Fixed Wireless Business

At This WISP, a Focus on Building

The WISP business model has three key flaws, all of which show up as expenses: CPE, backhaul, and access point locations. But one WISP claims to have solved all the problems.

by Alex Goldman ISP-Planet Managing Editor [March 2, 2004]

"California is the fifth largest economy in the world," David Williams reminds us. Williams is vice president of marketing and business development at Fremont-based NextWeb, a business oriented WISP that serves high tech areas around San Francisco and Los Angeles (including, of course, Silicon Valley) and claims to be the largest business WISP in the nation.

With 2,000 business customers in 120 Californian cities, the company has a large base, but the cities it serves are home to "about half a million businesses," Williams estimates.

Williams knows the territory. As co-founder and COO of Whole Earth Networks, he started one of the world's first ISPs, which was sold in 1998. He is bullish about his company, and feels the press and the government have been ignoring the urban ISP phenomenon, focusing instead on rural WISPs.

"The FCC is following rural success stories, but here we are going head to head with the established telecommunications infrastructure in major business markets."

The company is funded by venture capital, but in a way which, Williams feels, reflects the lessons of the telecommunications bust. "There was a lesson learned. If you take a lot of money and build in all the NFL cities, you may be able to build the infrastructure, but the difficult part to scale is the national sales organization, the national advertising organization, and the national backbone."

The NextWeb plan is more gradual and more focused. "We've pursued a regional, focused effort, deploying deeply. We have almost ubiquitous coverage in the markets we're in. The regional approach has worked well for us. It allows us to deploy all the resources we need to build the business."

The company had, however, considered trying to start nationwide. "We did talk about a bigger investment round, but agreed that we should prove to investors and the business community that we could take a single market from startup through positive earnings," says Williams.

The two big selling points of wireless broadband are speed of deployment and scalability. The company claims it can deploy service in three to seven business days, and the technology allows it to offer a greater variety of bandwidth increments than the telco monopoly. "You cannot grow a business class T-1," says Williams. "Either you add T-1s, or you jump to a T-3. We can scale to, for example, 10 Mbps without a truck roll."

Asked about the three perils of the residential WISP: backhaul prices, CPE costs, and landlords, Williams is dismissive. CPE prices are not a problem when selling bandwidth at business prices, but NextWeb will do better as CPE prices fall. "Lower CPE prices do lower our cost burden, but we still generate a phenomenal margin," he claims.

Backhaul, Williams admits, is expensive. "There's a misconception that this is an inexpensive business to be in," he complains. "If you're building a carrier class network, you need to use licensed backhaul." That's expensive. The company has licenses for 18 GHz spectrum and its backhaul links are high capacity, at 100 Mbps or 155 Mbps depending on the location. Each node has backhaul from at least two different directions.

"The goal is to make sure every base station has some level of redundancy," explains Williams. "Also, we can isolate any failure at one base station. We do have some spurs on our network as we grow, but the core network is designed on the concept that it has to be up 100 percent, not four nines or five nines."

He says that many other WISPs made the mistake of designing for the present, not the future, and have had to add capacity at nodes just as those nodes were about to break even.

By now, NextWeb has, to some extent, standardized the process of building an access point. It is a two stage process (but a planned one) in which the company starts out with a single base station and adds the full NextWeb infrastructure over time. "It's been a very successful model for us," he says. "But it's more expensive than what many smaller players would do."

It's all about competing. "If you try to compete with the monopoly carriers, you have to build a network with the same class of service and the same scale," explains Williams.

Obtaining locations for base stations remains a real challenge. Listed fourth on the company's Management Team Web page is Beth Rautiola, director of site acquisition and corporate counsel (Williams is second). The title makes it clear that she's the company's top lawyer, and also that most of her day to day work is dealing with agreements about base station locations.

In addition to a dedicated site acquisition team, NextWeb has a key relationship with one of its investors, Oakland-based nonprofit health industry giant Kaiser Permanente.

"Not everybody knows this, but Kaiser is the second largest private commercial real estate owner in California," says Williams. "They own a tremendous amount of real estate assets that we have exclusive access rights to."

With Tenet Healthcare and ACPI (a health care REIT), both of which were introduced to NextWeb by Kaiser, Nextweb has access to about 500 buildings in key business areas in California, plus 500 more buildings in 43 states, should it choose to expand beyond California. This helps.

With the central problems of the WISP business model licked, NextWeb looks sure to grow, organically, and through mergers and acquisitions.

—End