdoch criticized the press for "portraying a falsely negative portrait of China" by focusing on "destabilizing" issues such as human rights and Taiwan.
Closer to home, that Times article accused Google of invading users' privacy. This proved prescient if a tad paranoid given the fears of privacy-invasion when Google proposed Gmail. The raw common sense that services have to be paid for, and if you find contextual advertising intrusive, then there are plenty of other (non-searchable) email services around, that you can pay for, seems to have eluded some people.
"All this should not, of course, prevent the ordinary web surfer from using Google in preference to some of its increasingly effective rivals, such as Teoma.com and Vivisimo.com. But if you do, you might like to know that Google stores for years a detailed note of everything you search for, and at what time of day, which it logs according to your computer's address. It may, it says, "release specific personal information about you" to the powers that be. You're not feeling quite so lucky now, are you?"
Serious issues, but that last comment, a swipe at Google's much-liked 'I'm Feeling Lucky' button, verges on the malicious. If a user is making repeated searches on for example 'child sex pain' then maybe Google or Yahoo! should be asked to tell the police in his/her area. Privacy is important, but some issues, like child protection, matter even more. We Europeans sometimes like to duck responsibility. We can now start to set public standards of what we expect from the Internet in open, democratic debate or we can leave it to very large commercial companies and the US government to decide on our behalf, and then complain about it afterwards.
Out of scientific idealism more than anything, Google set the Internet benchmarks on honest information searching in 1997. Or as Brin put it: "Basically, our goal is to organize the world's information and to make it universally accessible and useful. That's our mission." Since then Google has become a universal tool and a global media brand. Conflicts between public service and the need to grow profits are going to be inevitable. So maybe this is now a job for a W3C and/or a UN/International Atomic Energy Authority task force to work on.
This is not frivolous. IAEA experts showed that Niger's 'sale' of Yellow Cake (uranium ore) to Iraq was a forgery by getting comparable documents from Google. "What? You Googled them?" The BBC2 Newsnight presenter was almost incredulous. "But, naturally." the expert replied. Google's idealism has been crucial, but we cannot rely on it forever. Time, maybe, for Google's founders to set up an Open Search research foundation, perhaps with Gloria Page or Professor Michael Brin as trustees.
In commercial search, Google are still highly unusual in declaring their belief in open, unbiased standards, AskJeeves is another more recent, exception. Google have been unique, perhaps, in publishing the methodology of some of their central ideas and core methods; (there's also now a fully open source alternative, but it is not very good yet). No good deed goes unpunished, and Google's core ideas, if not the actual maths and engineering, are almost invariably incorporated into the new generation software of spammers, competitors, clones and other developers.
Google's original open standards are now subject to a continuous war of attrition from link-farms - a form of spam designed to twist the technology until it breaks. To stay on top, Google have to continue to outwit and outthink the spammers, Microsoft and their commercial competitors.
Google's users finally only care about how well Search works and that quality is now subject to erosion, mostly malicious but also due to the exponential growth in the size of the web and its increasingly adept use of dynamic output. For Google, the serpent in the garden is currently not Microsoft but spam, spam, spam and spam. Not the unwanted emails which Bill Gates wrote to the US Senate Commerce Committee about, because they currently constitute more than 50% of all email traffic, but link-farms which jack up a site's rating in a Google Search list by creating artificial, meaningless links to inflate the apparent popularity and hence relevance of a client's site.
One spammer even tried to cut out the middleman and extort $100,000 directly from Google in exchange for not releasing his Adsense-cheating software. Google got the police to record the calls and he was arrested. His lawyer will probably have a field day because it is clearly blackmail but even Californian law may not yet define it as such.
In order to counter the sheer, relentless scale of spam attacks as fast as possible, Google applied an arguably more erratic and unpredictable technology (Bayesian probabilities) as spam filters, beginning with the notorious Florida 'dance' (monthly index update) in November 2003. Sites claiming to be blameless and innocent subsequently disappeared from their most cherished search listings perhaps because this implementation of the Reverend Bayes's theories was not quite up to the job.
Google can overcome this - they have the tools - but they do need the time and space to work the old magic, because their opponent is now hydra-headed and targeted at Google's methods, which the original data (Internet pages) was not - just very full of innocent mistakes. But Google's usefulness since 2000 has already built up a tremendous fund of goodwill amongst millions of ordinary users. This loyalty can cushion you if your technology goes through a bad patch, as Yahoo! also knows.
Seasoned industry savants are still amazed by the loyalty that Google evokes from consumers and concerned that this might become a liability. "When you're God, you have quite a long way to fall if you do something wrong,It's hard to think of a company that's been put on a pedestal as much as they have." According to Danny Sullivan, quoted in the Washington Post . (Sergey Brin and Larry Page used Search Engine Watch as their benchmark in 1997 when they first described their new architecture so this book follows their example.)
This status has been achieved by a combining a new mathematical solution to an ancient philosophical problem, developing an engineering distribution system with an unusual degree of speed, reliability and integrity, and then providing the results for free. That matters to most users a lot more than any mission statement even when Google's founders clearly believe every word of their 10-point credo. Ten things Google has found to be true
1) Focus on the user and all else will follow.
2) It's best to do one thing really, really well.
3) Fast is better than slow.
4) Democracy on the web works.
5) You don't need to be at your desk to need an answer.
6) You can make money without doing evil.
7) There's always more information out there.
8) The need for information crosses all borders.
9) You can be serious without a suit.
10) Great just isn't good enough.
(But, of course, if you want really grand statements of the highest business ethics, those of both Enron and BCCI are most impressive: - It ain't what you do, it's the way that you sell it - that's what gets...)
A technology that began with Sir Tim Berners-Lee's free distribution of a common Web language (HTML) developed into the corruption of online fraudsters aided and abetted by the efforts of Wall Street's, and the City of London's, finest. Google is one of the very few companies to be seen by millions of users to have retained its integrity and its potential by not only promising, but actually delivering a new kind of free service only made possible by the new technology of the Internet.
Linux, php, the Open Directory Project, Wikipedia, eBay and Amazon have a similar cachet. The first four are some of the collective, voluntary manifestations of the Open Source movement, a strange new form of politics which tends to be understood only by techies and developers, whilst Amazon and eBay's achievements are obvious to any ordinary punter who buys books or wants to auction consumer goods. But both eBay and Amazon are primarily business successes, using established technology. Google's combination of sound business and revolutionary technical innovation has not been seen since the earliest days of Apple, when Microsoft was still called Traff-0-Data.
As a UK citizen, I was not entitled to participate in the IPO and, to avoid any conflict of interest, as long as I am writing about Search technology and Google I will not buy shares in them or any other technology company. This book's editors do not have any connection with Google, although we have asked, so far without a response, if the AdSense program can be added to this site; partly so that we can track how it responds to the text.
Google's competitors may complain that this book is the most-heavily annotated fan letter in history, and I admit that I am a fan, partly because I have been working with the ideas behind the technology for a long time and Google's technology has my unqualified professional respect; by raising the bar in Internet search they have made my own work much easier and more interesting. Five years ago I would have paid to subscribe to a Google service, and I believed that they would have to go that way. Keeping Search free by rethinking contextual advertising leaves me, and millions of what the French call informaticiens in their debt. That helps create a powerful brand loyalty.
Google's technology is central to its present and potential success as a business but it is also part of a much wider and older story. Brin and Page have only made the most modest and realistic claims for Google, but there is no doubt that what they are really after - and what Gates also wants most - is the Big One: a program that understands meaning.
Never settle for the best
"The perfect search engine," says Google co-founder Larry Page, "would understand exactly what you mean and give back exactly what you want." Given the state of search technology today, that's a far-reaching vision requiring research, development and innovation to realize. Google is committed to blazing that trail
Ok, it is visionary. Even with Google's significant contribution we are still a long way from the perfect search. Craig Silverstein, Google's head of technology thinks it could be two, or even three hundred years before our computers can understand the real meaning of a query. But bear in mind that Google was developed at Stanford, in what looks from London to be the best school of artificial intelligence in the world, led by Professor John McCarthy.
In 2004 Brin and Page were made Fellows of the Marconi Foundation at Columbia University, New York. Sir Tim Berners-Lee, who won the same prize in 2002, said in a statement: "Google held a mirror up to us, reflecting the myriad little actions of linking as a set of concepts which society has discussed and sought." The meanings generated have been relevant enough to save lives and to move the whole Internet forward. Now there's Glory for you.
© 2004 Fleet Books London, UK. Permission required for any republication.
|