從
谷歌公司(Google Inc.)一份七頁的機密“願景陳述”文件可以看出﹐這家信息時代的巨頭正在仔細琢磨一個基本問題：守著大量有關人們上網活動的數據﹐谷歌在從這個寶藏獲利方面應該走多遠。
要不要進一步利用對Gmail用戶的知悉獲利？要不要建立一個龐大的“交易平台”﹐用於買賣網絡數據？要不要讓人們只有在付費的情況下才能看不到任何廣告？
包 括這幾個想法在內﹐文件討論了各種大大小小的想法（上面第三個想法被打入“愚蠢”(wacky)之列）。這份被《華爾街日報》看到的文件﹐是谷歌興趣廣告 部門的高級產品經理韋恩伯格(Aitan Weinberg)在2008年年底編制的。和十幾位谷歌前任和現任員工接受採訪時的陳述一樣﹐這份願景陳述文件對谷歌力保自己信息經濟先鋒地位的努力提 供了一種坦白、自省的寫照。
谷歌正在挺進自己未曾涉足的用戶隱私領域。它在過去有所克制﹐沒有大規模地利用它手中有關網民的數據賺錢﹐擔心這樣做會招致不滿。但由於跟蹤人們上網行為並出售相關數據的競爭對手迅速湧現﹐加上Facebook Inc.的成長﹐谷歌被迫開始轉變。
一位知情人士把這份原景陳述稱為一份“頭腦風暴文件”﹐並說它並沒有呈遞給谷歌的管理高層。這位人士說﹐其中的部分想法“完全行不通”。
但有幾個想法得到了實施。例如在去年﹐谷歌首次開始搜集一種有關人們訪問哪些網站的新數據﹐然後利用這種數據在整個互聯網上追蹤網民並向他們展示廣告。
上個月﹐在被一位記者問到谷歌有關其數據的計劃時﹐谷歌創始人之一佩奇(Larry Page)回答說﹐我們正在非常迅速地行動﹔數據有很多新的用途﹐對消費者受到損害的擔心是沒有根據的﹔人們總是很容易對可能發生的事情感到擔心﹐對不對？
隨 著谷歌做出改變﹐它可能會裹挾整個互聯網世界隨它前行。在用戶數量首屈一指的情況下﹐它比其他任何互聯網企業都更有能力把定向投放廣告的新方式變為主流方 法。這家公司還積極參與一些經常在內部制定新的隱私慣例的行業組織﹐這些組織的此類舉動旨在阻止國會出台隱私權方面的監管法規。美國聯邦貿易委員會 (Federal Trade Commission)去年說﹐只要企業向用戶披露它們的隱私慣例等內容﹐它們在這一領域是可以實現自我監管的。
Associated Press
谷歌創始人之一佩奇
谷歌對網絡隱私問題的重要性無以復加。市場調研公司comScore提供的數據顯示﹐全世界大約75%的互聯網用戶﹐即9.438億人﹐6月份都使用了谷歌的服務﹐超過其他任何互聯網企業。
谷歌的願景陳述文件在談到如何使廣告與互聯網用戶更有相關性時﹐把自己巨大的搜索數據庫說成是“在互聯網上發現用戶個人興趣的最佳源泉”。它說﹐沒有其他哪家公司可以與之競爭。文件後面的部分又警告說﹐一些想法有的“安全”﹐也有的“不安全”。
最大膽的那些想法﹐將讓谷歌在跟蹤人們上網行為並從中獲利的這塊業務中佔據優先地位。比如﹐如果是建立一個數據交易市場﹐那麼來自很多源頭（包括谷歌）的個人信息將可以被融合起來﹐並被用於對個人上網行為實施高度個性化的跟蹤。
BlueKai Inc.和eXelate Media Ltd.等微型公司已經在提供這樣一些服務﹐迫使谷歌追上來。《華爾街日報》一項名為“他們知道哪些”(What They Know)的調查﹐就旨在考察這類數據不斷增多的交易﹐及其對個人隱私造成的後果。
谷 歌有意在這方面的某些技術上落後於人。在創始人布林(Sergey Brin)和佩奇的強烈堅持下﹐谷歌多年來秉持廣為人知的非正式口號“不作惡”(Don't Be Evil)﹐沒有使用任何手段在網民不知情的情況下跟蹤他們的上網行為。但《華爾街日報》對谷歌十幾位現任和前任員工的採訪顯示﹐布林和佩奇已經逐步地認 定﹐他們可以在不利用網民的情況下開始從公司自己所掌握的數據當中獲取利益。
谷歌隱私工程的負責人惠滕(Alma Whitten)今年6月接受採訪時說﹐兩位創始人相信﹐他們是在提升網民的體驗﹐對用戶有利的東西﹐對廣告客戶也有利。
《華爾街日報》對網上跟蹤行為滋生現象的最新調查發現﹐谷歌的跟蹤代碼出現在了美國50家最熱門網站中的45家。（可點擊WSJ.com/WTK瞭解調查的詳細情況）
2008年願景陳述﹐加上《華爾街日報》看過的其他十幾份內部文件﹐都揭露了谷歌在創始人的看法不斷發展變化之際﹐如何拖拉不合作的內幕。
廣告銷售是谷歌的一大搖錢樹﹐而在線廣告業務開始擴大到谷歌最擅長的領域之外﹐銷售與用戶使用的搜索引擎詞匯相關的廣告。相反﹐廣告客戶希望根據愛好、收入、疾病或朋友圈等更具體的個人資料來針對不同的用戶。
谷歌的變化反映出在線領域的實力再調整。多年來﹐互聯網領域實力最強大的公司都是一些訪問量最大的公司。而如今﹐實力則蘊藏在那些擁有最豐富數據並知道如何加以利用的公司中。
Associated Press
這將互聯網廣告公司推入了一場“軍備競賽”﹐速度如此之快﹐即便是谷歌也擔心會落在後面。《華爾街日報》看過的、2008年年中的一份內部報告幻燈片的題目就很直白：參與到這場遊戲中來。
該幻燈片描述了進軍“展示”廣告銷售這一有利可圖的業務領域的重要性。展示廣告是一類配有圖片的較大廣告﹐與之對應的是較小的文字廣告。據分析人士說﹐如今﹐在美國展示廣告收入方面﹐谷歌仍落後於市場領頭羊雅虎(Yahoo)。
當然﹐總體來講﹐谷歌仍在互聯網業務中佔有頭把交椅。谷歌2009年實現收入237億美元﹐比排在第二位的雅虎高出三倍以上。谷歌的在線廣告業務比公開上市的美國競爭對手增長得更快。
不過﹐谷歌的收入增長已經大幅放緩。社交網絡巨擎Facebook向逾五億用戶銷售具有高度針對性的廣告﹐對谷歌來說﹐它是個越來越大的威脅。
對Facebook的恐懼深入谷歌。谷歌目前正在設計自己的社交網絡服務。相比之下﹐谷歌2008年願景陳述中還很少提及社交網絡﹐這也顯示出形勢的瞬息萬變。
很 少有在線公司具備像谷歌那樣瞭解用戶的能力。以住在曼哈頓東村的26歲男演員阿里•布蘭德(Ari Brand)為例。谷歌可以知道他花733美元買了一台平板電視﹐因為他把自己的預算上傳到了在線文字處理器和表格軟件Google Docs上。谷歌可以訪問他通過Gmail發送的2.3萬封電子郵件。谷歌還存儲了布蘭德的谷歌搜索信息。
不過﹐重要的是﹐谷歌並沒有將這些不同類型的個人信息混雜在一起。舉例來講﹐谷歌不會用從一個人的Gmail帳戶中獲得的信息在其他在線業務中發佈針對這個人的廣告。不過﹐谷歌的電腦確實會掃描Gmail郵件﹐在電子郵件旁邊放置內容相關的廣告。
谷歌還說﹐它的大部分信息是無法根據人名與一個具體的人掛鉤的。
高管們早就認為隱私風險相對於商業回報來說太大了。據瞭解谷歌想法的人士說﹐他們感覺人們給谷歌設定的標準比其他不那麼出名的公司要高﹐他們希望讓更加強勢的競爭對手來檢測這些標準。
在從谷歌獨家數據中尋找獲利新方法方面﹐對反壟斷審查的擔憂也使其風險增大。
分析人士說﹐2006年前後﹐谷歌的視線還沒有放到Facebook身上──而是在美國在線(AOL)和雅虎身上﹐這兩家公司共控制了展示廣告業務的約40%。
在展示廣告上贏得更多業務所面臨的一大障礙是谷歌創始人佩奇﹐他反對出於發佈針對性廣告的目的讓谷歌在人們的電腦上安裝“cookie”。Cookie是一種小文本文件﹐它可以用來幫助跟蹤人們的在線活動﹐以向其顯示針對他們興趣的廣告。
這 樣的政策打擊了谷歌的展示廣告銷售﹐廣告有效性下降。沒有廣告cookie﹐谷歌僅僅可以根據一個網頁的名字或內容銷售相關廣告。比如﹐在有關鞋類的頁面 上放置一個鞋類廣告。這就是所謂的“內容”定位﹐很多廣告客戶認為它沒有“行為”定位有效。行為定位會識別具體的用戶和他們的興趣。
在谷歌內部各方就此爭論不休之際﹐有個機會出現了。在網站上放置展示廣告的巨頭DoubleClick掛牌出售﹐而谷歌主要競爭對手微軟(Microsoft Corp.)在覬覦這項業務。
據知情人士說﹐谷歌高管對DoubleClick在很多用戶不知道自己被跟蹤的情況下﹐利用cookie來跟蹤人們在線活動的做法心存戒心。
他們認為﹐收購DoubleClick將立即帶來展示廣告技術和客戶。
2007年谷歌決定以31億美元收購DoubleClick。知情人士稱﹐與此同時﹐一些雇員開谷歌的玩笑說﹐公司花幾十億美元就是為了使佩奇喜歡cookie。
谷歌和DoubleClick高管聚在一起決定如何整合兩公司產品﹐他們有大量工作要去做。
根據題目為“參與到這場遊戲中來”的2008年7月的演示幻燈片﹐谷歌較之微軟旗下競爭對手Atlas在衡量廣告有效性方面方法更少﹐而且沒有美國在線旗下Tacoda分支那樣的行為定向能力﹐這意味著谷歌不能根據人們瀏覽的網站設定目標廣告。
谷歌高管最終決定使用cookie。由此而產生的結果是﹐在谷歌賣出的每一頁展示廣告上﹐都開始在用戶電腦上安裝DoubleClick的cookie。
谷歌第一次有能力針對個人電腦發放廣告﹐但正是因為有這種能力﹐谷歌沒有開始使用﹐它還有許多內部阻力。
韋恩伯格是2008 年出版的《願景陳述》的作者﹐他從DoubleClick加入谷歌。他及幾位產品經理和營銷高管開始討論谷歌更積極鎖定廣告目標的方法。
他標明“內部機密”的備忘錄承認這一主題的精確性﹐它的第一句話開宗明義地說﹐鎖定觀眾“本質上就很敏感”﹐因為用戶有可能“誤解”。
這一備忘錄概括出一個籠統的觀點﹐即谷歌可以聯合其他網站出於鎖定廣告的目的分享他們的數據。
該文件還說﹐谷歌可以從客戶的Gmail帳戶及他們對谷歌檢驗服務的使用﹐而知道的客戶的情況﹐通過網絡開始出售廣告。谷歌的“Checkout服務”是PayPal的競爭對手。
所有這些將是重大改變﹐當前儘管谷歌在用戶Gmail帳戶內放置上下文廣告﹐但它並不帶著那些廣告跟蹤這個人到其他網站上。
該文件顯示了對隱私含義的覺醒。該文件稱﹐“如果在內心沒有對隱私、法律和行業最佳實踐的嚴格考慮”﹐什麼也不會發生。文件稱﹐目標是在用戶看到精緻的目標廣告時﹐減少客戶的“緊張”感覺。在2008年底前﹐谷歌高管準備基於用戶興趣推出廣告﹐但細節仍在討論中。
大約18個月前﹐十幾位高管在谷歌位於加州的總部舉行了會議﹐會上佩奇和布林就谷歌該如何積極鎖定目標大聲爭吵﹐這使氣氛突然緊張起來﹐一位對這次會議有所瞭解的人士說﹐真讓人難堪﹐就像在看父母打架一樣。
這位知情人士說﹐布林比佩奇更不情願﹐最終他表示了默許﹐根據人們的興趣來銷售廣告的做法得以推進。
谷歌於2009年3月推出新的廣告產品──“興趣廣告”。這一服務目前只面向有限的廣告商﹐利用cookie追蹤用戶任一時間在谷歌銷售廣告的上百萬個網站中任何一個網站內的足跡。
為了打消創始人對cookie隱匿性的擔心﹐谷歌建立了一個網頁www.google.com/ads/preferences﹐人們可以選擇退出。
谷歌接受了願景陳述的其他想法。去年9月﹐谷歌推出新的廣告交易﹐它讓廣告商鎖定個人──例如鞋類市場的消費者──在他們上網的時候即時追蹤他們。谷歌在每筆廣告銷售中提成。
簡言之﹐谷歌試圖把自己打造為盡可能多的廣告交易的結算所﹐即使這些交易實際上並不涉及谷歌提供或看到的消費者數據。
對谷歌來說更進一步的是成為每個人數據的結算所。瞭解這一談話的人士說﹐這一想法也出現在了願景陳述中﹐但正在考慮之中。谷歌本已是消費者數據的最大倉庫之一﹐此舉將使公司成為其他人數據的交易中心。
Jessica E. Vascellaro
A confidential, seven-page Google Inc. 'vision statement' shows the information-age giant in a deep round of soul-searching over a basic question: How far should it go in profiting from its crown jewels -- the vast trove of data it possesses about people's activities?
Should it tap more of what it knows about Gmail users? Should it build a vast 'trading platform' for buying and selling Web data? Should it let people pay to not see any ads at all?
These and other ideas big and small -- the third one was listed under 'wacky' -- are discussed in the document, which was reviewed by The Wall Street Journal and compiled in late 2008 by Aitan Weinberg, a senior product manager for interest-based advertising. Along with interviews with more than a dozen current and former employees, the vision statement offers a candid, introspective look at Google's fight to remain at the vanguard of the information economy.
Google is pushing into uncharted privacy territory for the company. Until recently, it refrained from aggressively cashing in on its own data about Internet users, fearing a backlash. But the rapid emergence of scrappy rivals who track people's online activities and sell that data, along with Facebook Inc.'s growth, is forcing a shift.
A person familiar with the matter called the vision statement a 'brainstorming document' and said it wasn't presented to senior executives. Some of its ideas are 'complete non-starters,' this person said.
Still, several have been implemented. Among them: Last year, Google for the first time started collecting a new type of data about the websites people visit, and using it to track and show them ads across the Internet.
'We are moving very fast,' said Google co-founder Larry Page last month in response to a reporter's question about Google's plans for its data. 'There are lots of new uses for the data. Worries about harming consumers are 'hypothetical,' he added. 'It is always easy to be fearful of what could happen, right?'
As Google changes, it's likely bring the rest of the online world with it. With more users that any other Internet company, it has an unparalleled ability to make new ad-targeting methods mainstream. The company also actively participates in trade groups that regularly craft new privacy practices among themselves in hopes of thwarting legislation. The Federal Trade Commission said last year that the field can regulate itself as long as companies disclose their practices to users, among other things.
Google is overwhelmingly important to online privacy. Roughly 75% of global Internet users, or 943.8 million people, used its services in June, more than any other Web company, according to comScore.
Google's vision statement describes the company's immense search database as 'the BEST source of users' interests found on the Internet,' during a discussion of ways to make ads more relevant to users. 'No other player could compete,' it says. Later, the document warns that some ideas range from 'safe' to 'not' safe.
The most aggressive ideas would put Google at the cutting edge of the business of tracking people online to profit from their actions. A data-trading marketplace, for instance, would allow personal information from many sources -- including Google -- to be combined and used for highly personalized tracking of individuals.
Tiny companies like BlueKai Inc. and eXelate Media Ltd. already offer some of these services, pressuring Google to match them. A Wall Street Journal investigation, 'What They Know,' is examining the widening trade in this kind of data and the consequences for individual privacy.
Google trails in some of these techniques by choice. Famous for its unofficial corporate motto, 'Don't Be Evil,' for years it resisted using any method to track people online without their knowledge at the fierce insistence of founders and Sergey Brin and Mr. Page. But the two men have gradually decided they can begin exploiting the data their company controls, without exploiting consumers, according to interviews with more than a dozen current and former employees.
The founders believe they're improving the Internet user's experience, said Alma Whitten, who leads Google's privacy engineering, in a June interview. 'What's good for the consumer is good for the advertiser.'
A recent Journal examination of the proliferation of online tracking found that Google's tracking code appeared on 45 of the 50 most popular U.S. websites. (For details on those findings, go to WSJ.com/WTK.)
The 2008 vision statement, along with a dozen other internal documents reviewed by the Journal, tell the inside story of how Google dragged its feet while its founders' views evolved.
Selling ads is Google's big money-maker, but the online-ad business is broadening away from Google's sweet spot, selling ads tied to the search-engine terms people use. Instead, advertisers want to target people based on more specific personal information such as hobbies, income, illnesses or circles of friends.
The changes at Google reflect a power realignment online. For years, the strongest companies on the Internet were the ones with the most visitor traffic. Today, the power resides with those which have the richest data and are the savviest about using it.
That has propelled Internet ad companies into an arms race so swift that even Google fears being left behind. One slide from an internal presentation in mid-2008, which was reviewed by the Journal, is headlined bluntly: 'Get in the Game.'
That particular slide describes the importance of breaking into the lucrative business of selling 'display' ads, which are larger ads with pictures, as opposed to smaller text ads. Today, Google still trails market leader Yahoo in U.S. display-ad revenue, according to analysts.
Google still leads the Internet pack overall, of course. Its revenue, $23.7 billion in 2009, is more than three times Yahoo's, its closest competitor. Its online advertising business is growing faster than those of its publicly held U.S. rivals.
But Google's revenue growth has slowed dramatically. And social-networking powerhouse Facebook is a widening threat with its ability to sell highly targeted ads to its more than 500 million users.
Facebook fears run deep at Google, which is designing its own social-networking service. In a sign of how quickly things change, Google's 2008 vision statement scarcely mentioned social networks.
Few online companies have the potential to know as much about its users as Google. Consider 26-year-old Ari Brand, an actor living in Manhattan's East Village. Google could see that he paid $733 for a flat-screen TV, because he uploaded his budget to Google Docs, an online word processor and spreadsheet. It has access to the 23,000 emails he's sent through Gmail. Google also saves Mr. Brand's Google searches.
Significantly, however, Google doesn't mix those separate pots of personal data. For instance, it doesn't use data gleaned from a person's Gmail account to target ads to that person elsewhere online. Google's computers do, however, scan Gmail messages to place contextual ads next to the emails themselves.
Google also says much of its data can't be tied to a person by name.
Executives long considered the privacy risks too great relative to the business rewards. According to people familiar with Google's thinking, they felt the company was being held to a higher standard than less well-known firms, and preferred to let more aggressive rivals test the boundaries.
Concerns about antitrust scrutiny also heightened the risk of finding new ways to profit from Google's exclusive data.
As recently as 2006 or so, Google's sights weren't set on Facebook -- they were set on AOL and Yahoo, which together controlled roughly 40% of the display-ad business, analysts say.
One big obstacle in winning more of that business was Google founder Mr. Page, who objected to letting Google install 'cookies' on people's computers for purposes of targeting ads. Cookies are little text files that can, among other things, be used to help track people's activities online to show them ads targeted to their interests.
Those policies hurt Google's display-ad sales and made its ads less effective. Without the advertising cookies, Google could sell ads based only on the name or content of a page -- for instance, putting a shoe ad on a page about shoes. That's known as 'contextual' targeting, and many advertisers consider it less effective than 'behavioral' targeting, which identifies specific users and their interests.
As factions inside Google fought over the issue, an opportunity arose. DoubleClick Inc., a giant in the business of placing display ads on websites, put itself up for sale -- and Google archrival Microsoft Corp. was circling.
Google executives were leery of the way DoubleClick used cookies to track people online, on the principle that many users had no idea they were being tracked, people familiar with the situation say.
But an acquisition of DoubleClick would instantly bring in display-advertising expertise and clients, they thought.
In 2007, Google agreed to buy DoubleClick for $3.1 billion. At the time, some employees joked Google had to spend billions just to get Mr. Page to like cookies, people familiar with the matter say.
Google and DoubleClick executives huddled decide how to blend the two companies' products. They had a lot of ground to make up.
According to a resulting presentation slide, dated July 2008 -- the one headlined 'Get in the Game' -- Google offered fewer ways to measure an ad's effectiveness than Atlas, a rival owned by Microsoft. And Google had none of the behavioral-targeting capacities of AOL's Tacoda unit -- meaning it couldn't target ads to people based on websites they visited.
Google executives finally agreed it was cookie time. As a result, every page where Google sold a display ad began installing DoubleClick cookies on users' computers.
For the first time, Google had the ability to deliver ads targeted to individual people's computers. But just because it had the ability, Google didn't start using it. There was still too much internal resistance.
Mr. Weinberg, the author of the 2008 'vision statement,' came to Google from DoubleClick. He and a small group of product managers and marketing officials began discussing the ways Google could target ads to people more aggressively.
His memo, stamped 'INTERNAL CONFIDENTIAL,' acknowledged the delicateness of the subject. Audience targeting is 'of a sensitive nature,' it stated in the very first sentence, due to the possibility of 'mis-understanding' among users.
The memo then went on to outline a sweeping vision in which Google could get other websites from around the Internet to share their data with it for the purpose of targeting ads.
The document also says Google could start selling ads across the Web based on the things it knew about people from their Gmail accounts, and also from their use of Google's Checkout service, a PayPal rival.
All of that would be a significant change. Currently, although Google places contextual ads within a user's Gmail account, it doesn't follow that person to other websites with those ads.
The document shows awareness of the privacy implications. Nothing would happen 'without strong consideration of privacy, legal and industry best practices in mind,' it states. A goal should be to limit users' feeling of 'creepiness' from seeing finely targeted ads, it says. By late 2008, Google executives were preparing to launch ads targeted at users' interests. But the specifics still remained controversial.
Tensions erupted during a meeting with about a dozen executives at Google's Mountain View, Calif., headquarters about 18 months ago when Messrs. Page and Brin shouted at each other over how aggressively Google should move into targeting, according to a person who had knowledge of the meeting. 'It was awkward,' this person said. 'It was like watching your parents fight.'
Mr. Brin was more reluctant than Mr. Page, this person said. Eventually, he acquiesced and plans for Google to sell ads targeted to people's interests went ahead.
Google launched the new advertising product, 'interest-based ads' in March 2009. The service, currently available only to a limited group of advertisers, uses cookies to track any time a user visits one of the more than one million sites where Google sells ads.
To offset the founders' concerns about cookies' secretiveness, Google set up a page, www.google.com/ads/preferences, where people can opt out.
Google adopted other vision-statement ideas. Last September, it launched its new ad exchange, which lets advertisers target individual people -- consumers in the market for shoes, for instance -- and buy access to them in real time as they surf the Web. Google takes a cut of each ad sale.
In short, Google is trying to establish itself as the clearinghouse for as many ad transactions as possible, even when those deals don't actually involve consumer data that Google provides or sees.
The further step in that progression would be for Google to become a clearinghouse for everyone's data, too. That idea, also laid out in the vision statement, is still being considered, people familiar with the talks say. That would put Google -- already one of the biggest repositories of consumer data anywhere -- at the center of the trade in other people's data as well.
2010年08月11日08:56
Google Agonizes On Privacy As Ad World Vaults Ahead
