Part 3: Search for information - DISRUPTIVE INNOVATION AND ENTREPRENEURSHIP

💡 💲  📈 Innovation in information services. How Google became the best search engine! How Wikipedia reinvented Encyclopedia!

Retrospective:  from the first address books to yellow pages

Word of Innovator: Steve Kirsh

SEARCH FOR INFORMATION

Encyclopedias, directories, search engines
Book "Innovation 50 success stories" printed in French by Dunod - looking for a publisher for the English version

Do you know Google? So you know what a web search engine is. These engines have created an unprecedented service, used worldwide today, which allows the search of all kinds of information, in all fields. The search engines are used to find web sites, addresses of service providers, practical knowledge, commercial information, images, using key words. Search engines, or should we say now Google, are probably the most important service innovation for the last half century.

Until recently, the yellow pages of the telephone directory and word of mouth were the most effective means of finding a craftsman, a service provider, a business, the address of an institute or a professional. The search engines added a new dimension to this service by constituting the first directory with a universal dimension. We use them among others to find people, providers, to have information and opinions about them, to locate them. Google and Yellow Pages directories are well known. The innovation processes that allowed their creation are less known.

The search for encyclopedic knowledge has been also recently transformed. Wikipedia has revolutionized the production models of traditional encyclopedias and supplanted the Encyclopedia Britannica. Wikipedia is based on contents produced by a very large community: 130,000 "Wikipedians" are contributing each month . It improves knowledge by taking advantage of an increasingly educated population and values an interconnected world.

Search engines, yellow pages and Wikipedia are examples of information services delivered via a media (a website or a directory). They are supported by methods for collecting and updating information that can be manual, collaborative or automated. Several services of this type preceded the yellow pages. These were called: address books, almanacs and city directories.

 GOOGLE, THE RESULT OF SEARCH ENGINES
Google offers a very exhaustive information retrieval service. It is difficult to take it in default. It has an answer to everything, or almost. Curiously, it will not tell you the current number of unique Google visitors per month! It is in the order of $ 1.5 billion. This invention was built in just twenty years. The documents available on the Web allow us to trace this story.

The paradigm of automatized information retrieval is mentioned for the first time in 1948 by Calvin N. Mooers, who introduces the principle in a report of the Massachusetts Institute of Technology. The object is to find information whose location, or even existence, is a priori unknown.

The first attempts to retrieve files and to index texts deposited on the Web in an automated way appeared in the 1990s. Initially, they were not the subject of very intense attention in the academic field. Perhaps, because they were not seen as a fundamental problem sufficiently complex at that time. These are young computer engineers with entrepreneurial spirits who will take ownership of this topic. Prior to Google, several search engines had success in the 1990s: WebCrawler, Lycos, Yahoo, Infoseek, Excite, AltaVista. Their utility for searching research papers is quickly perceived in scientific communities. Users were evaluating themselves the performance of the different available search engines in comparing their answers. They recommended the best one through word-of-mouth  The best performing search engine seems to have to impose itself. There is, however, a way to distort this comparison by setting a default search engine in web browsers or in the main Web access portals. At that time, the various search engines provided sometimes complementary answers. Meta-engineer, such as MetaCrawler, appeared. They automatically integrated the additional answers of several search engines. Their success was short-lived because they were depending on the same sources of incomes as the search engines and thus competing with them. Their internet connections were blocked by the engines platforms which preserved like this their source of revenue: the audience.

#5 Google, best in class
The Google search engine was created in September 1998, four years after the first search engines. It is the result of research work started in January 1996 by two PhD students from Stanford, Larry Page and Sergey Brin. Google has introduced innovations not only at the algorithmic level but also at the business model level.

• SERIAL ENTREPRENEURS: Sergey M. Brin, born in Moscow and emigrated to the United States in 1979, and Larry Page, American, are the two founders of Google. Sons of scientists, they had been at Montessori schools and met during their PhD in Stanford. L. Page had for thesis director T. Winograd, a renowned professor in AI and natural language. S. Brin was in Data Mining in a research program on the mathematical and statistical software Mathematica.

💡 Innovation: Google's first innovation was to implement a new algorithm  that evaluates differently the relevance of web pages in response to a user's search. This algorithm determines the responses to be returned and their presentation order ("i.e. their ranking") according to their referencing by web sites which themselves are relevant to the search. The algorithms used before were based on the computing of the number of iterations of a word or a concept in each Web page, and in the Web sites referenced by it. The approach is thus reversed.  It was easy to falsify the web page relevance with the previous search engines by adding hidden iterations of a keyword. With this new algorithm, Google  became more powerful and more reliable than the previous search engines. A patent filed in January 1998 by Larry Page is granted to Stanford University. In exchange for the transfer of rights of use, Google sells a portion of the capital, corresponding to 1.8 million shares, to this university. Stanford will sell its listed shares in 2005 for $ 336 million! A second innovation introduced later concerns the advertising business model. It consists of displaying sponsored links corresponding to the search. It disrupts the usual model used in the Web, which solded advertising space on the search engine portal and display advertising banners. Google has also invested heavily in machine translation since 2007 and has developed machine learning programs. They use the multitude documents that have been translated manually as a machine learning base. This translation tools wide the scope of research and make possible interrogations in different languages. In 2010, more than 50 different languages ​​were integrated into Google. The latest innovations include deep learning, and neural networks. The results are convincing, especially for image recognition and machine translation.

💲 Business model: Since 2000, the Google  audience has been monetized by advertising; Google includes sponsored links who look like answers above the  search engine answers. They direct trafic to advertisers. Their content is relevant the search for information and their format is similar to the answers. In return, there are no traditional advertises. Google has like this a very clean graphical design. Adwords is the Google's advertising solution which consists of displaying sponsored links corresponding to the search. Advertising targeting is based on keywords provided by advertisers. Thereafter, geolocation criteria were used to target the local advertising market. In 2003, the Pay-Per-Click model is introduced. The advertiser pays only when the user clicks on the sponsored link, which allows him to display information at a lower cost. Advertisers compete through an automated bidding system. The algorithm realized by Eric Veach is innovative and is derived from the Vickrey bidding model, where the winner pays not the price of his offer but the price of the bid ranked second. This encourages advertisers to make proposals at their "fair price". A subtlety of this algorithm is to determine the winner of the auction by taking into account both the amount of the offer and  the advertisement (its rank) interest in the search of information. With this model, Google would have gained 1.3 times more than Yahoo's in advertising investments. This algorithm was patented, filed in January 2003 and published in 2008.

📈 Development: At its launch in 1998, the engine was for a short time used exclusively inside the Stanford University campus located in Silicon Valley. The notoriety of the engine quickly surpassed this campus. This search engine, originally called BackRub, is renamed Google in reference to the term "Googol" which designates the number 10100. L. Page and S. Brin tried to sell their innovation. AltaVista refused the proposal for the modest sum of $ 1 million and Excites that at $ 750,000. Google's success was fast. It took the leadership of search engines from 2000 and became the search engine by default of the Netscape web browser. With its innovation on the auction algorithm, Google was gradually taking precedence over Yahoo. Internet service provider AOL, which was a major player in the United States, opted for it; that was reinforced Google's position to the detriment of the advertising solution Ouverture, bought by Yahoo. Google then launched AdSense to extend its targeted publications to other websites. Alphabet now includes all Google's products (the search engine, YouTube, Gmail, etc.). By 2015, Alphabet's net income exceeded $ 15 billion and the Company has still growing. More than 180 companies have been acquired since the creation of Google. In 2016, it became the world's largest capitalized company with a rating of $ 520 billion.

BEFORE GOOGLE
The creation of search engines was at the begining developped by a couple of dozen of entrepreneurs. The first search engines appeared in 1993, they had for name: JumpStation, World Wide Web Worm and RBSE Spider. The existence of these sites has been rather ephemeral. The functional principles were in large part present, but technical elements allowing large scale deployment were still lacking. The first successes appeared the following year, in 1994, with WebCrawler and Lycos. The advertising gains of the time on the Web were still quite limited, one of the difficulties was to operate this type of service with a cost less than 1 to 2 cents of dollars per request. WebCrawler and Lycos shared the market for 3 years, with Infoseek, Excite, AltaVista, Inktomi and Yahoo.

In France, Christophe Dupont founded the Écho search engine in September 1996 at Sophia Antipollis. Rebranded Voilà in 1998, it was the second search engine in France behind Yahoo. France Telecom, with its subsidiary Wanadoo, acquired part of the shares in 1997 and then all of it in 2001. The development and running of this French search engine will continue until 2016 on Francophone web sites, in addition to a partnership With Google.

Let's look at the situation in more detail!

WebCrawler
WebCrawler was created in January 1994 by Brian Pinkerton in the scope of a Ph.D. B. Pinkerton prioritized the development of the product to the detriment of the publication of his thesis which will wait until the year 2000. Its approach of valorization follows a classical scientific approach. In 1994 he published a scientific paper in the "3w" conference. The announcement of WebCrawler opening was done in June 1994 in a list of news "Newsgroups: comp.infosystems.announce". It became like this the first search engine widely known. This search engine was acquired by the American Internet access provider AOL in June 1995 and then by Excite in April 1997. Its operation was rapidly supported by advertising revenues.

• SERIAL ENTREPRENEUR: Brian Pinkerton is an American. He has a Ph.D. in Computer Science from the University of Washington. Founder of the Webcrawler search engine, he went through NexT, a company of Steve Jobs that realized the "cube" machine combining the spirit of Apple with the opening Unix. B. Pinkerton was latter the director of A9.com, the "Research and Advertising" division of Amazon. It developed Amazon's search technologies, and produced and operated Amazon's Cloud services. The Amazon engine is number one in product search. B. Pinkerton becomes the CTO of the Chan Zuckerberg Foundation in 2016.

💡 Innovation: WebCrawler was the first search engine indexing the full text of web pages. The business model was based on advertising revenue. It was the first search engine quite widely known.

Lycos
Lycos was developed at Carnegie-Mellon University from May 1994 by a research team led by Michael L. Mauldin, Doctor in Artificial Intelligence. The company's initial funding was $ 2 million. The company was not runed by the creator of the engine, which only acted as scientific director. Intellectual property is shared between Carnegie-Mellon University and the inventors with a rare equity. The business model was supported by advertising. The growth of its web coverage was very rapid: it indexed 1.5 million pages in January 1995 and 60 million pages in November 1996. Lycos thus had the most extensive indexing at the end of 1996. It was also the most used search engine on the Netscape Internet browser, while it was not its default search engine. Lycos next proceeded to multiple acquisitions for tens of millions of dollars in order to become a general portal like Yahoo integrating email and news services in addition to search facilities. In 1999, it was the most visited website. In May 2000, Terra Networks, a subsidiary of Telephonica, bought Lycos for $ 5.4 billion and created Terra Lycos. This serach engine will be resold several times. Its European subsidiary, which was independent, has been very successful. In 2001, it has more than 1,300 employees and generated revenues of € 139 million. In France, it became the second web site behind Wanadoo. Lycos Europe, which became a loss-making company, was arrested in 2009. Lycos holds several patents. Those on the integration of customer satisfaction feedback have been valued nearly $ 30 million after a lawsuit against Google. In 2016, an attempt is made to revive the Lycos trademark in connection with connected objects

💡 Innovation: Lycos extents the amount of indexed pages and improves the performance

Excite
The Excite search engine was created by five students in computer science and one student in political science from Stanford University after their Master's degree. They started in February 1993 with the aim of launching an enterprise. Joe Kraus was the entrepreneur, Graham Spencer the visionary and the specialist in algorithms. They first planned to develop technologies to conduct research in large databases. The treatments were based on the statistical proximity of words, it allowed search from words that do not appear in the texts. The realization took a year. The service was started in 1994, the same year that Yahoo, also from Stanford. Excite developed a strategy competing to Yahoo, integrating a search engine, thematic news, stock quote information, weather information, communication tools. While its annual advertising revenues were in the order of $ 4.5 million, Excite acquired the Magellan search engine for $ 18 million in 1996, with the goal of strengthening its audience and positioning itself as a competitor to Yahoo. In May 1997, the Stanford Alumni letter reported on its success: "Excite is behind Yahoo, the undisputed leader in search engines." In 1999, Excite was sold to the bandwidth provider Home.com for $ 6.7 billion. The service stoped two years later due to bankruptcy.

💡 Innovation: Excite extends the searches in using statistical proximities between the semantics of the searched words and other words. Their innovation is reinforced by the filing of a patent.

#6 Infoseek
Infoseek was founded in 1994 by Steve Kirsch, who in the past had created several companies on technological innovations. His idea was to transpose to the Internet the principles of an information retrieval tool, the Computer Library, provided on a CD. This software allowed access to commercial transaction data of 300 computers. The creation of Infoseek was not due to chance. S. Kirsh anticipated the interest of such an invention. He recruited a dozen people and invested several million dollars in the business to realize a search engine.

💡 Innovation: The system introduces many operating mechanisms that are still relevant, such as cache systems or audience measurement. That is the first to sell advertising to CPM (cost per thousand page views); with its Ultramatch algorithm, it is also the first to make targeted advertising based on internaut web browsing. Several patents are filed, including a patent filed in 1995 and issued in 1997 prior to Excite's patent. A patent also covers a redirection system that is still widely used by websites to measure audiences and count clicks on links. The system integrates other innovations: pre-calculating frequently requested queries and storing response elements to these queries in cache systems. Caches are storage spaces that have fast response times.

💲 Business model: At the beginning, it was a paid service: the price was fixed at 10 cents per question and a package of 10 dollars per month was also proposed. This business model was abandoned in August 1994. Ultimately, the operational cost will be supported by advertising. The development of Infoseek was accelerated in 1995 following an agreement with the Netscape browser which it became its default engine. In September 1996, Infoseek had 7.3 million visitors per month. It offered free web page hosting with no limit to file storage. An Infoseek engineer, Robin Li, will later found Baidu the Chinese search engine. Infoseek was acquired by Disney partially in 1998 and then fully in 1999, for an amount in shares of $ 2.5 billion.

AltaVista 
AltaVista was developed at DEC's Palo Alto laboratory in the spring of 1995 to demonstrate that the DEC's Alpha 8400 server was more powerful than concurrent servers. It was realized by researchers, Louis Monier (PhD at the University of Orsay in 1980) and Michael Burrows (PhD at Cambridge University, on security, distributed computing). AltaVista supported high traffic and allowed queries with "and, or, negation" logical operators. In August 1995, it processed 10 million pages. In 1996, AltaVista produced search results exclusively for Yahoo. At the end of 2002, AltaVista was the first engine to extend its research to multimedia: images, audios and videos. He was the first to introduce multilingual research. L. Monier and Mr. Burrows joined Google. AltaVista encountered financial difficulties in 2002 with the burst of the Internet bubble and the slowdown in the advertising market on the Web. He was officially arrested in 2013, ten years after the end of his success.

💡 Innovation: The system supports big traffic; It allows the expression of queries with logical operators and multimedia searches.

Inktomi 
Inktomi is not well known to the general public. It was founded in February 1996 at the University of Berkeley by Eric Brewer, assistant professor, and Paul Gauthier, student. Both are involved in a government-funded project on the computing distribution on networked clusters of computers. In this context, they found a company to value their search engine. Inktomi temporarily became the Yahoo search engine to replace AltaVista. It will later be bought back by Yahoo.

OUTSOURCING OF SEARCH ENGINE
In the 1990s, all main internet actors did not perceive the strategic issues of search engines. Developing and exploiting a search engine are far from simple, some players will outsource them. This was the case of Yahoo and Microsoft.

Yahoo!
Yahoo was founded in 1994 by Jerry Yang, a graduate of Stanford University, and David Filo. Their goal was to build the first Internet directory. Initially, this directory of websites was called Jerry's Guide to the World Wide Web. Yahoo quickly popularized the concept of a Web access portal grouping together editorial contents, email and search services. The positioning of Yahoo was singular; Yahoo integrated a traditional approach based on a manual categorization of websites to an automated indexing. Thus the construction of the directory was not totally disruptive to what was done by the Minitel directory in France. A team maintained a tree-based categorization of websites. The answers were, in the case of Yahoo, completed by the use of an automated search engine. Combining manual and automated processing allowed Yahoo during few years to provide better answers before being overtaken by Google. In 2002, Yahoo abandoned its manual approach. During a couple of years, It will next outsourced its search engine service to various successive partners: AltaVista, Inktomi, Google. Given the growing business impact of the search engine on global audience, Yahoo changed its strategy. It bought Inktomi at the end of 2002 and Overture in 2003 to have its own solution. Overture was acquired a year earlier by AltaVista.

In 2005, Yahoo dominated the Internet; it generated revenues of $ 5.2 billion, a net profit of $ 1.9 billion and employed 11,000 people. But Yahoo, unlike Google, did not take a real advantage of the development of sponsored links. In 2008, Yahoo outsourced to Bing (Microsoft) for ten years its search engine service.

💡 Innovation: The portal has been the reference as integrated web portal: that includes free services such as email, search engine and editorial content to maximize advertising revenue.

MSN Search, Window search, Live search, Bing
In 1995, Microsoft was positioned as a provider of Internet access; it developed MSN Search its first search engine. Thereafter, from 1998, this service was outsourced to various partners: Inktomi in 1998, AltaVista in 1999 and later Looksmart. It adopted a new strategic in 2004; considering the relevance of the business of Google, Microsoft developed again its own technology. Several versions are launched: Live Search in 2006, Bing in 2009.

The performance of the Bing engine is far behind that of Google. However, Microsoft is still present at the commercial and provides its engine to partners. In 2016, AOL chooses it as a research and advertising solution for the next ten years. In this business, Bing would have 20% of the market share in the United States.

WIKIPEDIA REINVENTED THE ENCYCLOPEDIA
The aim of encyclopedias is to make the rounds of knowledge and disseminate knowledge. Until the early 2000s, encyclopedias, like the Encyclopedia Britannica, were established in a pattern set up in the 18th century. The realization of the Encyclopedia by Diderot in 1750-1770 illustrated this process well. The edition is based on a community of experts and a publishing committee that review rigorously the contents. The access of knowledge become available for a fee. Wikipedia will revolutionize this scheme: access to encyclopedic knowledge becomes free, the production of content relies on a very large community of contributors which itself support the validation and revision of the contents.

It is interesting to have a few orders of magnitude in mind. The realization of the Diderot Encyclopedia takes twenty years. It includes 76,000 articles provided by 134 authors. In 1772, just over 4,000 copies were sold by subscription; About 25,000 copies were distributed from 1751 to 1782 - which for the time was remarkable. The purchase price was then around  the equivalent of 6 months' salary of a manual worker. By way of comparison, in 2015, Wikipedia provided 4,500,000 articles (60 times more) realized by 100,000 to 130,000 contributors (almost 1,000 times more). Each month, 500 million different people consulted Wikipedia for free.

#7 Wikipedia, the free encyclopedia
The idea of creating an online encyclopedia whose contents are produced by the users themselves was already the subject of discussions in Newsgroup forums in 1993 around the Interpedia concept. Wikipedia was created in January 2001 by Jimmy Wales and Larry Sanger. Their aim was to apply the software strategy of open source, in vogue since the 1990s, to the realization of an online encyclopedia. The watchwords were: a collaborative production, open, free to use. One year before, J. Wales launched Nupedia, a free-access encyclopedia, supported by a classic publishing process based on an editorial board. L. Sanger was the editor-in-chief. Only 25 items were produced in one year. The process of reviewing articles and selectively recruiting contributors slowed down their production. Wikipedia was created to broaden the community of content producers and to eliminate the centralized supervision. It was based on Wiki, an emerging collaborative technology. Investment was focused on contents generation and integration. To do this, Wikipedia did not apply to develop a new dedicated computer technology; It reused existing MediaWiki open source technology. Unlike the competing GNUpedia project, Wikipedia is consequently focused on service. At first, Wikipedia was perceived as a content production source for Nupedia, the contents have to be submitted to the Nupedia committee for approval. The community created in the context of Nupedia facilitated the launch of Wikipedia. Quickly the production of Wikipedia contents became too prolific for a centralized control. In May 2001 Wikipedia already has 3,900 contents; Nupedia was supplanted by Wikipedia.

💡 Innovation: Wikipedia has created a new editing process: the content validity and the reviewing are accomplished directly by the community of content publishers; the centralized supervision is thus limited; it provides recommendations, enacts editorial guideline and takes part in case of repeated fraudulent actions intended to falsify content whose validity is recognized. Anyone can correct founded errors. This becomes a citizen's initiative. The organization of contributors remains the open source communities. There are different layers like in open source: a heart comprising a few hundred very active people, to which are added a few thousand people who produce about a hundred editions per month, and then about 10 000 people who make between five and ten contributions per month, tens of thousands of casual publishers.

• SERIAL ENTREPRENEUR: Jimmy Donal Wales founded the Wikipedia encyclopedia in 2001 and  the Wikia Company in 2004. He was educated in a Montessori school. He started a PhD in finance and made a fortune through stock market investments. Before founding Wikipedia, he co-founded Bomis.com, a search engine for adult erotic content in 1996, and then the free online encyclopedia Nupedia in 2000.

💲 Business model:  Nupedia and Wikipedia in their beginnings were financed and hosted by Bomis Company. The mix of genres is a bit strange, because Bomis.com is an adult site. When creating Wikipedia, advertising funding is not excluded, hence the creation of a commercial site "Wikipedia.com" at the same time as Wikipedia.org. Wikipedia will however be developed without advertising. Following the stop of Bomis support, due to the bursting of the Internet bubble in 2002, a new business model was established. Since 2003, Wikipedia.org has been supported by the creation of the nonprofit Wikimedia Foundation.

📈 Development: At the end of 2004, the success of Wikipedia was already proven. The encyclopedia provided a million articles for a cost of realization 100 times less than that of a dictionary with encyclopedic vocation. The control and the effectiveness of the validity of the contents by the contributors, Wikipedians, was effective. The quality and precision of content was growing with the widening of the community of contributors and readers. Since 2012, Wikipedia has oscillated between 5th and 7th place of the most visited sites in the world. One million donors around the world subsidize the service. The budget for operating this service has been approximately $ 60 million to $ 80 million per year, which is comparable to the cost of operating commercial web portals with a broad audience. It is a very good example of content production by users, with more than 100,000 active publishers each month. Operating costs remain measured. The company had only 125 employees in 2012 to support and manage the infrastructure. In the years 2010, The Wikipedia Zero program has established partnerships with mobile operators to give free access to Wikipedia via a mobile phone in developing countries. Combined with the fact that Wikipedia is available in more than 200 different languages, this gives unprecedented access to encyclopedic knowledge in these countries. 75 operators in 57 countries offer this facility, which concerns 600 million people. Wikipedia content is also displayed by the Google engine on the front page of user searches. Undoubtedly, it is a success!

FROM ADDRESS BOOKS TO YELLOW PAGES
In the case of Wikipedia, disruptions and inheritances of past innovations are easy to discern. The guide lines of this service are the same as that of classical encyclopedias; Its accessibility is expanded and the articles are multiplied; The production and maintenance processes are completely revised.

For search engines, we are rather in the rare case of the creation of a new service in its own right. Indirectly, however, there are relationships with innovations of the past. Yahoo illustrates quite well the transition between yellow pages type directories maintained by a dedicated team and search engines. Yahoo also learned to its expense the surpassing of the first by the second. The directories have been an innovative mediation tool in the past. Innitialy, when they were not necessarily updated regularly and even less annually, they were called: Address Book, Almanac Trade and City Directory.

Address book, Almanac of commerce in France in 18th century
The practical book of addresses of Paris ‘le livre commode des adresses de Paris), edited in1692 by Nicolas de Blegny, is the first edition of a directory of professional addresses in France. It did not constitute a service for lack of updating and maintenance over the years. In the absence of updates, after a few decades, the usefulness of this type of directory decreased. The book lists the addresses of administrations, doctors, surgeons, nursing homes in Paris. It also refers to colleges, masters in art, mathematics and letters. The number of addresses is quite limited. A few well-informed persons were probably sufficient to establish it. Its reissue by Edouard Fournier is interesting, because it introduces it in retracing the evolution of intelligence offerings of this type since Antiquity. This is interesting for an historical analysis of innovation!

An annual Almanac of the Paris trade was realized one hundred years later, in 1798. The opinion published in the introduction gives us details on the stakes and the context of its realization. "The communes of Lyons, Marseilles, Rouen and Bordeaux each have an Almanac of commerce ... This work was lacking in Paris ... The public itself has expressed the desire to enjoy such a product ...". It contains the names, addresses of all kinds of professions and business (horse rental, macaroni merchant ...). The dimension of the book is without comparison with the address book realized a century before. The inventory of addresses has grown, and it has been systematized from the publication of a statement of the roles of patents of the year VI (year VI in the french revolutionar calendar). Before the Revolution, this information, which may contain certain privileges, was not public. In the book, people are also asked to communicate the changes that have occurred in their activity and the address of their institution.

CITY DIRECTORY IN UK
In England, it is nearly the same story with a few decades ahead. The formation of the Common Wealth in 1650 and the exchanges with foreign countries developed the demand for directories facilitating trade. Cities in the United Kingdom with foreign trade were the first to have a trade directory.

The Little London Directory is a register published in 1677. It is a list of names and addresses of professionals classified in alphabetical order, without specifying the occupation. It can be used to find the address of a recommended professional. The editions of this type of work multiplied in the following century.

#8 The Kent Directory
London has a directory from 1732, 60 years before Paris. The Kent Directory was published from 1732 to 1828. It was sold 1 shilling in 1740, the equivalent of 5 pounds sterling these days. This directory was founded by James Brown. It was put into operation by Henry Kent, of which he will ensure the fortune. The Kent Directory information is not exhaustive. Thus in the 1800s, the directory only covers 20% of the addresses of a competing directory. The operational modes for establishing and maintaining it are unfortunately not reported. The first English directory covering several cities appeared in 1781; the first to include residents in 1792. Note that Scotland also had directories from the 1770s, with a ranking by names and occupations.

• SERIAL ENTREPRENEUR: James Brown (1709-1788) trades through Russia and Persia. He is fluent in several languages: Turkish, Latin, Greek, Italian and Spanish. He created the first directory of London which became the Kent Directory when it was put into operation by Henry Kent.

The online scanning of Kent Directory (see "Your Internet searches" at the end of this chapter) gives a clear idea of ​​their content. Thus the first page of the 39th edition, dated 1771, indicates that the yearbook includes "the alphabetical list of names and places, directors of enterprises, persons in public affairs, merchants and other eminent merchants In the cities of London and Westminster ... also lists of Lords ... to which is added a correct list of all the bankers ". The directory is printed and sold by Henry Kent and distributed in the bookstores and brochures shops of London and Westminster.

The mail distribution requires a good knowledge of the addresses; it is not surprising to see appear a directory of Posts. The first edition in England dates from 1800 and is entitled The New Annual Directory. It is a draft, with only 250 addresses. It became The Post Office Office Directory in 1801 and Post Office London Directory in the mid-1810s. Frederic Festus Kelly, Post Inspector, is the creator of this directory is. Benjamin Critchett became his associate in 1803. In 1839, the directory contained 1187 pages, one third of which were advertisements. The directory is not innovative on the principles of its design in comparison to the Kent Directory. However, its accessibility and coverage are more important. The directory extends to other cities from 1845. It will benefit from a wide distribution. The business model is based not only on sales but also on advertising. The price is 1 pound in 1850, or 45 pounds sterling present.

DIRECTORY PUBLISHERS IN THE USA
In New York, New Amsterdam at that time, Dutch magistrates established as early as 1665 an initial list of residents of the streets. The first real directories were published more than a century later, in 1785 in Philadelphia, then in Boston and New York. They mention the name, occupation and address of the inhabitants. A few other restricted initiatives preceded them in Baltimore and Charleston.

The first directory publishers in the United States done it in addition to their current business: master of post, commercial business, newspaper publishers. Local newspaper publishers, whose activities are already related to printing, have operational facilities for the production of directories; they are interested in knowing the residents who constitute their clientele. By the mid-1800s, full-time directories publishers appeared. They realize the directories of several cities, with various release dates to smooth the workload over the year. These companies are gaining momentum. To capture more subscribers, publishers add new information and the organization of the directory is evolving, thus providing a better service.

Yellow Pages and phone books
The telephone directories, the Yellow Pages, are very much inspired by the City Directories and the Post Office directories. The business model is based on the sale of advertising space. In England, a phone book appeared in 1881, one year after deploying the phone. There are several telephone operators, but their telephone services are not yet interoperable.

In the United States, the first repertory was created in Connecticut in 1878, two years after the invention of the telephone. This is a simple list of 47 subscribers. Ruben H. Donnelley, in 1886, categorizes phone numbers by types of activity and locations. It develops the sale of advertising space and contracts with Bell System twenty years later.

#9 L.M. Berry and Company
Loren Murphy Berry creates an agency in charge of the sale of advertising space of the yellow pages. The idea takes shape in 1909 in a city in Indiana, having valued the creation of a directory by selling advertising for the telephone company owned by his uncle. It gradually extends the offer to other cities in that state, each with its own local telephone service, and then to Ohio, a neighboring state. In 1921, the telecommunications market legislation was amended. It authorizes the national consolidation of this market, hitherto not allowed. The interest of national telephone directories increases all the more. In 1930, the Berry directories reached a million copies. Between 1945 and 1950, Bell doubled its subscribers. It deploys 14 million phones. For its part, the company Berry is extending its business and now covers not only the management of advertising space but also the publishing and publication of directories. The business was expanded internationally from 1966, following a partnership with ITT (International Telephone and Telegraph). Berry provides a common directory to the four operators in the United States and extends its service to Europe. The Yellow Pages monopoly in the United States will continue until the spin-off of AT & T in 1984. AT & T revenues from Yellow Pages were $ 5 billion a year. During the split, the seven US operators recover the Yellow Pages directories. They have an obligation to provide minimal access to data. Berry was purchased in 1986 by BellSouth Corp. for several hundred million dollars to extend its Yellow Pages service while it is present in 31 states of the United States. The subsidiary became Berry Company; it is responsible for more than 800 national directories and serves more than 100 US telephone companies and one million advertisers.

• SERIAL ENTREPRENEUR: Lorren Murphy Berry (1888-1980) is known as "Mr Yellow Pages". He created a regional, national and then international company specialized in the sale of advertising space and the production of Yellow Pages.

WORD OF INNOVATOR - Steve Kirsh: the ingenious founder of Infoseek
Steve Kirsch, American, born in 1956, graduated from a master's degree in computer science at MIT. Ingenious, he has several successes to his credit. He founded 6 companies on subjects that became very promising. He created an optical mouse system in 1982 during his Master at MIT; it will propose it to Apple, before setting up its own company to produce and distribute the invention; the company sold $ 12 million in 1988. It co-founded in 1986, several years before Microsoft's Word solution, a company on the FrameMaker suite Editor, competing Macwritte in the years 1986. The company is bought by Adobe in 1995 for $ 500 million. He created the search engine Infoseek in 1994 bought in 1999 by Disney for some 2.5 billion; He founded Propel in November 1999, which aims to develop a standard e-commerce solution with the slogan "put Amazon in a box". It then created the anti-spam Abaca in 2000 sold in 2013, then One Id in June 2011 a single authentication solution that replaces passwords with elliptical cryptographic curves; He develops a Cointrust payment service. All of its businesses represent a capitalization of $ 6 billion.

While you were a teenager, you went to the the Vint Cerf lab at UCLA and got a computer account and create. I would like to come back to your motivations at the time, were you so fascinated by computers to dare this approach? How did it go?

I had an introduction to computers in grade 6 at elementary school and found it fascinating; So I tried to learn all I could about the subject. UCLA was near my home and I was lucky to get in touch with the right people while I was in high school.

You are rather (on) gifted. You bounce back from solving one problem to another, your need to create seems insatiable. Were you a particularly brilliant child at school?

Yes, I always got good ratings (especially A) and I skipped a class. I was the best in class at mathematical level.

Google is now almost in a monopoly situation. What enabled Google to dominate InfoSeek is it its page rank algorithm? Is it its business model based on sponsored links?

Concentrate on the search engine. Very early we learned that there were 2 things that customers wanted: a quick response, and finding what they were looking for. Unfortunately, the people we recruited felt that we had to limit the search database to the 10 Million most interesting and useful web pages and spend resources in other developments than the search engine. That did not sound like the right strategy at the time; Retrospectively, I was right. However, I am glad that Google has succeeded; they have done an enormous job.

Innovation in business models seems to be a key element in establishing a success story, does it?

Yes, you need both a good idea and a way to put it on the market and monetize it. The adequacy between the product and the market is essential. If you have it, you can make many mistakes and continued to succeed.

About your last two innovations OneId and Cointrust where are you at? Are these successes in the making?

Yes I think so. Cointrust was renamed Token. OneID is about to be acquired by a large company.

INTERNET SEARCHES TO FIND OUT MORE
With simple Google search find the following Expressions with and without quotes:

• Zator Technical Bulletin No. 48
•  “The Anatomy of a Large-Scale hypertextuels Web Search Engine”
•  “keyword auctions”
• “searching the world wide web IEEE”
•  “Kent’s directory” 1759, 1774
• "WebCrawler history", “WebCrawler: Finding What People Want”
• "father of the search engine" (BBC)
• "Why did Wikipedia succeed" (nieman lab), “early history of nupedia”
• Wikimedia Foundation Annual Report
•  “advertising; Donnelley’s Yellow Pages” (the new York times)
• “City Directories: The Phone Book Before the Telephone” (Kenton library)
•  “Short History of Early Search Engines”
• History of L. M. Berry and Company
• Steve Kirsch ieee spectrum
• Google data center
• patent 6285999 (Lawrence Page,  Stanford), patent 7349876 (Eric Veach, Google), patent 6314420, 6775664 (Lycos), patent 5659732 (Steve Kirsh, Infoseek,), patent 5915249   (Graham Spencer, Excite)

At Wikipedia.org:

•  “Encyclopedia”

At Archive.org:

• London directory of 1677.
• History of English journalism 1908; Henry Walker.

At Alumni.stanford.edu:

• Excite

At LinkedIn:

• Steve Kirsch (profil)