18.6 Moderate Speeds Over Wireless WANs

In 2000, Metricom surfaced with a nationwide advertising campaign convincing individuals to use their Ricochet wide area networks. The goal of the Ricochet system was to provide Internet access through wireless mode at moderate speed close to 100 kbps at competitive rates. The Ricochet packet radio network enabled coverage through the deployment of a large number of inexpensive packet radios on pole tops, which routed modem packets to a wired access point (WAP), which then connected to the wired Internet. Ricochet employed spread spectrum, frequency hopping technology across 160 channels in the license-free Part 15 902 to 928 MHz ISM band.

The company was founded in 1985 by Paul Baran, who in 1962 helped create the Internet. Metricom was able to attract significant capital investment, including over $500 million from MCI WorldCom and Microsoft cofounder Paul Allen, who once talked Bill Gates into dropping out of Harvard to begin a software company. Originally, Metricom provided utility companies with a way to automatically read gas and electric meters; however, the company soon changed course, choosing to offer wireless Internet access to mobile users. In 1999, the company began to upgrade its 28.8 kbps wireless network to 128 kbps speeds, more than twice as speedy as the fastest dial-up Internet connection. Metricom managed to wire 17 cities/markets, including Manhattan, San Francisco, and other major metropolitan areas, totalling about 30 million people under the coverage umbrella.

The first version of Ricochet, sold by Metricom through its Web site, operated at 28.8 kbps and cost subscribers $29.95 per month. The faster system was later sold through a number of providers, such as MCI WorldCom, Juno, SkyTel, and UUNet, and cost subscribers between $60 and $100 per month. Modems were priced between $220 and $250. While most users enjoyed the system's excellent coverage within the 17 metropolitan areas in which Metricom operated, and found the performance to be adequate for most Web-based applications, Metricom was never able to overcome negative market perceptions. By the end of the first quarter of 2001, Metricom counted only about 40,900 subscribers. Among the long laundry list of problems were limited coverage, very high costs to expand infrastructure, submega-bit performance, and consumer price sticker shock at $80 per month.

It was clear that Metricom's wireless Internet access product was viable, and its Ricochet service offering provided the fastest mobile wireless communications solution on the market at the time, but unfortunately for Metricom, the company turned into another lesson in mobile Internet history, filing for Chapter 11 in late 2002. Consumer markets, which Metricom heavily targeted, were not willing to pay the $80 per month, and Metricom could not find anyone to finance the already-high $1 billion debt accumulated only 2 years after the new mobile Internet services launch.

Other experts in the field suggest that the slow launch and slow spread of Metricom wireless Internet access services is not a problem of the technology and its high cost, but unsuccessful execution by Metricom's executives. In other words, the technology was ready for launch and the right customers existed, but the company failed to find the right market and distribution strategy. Many blame Metricom's dismal performance on the company's wholesale model, which left Metricom at the mercy of a handful of resellers, such as WorldCom and Juno. Others said Metricom was going after the wrong customers the whole time. Metricom was trying to convince consumers to use mobile Internet services in place of their broadband Internet connection services in their homes and offices. The problem was that those wired Internet services were cheaper and offered higher speed at that time. Metricom would have been more successful in positioning their service as a mobile extension and field service for mobile workers. So often the success or failure of the mobile Internet services depended not on the quality and cost of the technology, but company execution.