3
The User Experience

IMPACT ON BUSINESS IN THE IMMEDIATE AREA

Wall Street-area financial institutions were of course significantly affected by the September 11 attacks, though in varying degrees. Some companies were severely hurt—Cantor Fitzgerald, for example, lost a majority of its employees and all of its facilities in the World Trade Center. Other firms suffered primarily from the loss of their physical offices. At Morgan Stanley, most employees escaped the area before the Twin Towers collapsed, but the firm’s offices, along with all the information technology (IT) equipment in place there, were completely destroyed.

Other companies located near the World Trade Center, such as Merrill Lynch and Lehman Brothers Holdings, Inc., also could be categorized as “headquarters rendered unusable”;1 and a great deal of additional space in the areas abutting Ground Zero was rendered either temporarily or permanently out of commission. The New York Stock Exchange itself was shut down for almost a week, in part because many firms did not have the communications capability for completing trades and in part because the Exchange had communications and physical-damage issues of its own to contend with.

Some financial firms faced power and communications disruptions even though their office space and IT infrastructure had not been directly

1  

Randall Smith. 2001. “At Morgan Stanley, Readiness Saved Lives,” Wall Street Journal, September 14, p. C1.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 40
3 The User Experience IMPACT ON BUSINESS IN THE IMMEDIATE AREA Wall Street-area financial institutions were of course significantly affected by the September 11 attacks, though in varying degrees. Some companies were severely hurt—Cantor Fitzgerald, for example, lost a majority of its employees and all of its facilities in the World Trade Center. Other firms suffered primarily from the loss of their physical offices. At Morgan Stanley, most employees escaped the area before the Twin Towers collapsed, but the firm’s offices, along with all the information technology (IT) equipment in place there, were completely destroyed. Other companies located near the World Trade Center, such as Merrill Lynch and Lehman Brothers Holdings, Inc., also could be categorized as “headquarters rendered unusable”;1 and a great deal of additional space in the areas abutting Ground Zero was rendered either temporarily or permanently out of commission. The New York Stock Exchange itself was shut down for almost a week, in part because many firms did not have the communications capability for completing trades and in part because the Exchange had communications and physical-damage issues of its own to contend with. Some financial firms faced power and communications disruptions even though their office space and IT infrastructure had not been directly 1   Randall Smith. 2001. “At Morgan Stanley, Readiness Saved Lives,” Wall Street Journal, September 14, p. C1.

OCR for page 40
damaged. Most significant for companies in Lower Manhattan, the collapse of Building 7 of the World Trade Center and the consequent damage to Verizon’s central office across the street disrupted voice and data lines that linked Wall Street to the world.2 Another nearby Verizon office that served the New York Stock Exchange was also affected, with 20 percent of its high-speed data lines “out of action” and the rest “operating only sporadically.”3 Some local firms reported not being seriously affected. For example, the director of infrastructure at Blackwood Trading LLC was quoted as saying that “if he hadn’t seen the attack, he wouldn’t have known it happened.” Blackwood was relatively well prepared for such a disaster: while its data center is housed on Wall Street, the firm backs up all its trade data to remote centers in New Jersey, which is on a separate power grid; it was able to execute more than a “half-million trades before the NASDAQ voluntarily shut down,” according to that executive.4 Indeed, data loss was less of a problem than one might think. Most large Wall Street firms had responded to the earlier World Trade Center bombing (in 1993) by focusing their attention on crisis management,5 which resulted in the institution of thorough data-backup or co-location procedures. Cantor Fitzgerald’s eSpeed, for example, had mirrored the firm’s entire operations at other sites.6 PEOPLE ON THE NET Data on people’s usage of the Internet following the terrorist attacks came from a variety of sources. Probably the most detailed information available to the committee was from the Pew Internet and American Life 2   Shawn Tully. 2001. “Rebuilding Wall Street,” Fortune, Vol. 144, No. 6: pp. 92-100. Available online at <http://www.fortune.com/indexw.jhtml?channel=artcol.jhtml&doc_id=204166>. 3   Emily Thornton et al. 2001. “The View from Ground Zero,” Business Week Online, September 13. Available online at <http://www.businessweek.com/bwdaily/dnflash/sep2001/nf20010913_005.htm>. 4   Mark Hall and Lucas Mearian. 2001. “IT Focus Turns to Disaster Recovery,” IDG, September 11. Available online at <http://www.cnn.com/2001/TECH/industry/09/11/disaster.recovery.idg/index.html>. 5   Shawn Tully. 2001. “Rebuilding Wall Street,” Fortune, Vol. 144, No. 6: pp. 92-100. Available online at <http://www.fortune.com/indexw.jhtml?channel=artcol.jhtml&doc_id=204166>. 6   Edward Cone and Sean Gallagher. 2001. “Cantor Fitzgerald—Forty-Seven Hours,” Baseline, October 29. Available online at <http://www.baselinemag.com/article/0,3658,apn=2&s=2101&a=17022&ap=1,00.asp>.

OCR for page 40
Project, which surveyed Internet users immediately after September 11.7 The project had been conducting telephone surveys of Internet users for some time before the crisis, and it continues surveying users even now; thus its data not only present a picture of how users behaved that day but also allow comparison with their behavior both before and after the attack. Other useful sources of data on user behavior included Web-usage measurements from Webhancer, search statistics from major search sites such as AOL and Google, and data from content providers such as CNN.com and Akamai. Together, these data provide a very telling portrait of what people wanted, needed, and expected from the Internet in those extraordinary circumstances. The Internet as a Source of News Many people learned of the terrorist attacks on the World Trade Center and the Pentagon while they were at work or on their way to work.8 And because people often do not have access to television sets at their place of work, there is reason to believe that they then turned to Internet news sites for information. In what is sometimes referred to as a “flash crowd”9 event, national and international demand for timely information soared, and many news Web servers—those of CNN, MSNBC, and the New York Times, for example—experienced unprecedented loads. An anecdote regarding CNN’s Web site <www.cnn.com> gives a vivid example of just how fast the demand for Internet-accessible news grew. When the director of the facility saw on TV that the second plane had just struck, he stood up in his cubicle and shouted to other staff members to take steps (such as bringing extra servers online) to prepare for an increased demand for news. By the time he sat down, that spike had already arrived. (Box 3.1 discusses the CNN experience, including steps that the network took to keep up with demand, in more detail.) 7   Lee Raime and Bente Kalsnes. 2001. The Commons of the Tragedy: How the Internet Was Used by Millions After the Terror Attacks to Grieve, Console, Share News, and Debate the Country’s Response. Pew Internet & American Life Project, Washington, D.C., October 10. Available online at <http://www.pewinternet.org/reports/toc.asp?Report=46>. 8   While it is difficult to get a precise estimate, data from the Census Bureau suggest that almost half of the U.S. workforce was at work or en route when the planes hit. Information available online at <http://www.bls.gov/news_release/flex.t07.htm>. 9   The term “flash crowd” was coined by science fiction writer Larry Niven, who wrote a short story by that title about masses of people teleporting to see exciting events they see reported in the news.

OCR for page 40
During the remainder of the month after September 11, the number of Internet users who sought to get news online increased by about 25 percent, even though Internet use for some other purposes (such as shopping or sending e-mail) declined. Indeed, survey data indicate that the total number of Internet users declined by about 10 percent in the week immediately following September 11 (see Table 3.1). Even given the surge in demand for online news, all the evidence is that Internet users, in the same proportion as the general population, preferred to get their news from television. A poll by the Pew Project showed that in the week after September 11, television was the main source of information for 79 percent of Americans and for 80 percent of the heaviest Internet users (see Table 3.2). Heavy Internet users relied on the Internet as much as on radio and newspapers, while Americans overall relied on the radio and newspapers far more than they depended on the Internet. One possible reason for this seeming contradiction—high online demand for news and high reliance of Internet users on television—is that once they were home from work (where they relied largely on the Internet) on September 11, most people turned on their television sets and got the latest news without having to go online for further information. Another possible reason was frustration with the Internet: 43 percent of Internet users reported at least some trouble accessing Web sites in the first hours after the attacks, and 15 percent reported great difficulties.10 Yet another possible reason is that news organizations generally do not provide live streaming video programming.11 In the end, about a fifth of those who had difficulty reaching a site gave up on using the Internet for news during that period.12 Another important point is that many people appear to have used the Internet not as a replacement for regular news sources but as a supplement. Major search engines reported that the information sought by users changed dramatically on September 11 and in the following days. For example, on September 12, a number of talk shows mentioned Nostradamus, a Renaissance writer renowned for his prophecies. Thereafter, “Nostradamus” was at the top of the list or near the top at many popular search engines; at Yahoo, for example, it was number 1.13 Google 10   Raime and Kalsnes, 2001, The Commons of the Tragedy. 11   A scalable technology known as multicast could support streaming video to large numbers of viewers, but it is not commonly employed by content providers. 12   Raime and Kalsnes, 2001, The Commons of the Tragedy. 13   Google <http://www.google.com/press/zeitgeist/9-11-search.html>; Yahoo <http://websearch.about.com/gi/dynamic/offsite.htm?site=http://buzz.yahoo.com/>; Lycos <http://websearch.about.com/gi/dynamic/offsite.htm?site=http://50.lycos.com/091101%5FSpecial.html>.

OCR for page 40
BOX 3.1 CNN.com on September 11, 2001 September 11, 2001, has certainly not been the only high-demand period experienced by Internet news sites. For example, interest in the results of the 2000 general election fueled a steep rise in demand. But September 11 set new records, and consequently the ability to reach major news Web sites that day was reduced for some people.1 To use this experience to better understand the demands on news servers during a crisis event and to identify measures that can help deal with that demand, a representative from CNN’s Internet division was invited to participate in the workshop held by the Committee on the Internet Under Crisis Conditions. Key elements of CNN’s experience follow. On September 11, CNN’s overall demand surged greatly, with the measured daily load (as expressed in page views) increasing on September 11 to 132 million— nearly 10 times the more typical load of 14 million on September 10. The measured demand of September 11 probably underestimates total user demand, however, because not all users were able to successfully load the Web page as the demand initially surged after the crash of the first airplane. The number of hits (pages or images requested) doubled every 7 minutes, resulting in an order-of-magnitude increase in less than 30 minutes. Demand for news continued to grow in the hours following the attack, with the load on September 12 reaching 304 million page views—more than twice that measured on September 11. Keeping up with demand after the first airplane crash was very challenging for the CNN operations staff, who employed a combination of several techniques to deal with the load: Reducing Web page complexity. The CNN.com main Web page was significantly reduced in size (i.e., as measured by the number of separate elements such as reported that “Nostradamus,” “cnn,” and “World Trade Center” were the top three terms among people whose search-engine usage increased during the week ending September 13.14 The Internet as a Means of Communicating Between Individuals On September 11, many people felt the need to communicate right away with family, friends, and colleagues. The purpose of these communications ranged from emergency responses (as officials in New York 14   Google. 2002. “Google Search Statistics from 9/11/01,” Google, Mountain View, Calif. Available online at < http://www.google.com/press/zeitgeist/9-11-search.html>.

OCR for page 40
headline pictures and graphical menu bars for selecting additional content), consistent with CNN’s in-place strategy for handling high-demand periods. In fact, the main page was stripped down to the bare bones—even further than the usual minimum—to increase its ability to serve pages. Indeed, at its minimum complexity, the page could fit into a single IP packet. Adding more servers. A number of other server systems are colocated with the servers assigned to CNN.com. These systems, which normally are used for other CNN and Turner Broadcasting content, were for the most part experiencing significantly reduced volume that day. Thus, a number of them were reconfigured and added to the CNN.com server pool. (Interestingly, CNN did retain server capacity for the Cartoon Network, which saw an increase in volume—likely reflecting parents’ desire to provide children with an alternative to the disturbing news.) Temporarily employing a third-party content-distribution network. CNN arranged to significantly increase its use of the Akamai content delivery network in order to reduce the load on the CNN servers themselves. That is, the CNN Web pages temporarily pointed Web browsers to retrieve images from Akamai servers instead of from the usual CNN systems. The net effect of all of these efforts was to enable overall capacity to increase over an order of magnitude within hours of the event, permitting CNN to cope with the greatly increased demand. 1   For example, according to a report in ComputerWorld, the Web-measuring company Keynote observed that the availability (responsiveness to requests to download Web pages) of the Web sites of CNN, the New York Times, ABC News, MSNBC, and USA Today were all significantly reduced following 9:00 a.m. on September 11. (Todd R. Weiss. September 11, 2001. “News Sites Simplified After Performance Bogs Down.” ComputerWorld. Available online at <www.computerworld.com/managementtopics/ebusiness/story/0,10801,63729,00.html>.) City and Washington, D.C., sought to deal with the crisis) to trying simply to make sense of what was happening. Although the Internet was one medium by which people chose to communicate, it is important to emphasize that the preferred mode of personal communications was the telephone. Indeed, even heavy Internet users reported using the telephone more than the Internet (and at a higher rate than the national average). While 63 percent of Americans phoned a family member about the attacks on September 11 or in the following days, 75 percent of heavy Internet users called a family member during that period.15 15   Raime and Kalsnes, 2001, The Commons of the Tragedy.

OCR for page 40
TABLE 3.1 Internet Use by Activity, August Through September 2001 Activity Aug. 13–Sept. 10a (percentage) Sept. 12–19b (percentage) Sept. 20–Oct. 1c (percentage) Going online for any purpose 56 51 57 Sending or reading e-mail 51 42 49 Getting news online 22 27 26 Seeking hobby information 20 10 22 Browsing for fun 20 13 20 Doing work-related research 17 13 15 Seeking medical or health information 5 3 5 Buying products 4 2 2 aN = 1,351; margin of error is ±3 percent. bN = 1,138; margin of error is ±3 percent. cN = 525; margin of error is ±6 percent. SOURCE: Lee Raime and Bente Kalsnes. 2001. The Commons of the Tragedy: How the Internet Was Used by Millions After the Terror Attacks to Grieve, Console, Share News, and Debate the Country’s Response. Pew Internet & American Life Project, Washington, D.C., October 10, p. 7. Available online at <http://www.pewinternet.org/reports/toc.asp?Report=46>. TABLE 3.2 Main Source of Information Following September 11, 2001 Main Source of Information All Americansa (percentage) Heaviest Internet Usersb (percentage) Television 79 80 Radio 7 6 Newspaper 7 7 Internet 2 6 Talking with others 2 1 aN = 1,029; margin of error is ±3 percent. bN = 260; margin of error is ±7 percent. The Pew study defines the heaviest Internet users as those who have more than 3 years’ experience online and who log on from home every day. This group constitutes about 20 percent of all Internet users and about 11 percent of the U.S. adult population SOURCE: Lee Raime and Bente Kalsnes. 2001. The Commons of the Tragedy: How the Internet Was Used by Millions After the Terror Attacks to Grieve, Console, Share News, and Debate the Country’s Response. Pew Internet & American Life Project, Washington, D.C., October 10, p. 10. Available online at <http://www.pewinternet.org/reports/toc.asp?Report=46>.

OCR for page 40
At the same time, about one-third of Americans had difficulty placing a phone call on September 11 (see “The Experiences of Other Communications Networks” in Chapter 2 for more detail on the telephone system and its performance), and about one in eight turned to the Internet to communicate with friends and loved ones. Much of the communication was through e-mail, which was used almost as soon as the attacks began, though a modest fraction of Internet users (13 percent) reported using instant messages.16 Anecdotal reports both from Washington, D.C., and New York City suggest that instant messaging proved a viable alternative for office workers who were unable to use their phones but still had Internet access. Those directly affected by the attacks also made use of Internet communications. Some people trapped at the top of the Twin Towers were able to e-mail colleagues and family.17 Some communications from the Twin Towers were from people who used wireless PDAs, such as those from Research in Motion (Blackberry), to send messages even after in-building infrastructure had been knocked out. Finally, Internet telephony provided a useful alternative communications channel for some people who had lost telephone service, though apparently the total number of such calls was small compared to those placed through the conventional telephone network. The Internet and Community In the hours and days following the attacks, a number of Web sites were created (or adapted from existing sites) to help fill various disaster-related needs. They included the following: Missing person and “I’m alive” lists. For example, Prodigy Communications created an “I’m ok” online message center to help people find information about loved ones. Relief supply requests. Solicitations for relief contributions. Companies such as Amazon.com and Yahoo used their Internet billing systems to facilitate people’s donations to the American Red Cross. 16   Raime and Kalsnes, 2001, The Commons of the Tragedy. 17   New York Times. 2002. “Fighting to Live as the Towers Died.” May 26, p. 1, col. 1.

OCR for page 40
Overall Use of the Internet Total use of the Internet declined, as discussed above. (Instances in which particular ISPs instead saw a rise in traffic levels appear to be attributable to their serving news and other content that were in higher demand on September 11.) The decrease in overall demand is apparent in both the Pew Internet users survey data and in reports of ISPs, including presenters at the workshop.