|
By the Numbers:
How UT and Cisco Teamed Up to Track the Internet Economy
by J.B. Bird
On the morning of June 11, 1999, UT Business Professors Anitesh Barua and Andrew Whinston found themselves in unfamiliar territory for most academic researchers. After toiling for months on a joint research project with Cisco Systems, they suddenly found their work moved out of academic obscurity into the national spotlight.
Television networks, radio shows, newswires, The New York Times, The Wall Street Journal - virtually all major media outlets picked up their story. "I woke up and saw UT on every TV station," says Whinston.
The rush had just begun. Over the next two weeks, interview requests, phone calls, and emails flooded in from foreign and national publications, the White House Office of Technology, Vice President Al Gore's office, the US Senate, and scores of companies, lobbyists, research firms, and brokerage houses.
The professors say that they were not prepared for the onslaught. In hindsight they might have been, given the nature of their project: in less than four months they had pulled together the most comprehensive measurements to-date of the U.S. Internet economy.
The numbers
The measurements, called "the Internet Economy Indicators," show that the size of the U.S. Internet economy exceeded all previous estimates, producing over $300 billion in annual revenues and more than 1.2 million jobs by the end of 1998. The latest indicators, released October 27th, push these totals even higher, projecting total revenues of $507 billion by the end of 1999.
The numbers are not an exact measurement of the Internet economy, but predictors, or best estimates, drawn from surveys of over 3,000 Internet-related companies and thousands of existing Websites.
"I don't think there's anything else as broad as their study," admits Barry Bosworth, who heads up e-commerce projects at the Brookings Institution. Previous studies of the Internet, conducted by firms like International Data Corporation and Forrester Research, offered partial analyses of the online economy, but the UT/Cisco project is the first to look at the Internet holistically, as an entire economic system.
The publicity from the first set of indicators has already shifted national thinking about the Internet. More people are now aware that infrastructure and applications are vital links in the online chain - and also huge sectors in their own right, comprising over half the Internet economy.
The indicators further reveal that the U.S. Internet sector is already larger than long-standing industries like energy and telecommunications. And with total revenues doubling annually in each of the last three years, the end is nowhere in sight. "As more and more business activity moves into the Internet stream," Barua told ABC News, "You'll see an explosion. Ultimately, online transactions will go through the roof."
Texas connections
Cisco, the world's largest internet infrastructure company, sponsored development of the indicators last spring as a public service. From the start, the company felt that UT would be a neutral place to host the research.
E-commerce may currently be in vogue at schools everywhere, but Texas actually comes as close as possible to having a tradition in the field. Whinston founded UT's Center for Research in Electronic Commerce in 1994, long before Jeff Bezos had dreamed of Amazon.com and years before other business schools were jumping on the e-commerce bandwagon. Whinston was likewise ahead of his time when he pioneered an economic approach to e-business with text books like the best-selling Frontiers of E-commerce (1996) and The Economics of E-commerce (1997).
In 1997 he and Barua once again struck gold when they established the Electronic Digital Economy (EDE), a living laboratory of e-commerce where students buy and sell products over the Web, with grades and cash prizes as incentives. The EDE continues to attract international attention, most recently receiving a $300,000 research grant from the National Science Foundation.
UT's connections to the indicators extend far beyond Barua and Whinston. MBA Program Director Ramesh K. Rao was the moving force behind bringing the indicators to UT, and he brokered the final deal with Cisco. Rao collaborated with two Texas alumni, lead scientist Pinnell and lead researcher Jay Shutter, MBA 92, vice president of research at Cunningham Communications.
Shutter oversees the 25-30 people working on the indicators at any given moment, a staff that jumps to 60 during peak times. Most of the actual number-crunching takes place under Pinnell at MarketVision Research in Cincinnati. (Interestingly, Shutter and Pinnell first met through another Business School alumnus, Peter Zandan, who employed them at IntelliQuest in the early 90s.)
Given the magnitude of the indicators project, Pinnell, Shutter, Barua, and Whinston have had to adopt a divide-and-conquer approach to their work. Pinnell credits the professors with providing much of the vision and structure, "which is critical when you're working with something as amorphous as the Internet."
The project has offered more than enough challenges for the market research professionals. "Just trying to identify all of the participants in the Internet economy is daunting," says Pinnell. He points out that key players like E-bay and Amazon.com may be easy to identify, but tracking everyone that produces revenue - from small ISPs to thousands of online content providers - is far more difficult.
Companies that generate revenue both on the Web and through conventional bricks-and-mortar markets present added difficulties, notes Shutter. "It's easier to measure the e-business of pure Internet companies," he says, but hybrid companies do not always keep separate records of Internet and non-Internet revenues. "Most companies, their accounting systems aren't tracking that way," says Shutter, "So we're pretty far ahead of the curve."
Measuring needs
Since the indicators first appeared, many organizations have wanted more detailed access to the study. "A lot of companies would like to get the data," admits Whinston, "but it's not available. We collected it on the condition that it can only be aggregated. We can analyze information for research purposes, but we will not release information collected from individual companies."
Government agencies have been particularly interested in the methodology of the survey. In September the Brookings Institution hosted a conference for analysts from the Census Bureau, Department of Commerce, and Federal Reserve, so they could meet with Barua, Whinston, and colleagues to discuss their research in detail. Census and Commerce vitally need assistance as they create official government metrics for e-commerce.
"The Cisco study helped us in terms of thinking about how to think about the Internet economy," says Tom Mesenbourg, in charge of e-commerce efforts at Census. According to Mesenbourg, Census analyzes the Internet a little differently than the Cisco/UT researchers, seeing three layers instead of four; nonetheless he credits their influence. ("We've also been looking at what the Canadians are doing," he says - which is interesting, since Industry Canada is adopting the Cisco/UT model for their official analyses).
The government agencies have to track hard numbers, not best-estimate predictors. As a result their studies, requiring greater statistical controls and accuracy, will inevitably have longer lag-times to market. The government is already circulating its first e-commerce surveys, for instance, but limited results are not due until 2001, with complete measurements anticipated in 2002.
"The private sector cannot wait until 2002 for numbers," says Barua. In Internet time, one-year lags are an eternity.
Barua and his colleagues know that eventually the government will be, and should be, the central authority for reporting e-commerce statistics. Until that becomes a reality, however, private-sector initiatives are the best options. Even Mesenbourg agrees, seeing a need for "more real-time numbers" and predicting that real-time reporting on the Internet "is something government's not going to be able to do."
Time is only one of obstacles the government faces. Just as the Internet is overturning old business models and practices, it thwarts conventional methods of categorization. Government surveys, for instance, have typically assumed that certain manufacturers sell to certain wholesalers who sell to certain retailers. But what if the manufacturer, like Dell, sells direct to the consumer?
"The Internet economy is changing longstanding business relationships," says Mesenbourg, which creates "a good opportunity for government to work with the academics and private sector to understand the change."
Barua agrees that the private sector can help define things better. "Ultimately it's the government's job, but we can help."
By the time government catches up, Barua and Whinston plan to be onto the next challenges - issues like productivity, efficiency, and value-added propositions, concepts which are all changing radically on the Web. These future projects will no doubt live up to Jon Pinnell's assessment of the indicators: "Another brilliant example of UT being in the forefront of important, ground-breaking research."
For information on specific programs at the McCombs School, consult our contacts page. For media information, contact the Communications Director by phone at 512-471-3314 or by email at CommunicationsDirector@mccombs.utexas.edu.
