Internet services: Who's smiling now?

A joint study by McKinsey and J. P. Morgan suggests that in the United States, Internet traffic will grow at a compound annual rate of just 88 percent through 2005.

Jan. 16, 2002 3:17 p.m. PT

4 min read

The shifting mix and management of worldwide Internet traffic will make services such as Internet hosting and content distribution increasingly important and put traditional information-technology players such as AT&T, IBM and EDS back in the driver?s seat--leaving behind smaller upstarts that have specialized in such offerings.

This power shift is the result of unexpected changes in the nature and growth of Internet traffic. Web pages, which drove Internet usage in the past, now represent a rapidly diminishing share of total traffic. Consumers and corporate users will increasingly turn to rich media and to streaming live audio and video. In addition, server-to-server traffic should become the single largest category of Internet traffic in both the United States and Europe by 2005 as businesses come to rely increasingly on extranets and other kinds of Internet-based communication.

Meanwhile, the volume of Internet traffic is increasing at a slower rate, which is bad for new companies that gambled on reaching profitable scale on the back of continued explosive growth. In the United States, such traffic will grow at a compound annual rate of 88 percent through 2005, according to a study by McKinsey and J. P. Morgan. A parallel McKinsey study showed that in Europe, Internet traffic will grow by 106 percent annually during these same years. The figures are high, but in 1999 it was growing at a rate of nearly 200 percent.

Such changes in Internet traffic patterns play to the strengths of large incumbents, at the expense of the upstarts. New entrants such as hosters (which offer secure outsourcing and guaranteed connections for Internet servers) and content distribution networks (which store content at locations close to end users) helped speed the delivery of Web pages to customers by enhancing the connections between local and national telecom networks. But many new entrants lack the scale and expertise to offer large businesses help in managing streaming media and the applications (such as databases and systems integration) that are now moving onto the Internet.

These are precisely the kinds of services that traditional IT companies such as EDS and IBM have long offered their corporate clients. Moving into hosting non-Web page applications online will allow IT incumbents, by making mostly incremental investments, to strengthen their customer relationships and to take a share of the growing revenue.

Doing so will be worthwhile: We expect the revenue of the U.S. hosting industry to grow to $30.6 billion by 2005, from $4.5 billion last year, and the revenue of the European industry to reach $16.7 billion.

The advantage in content-distribution services will shift in a similar way. The wide-ranging and stable networks and deep pockets of the long-distance carriers, as well as their existing relationships with large companies, will bolster their efforts to carry data and streaming media for their corporate clients. By 2005, when these traffic shifts have taken hold, long-distance incumbents such as AT&T, British Telecommunications and WorldCom and its UUNet subsidiary are likely to have deprived current players such as Akamai Technologies of more than half of the global $9.5 billion in revenue from content distribution.

Changes in Internet traffic patterns play to the strengths of large incumbents, at the expense of the upstarts. Falling prices for bandwidth and leased optical fiber--the result of fierce competition and, in the United States, of an oversupply of capacity--will also hurt the newer providers more than they will harm the deep-pocketed incumbents. In this setting, the traditional long-haul carriers, which were once dismissed as dinosaurs, now have an opportunity to move more of their services to high-performance networks and to do so at a lower cost than anyone ever expected. U.S. local telephone companies, sometimes seen as peripheral players in the Internet economy, should benefit from the increase in broadband connections and from strong economics that are based on continued local dominance. We expect that these companies too will try to increase the depth of their customer relationships and to gain more revenue by entering the content-distribution and hosting markets and, as regulators allow them to do so, the long-distance market as well.

Challenges certainly remain for the incumbents. Long-distance phone companies, for example, will see the revenue growth they have enjoyed from dial-up Internet access and other sources slow down. They will also have to compete against today?s hosters and content distribution networks as the latter increasingly take on a middleman role between them and their customers. But most of the large incumbents are well positioned to profit from the changing Internet landscape. For the first time in recent years, these stalwarts may be able to take a place at the head of the pack.

For more insight, go to the McKinsey Quarterly Web site.