But like so many other potential customers, the banished CEO soon became frustrated with Covad's inability to quickly install DSL (digital subscriber line) service, so he signed up with a competitor.
"I tried first to get Covad but it didn't work out," he said. "I have a cable modem."
Installation woes are just one reason Covad has gone from one of Silicon Valley's most promising upstarts to a poster child for the collapse of the Internet bubble.
In early 1999, Covad seemed to have everything: insatiable demand for its service and slow-footed competition from local phone companies such as Verizon Communications and Pacific Bell. At the helm was Knowling.
One of the few African-American CEOs in the technology industry, he had the resume (a successful stint at US West) and the political connections (he sat next to Hillary Clinton during her husband's State of the Union address in 1999) for success.
An initial public stock offering in January 1999 provided a financial boost that eventually pushed Covad's market value to well over $10 billion.
These were heady times in Silicon Valley and on Wall Street. Initial public offerings, especially from the high-technology industry, hit the trading floor in record numbers. Investment bankers tripped over themselves to lock up fat fees for stock offerings, and the Nasdaq composite index regularly hit new highs.
Intel, AT&T, Qwest Communications International, Crosspoint Ventures, Warburg Pincus and Nextlink Communications--now XO Communications--pumped hundreds of millions of dollars into Covad, knowing profitability could be three to five years away.
As it turned out, Covad needed every penny and then some to build its network of high-speed lines, an expensive project that's left the company with more than $1 billion in long-term debt.
In fiscal 2000, Covad posted a loss of $1.44 billion and finished the year with only $869 million in cash. To put that number in perspective, the company burned through more than $550 million in cash in the third and fourth quarters.
That huge loss and burn rate wasn't much of an issue in late 1998 when the company began preparations for its initial public offering. Covad's management at the time, headed by co-founder and Chief Executive Chuck McMinn, wasn't terribly concerned with profitability.
Before its IPO, Covad told the Securities and Exchange Commission it had only about 100 customers, including Cisco Systems, the former Concentric
Network and the former Verio. But it also was adding fledgling and established Internet service providers to its customer base, often picking up marketing expenses to lure them away from other independent DSL providers.
The strategy, expensive as it was, brought early success. Before its second birthday, Covad was able to deliver its TeleSpeed service to more than 1 million homes and businesses in the San Francisco Bay Area. It expanded into markets across the country as the Baby Bells remained apathetic about the DSL market.
Covad's annual sales jumped from $5.6 million in 1998 to $66 million in 1999. "We were told to grow, grow, grow and not to worry about profitability," Knowling said. "And that's what we did."
And it worked--at least initially.
In September, SBC Communications agreed to invest $150 million for a 6 percent stake in Covad. On the surface, SBC's move looked like a vote of confidence.
But there may have been other motives. Analysts said SBC's investment in Covad had more to do with creating the appearance of competition than holding out hope for a Covad recovery.
"The (Bells) are a lot smarter than people think," one analyst said. "It's a travesty and a tragedy in some ways. By keeping a handful of competitors
alive, they can show that they're leaving the market open to competition while maintaining higher prices."
After its early growth spurt, it didn't take long for Covad to realize that Baby Bells such as SBC, BellSouth and Verizon Communications were partners and competitors, mostly the latter. On one level, Covad needed the Bells to provide access to their infrastructure. But the Bells also wanted to provide DSL services themselves to boost revenue growth.
It soon became clear to the upstarts that the Bells weren't interested in making life easy for Covad, NorthPoint Communications, Rhythms NetConnections or any of the other independent DSL providers using their networks to resell service to ISPs and residential customers.
DSL wholesalers like Covad place their equipment in phone company central offices, using the copper phone lines owned by Baby Bells such as Verizon and SBC, which owns Pacific Bell, to deliver DSL service to their ISP partners.
From the beginning, there were myriad complications in establishing service from the Baby Bells to its ISPs and then on to the retail customers.
Covad and other independent DSL providers complained to anyone who would listen that the Baby Bells were slow in providing phone lines to them. Representatives from SBC and other Baby Bells deny those allegations, claiming they provide the same level of service to the DSL wholesalers as they do to their own direct customers.
Covad executives said the local phone companies were giving them the runaround, part of what Knowling once described as "planned incompetence," designed to keep independent DSL providers from providing fast and effective service to their customers.
"I saw firsthand just how horrible the phone companies were at delivering those (phone lines)," Knowling said. "Over time these relationships improved, but in the beginning, we had to use brute force to get them to comply with the regulations."
The Baby Bells have consistently denied those allegations and, at least in Verizon's case, have fired back with their own accusations.
Verizon recently filed a lawsuit against Covad, claiming it falsified complaints about Verizon's service to cover up its own technical shortcomings.
In a filing with U.S. District Court in San Jose, Calif., Verizon claimed that Covad managers instructed their employees to blame Verizon for a variety of technical difficulties it experienced trying to install and service DSL connections.
The suit further claims that "ill-equipped, untrained and overscheduled" Covad technicians were routinely told by supervisors to blame Verizon for problems that "systematically disparaged" the local phone company's reputation.
Covad general counsel and co-founder Dhruv Khanna denied those allegations, calling Verizon's suit an attempt to intimidate and harass Covad.
Industry analysts say this animosity illustrates the predicament Covad and
other independent DSL providers found themselves in from the beginning,
serving as both a customer of and a competitor to Baby Bells that are providing this access to high-speed connections only to satisfy regulatory demands so they can eventually offer long-distance voice and data service.
"At the end of the day, Covad is a middleman," said Jonathon Poe, an analyst at Meta Group, a market research firm in Stamford, Conn. "DSL providers are dependent on other companies' infrastructures. Not only will they have a higher cost to compete than the Bells, but they're at the mercy of when and how these companies choose to upgrade their existing networks."
According to IDC, another market research firm, Covad had 319,000 installed lines at the end of the first quarter, beating out local phone providers Qwest (306,000) and BellSouth (303,000), but lagging behind SBC Communications (954,000) and Verizon (720,000).
In addition to its flaps with the Baby Bells, Covad was hurt by its ISP customers, which were failing at a rapid clip.
Just as Covad's massive expansion reached a crescendo, the free-spending venture capitalists and institutions funding independent ISPs turned off the spigot. This caustic blend of economic realities resulted in many of Covad's customers filing for bankruptcy protection, delaying payments, or simply not paying their bills at all.
For a company burning through an average of $3.3 million a day, the loss of any cash was painful. Covad recently reported a 40 percent increase in nonpaying ISP customers. The company also reported a loss of $1.44 billion in fiscal 2000, or roughly a loss of $4,500 per installed line.
Wall Street has continued to lose faith in Covad. Not only is the company on the verge of being delisted by the Nasdaq exchange, but it had to restate its results for the first three quarters of fiscal 2000 and delayed the release of its first-quarter results for this year until the week of June 18.
"Covad's an interesting case," said Mark Langner, an analyst at Epoch Partners who confessed that he "doesn't pay that much attention" to Covad anymore. "There's no question they're going to have to restructure themselves a lot more if they're going to survive."
Meanwhile, cash-strapped ISPs and competing DSL wholesalers continue to go belly-up, eliminating much of the competition that was supposed to be created by the Telecom Act of 1996.
Last month, AT&T snapped up NorthPoint's assets for about $135 million out of bankruptcy court, giving it NorthPoint's network equipment, systems and support software, and two leased buildings in Emeryville, Calif., at a bargain-basement price.
Ironically, all this upheaval could turn out to be a boon for DSL providers, just not for small companies like Covad that lit the fire under the larger carriers. As the prices increase, there's more incentive for large carriers to expand their networks.
That other companies, and their investors, have bankrolled the expensive and laborious task of extending the lines to residents and small businesses is simply more salt on the wound.
"Carriers are finally figuring out how to make broadband profitable," Meta Group analyst David Willis wrote in a research report.
"For one thing, more than 80 percent of new installations are now done directly by the customer, without any site visit from telco technicians," Willis wrote. "This makes a huge difference in whether the carrier breaks even during the first year of service. But for the pioneers like Covad, these efficiencies came too late."
For many customers, Covad also has been a bust.
"Every other month or so, the DSL connection to the server for my Web site would go ker-blooie," one former customer said. "When that happened, a call was placed to Covad to come out and reconnect the server. Needless to say, it was an incredibly frustrating time for everyone involved, not knowing from day to day whether or not your server would be functioning, whether you'd get your e-mail or not."
Covad has tried to capitalize on the demise of other DSL companies such as NorthPoint by picking up orphaned DSL customers. The company launched its "Covad Safety Net Promotion," waiving installation fees for customers of defunct ISPs who could easily be switched over to Covad. The company also started offering a rebate on new modems for NorthPoint subscribers who send in their old DSL modems.
However, it's unclear whether Covad can keep customers happy. The company's reputation, which at least partially depended on now-defunct ISPs, has taken a hit. Some customers who used Covad's service, directly or indirectly, said the company was slow to react to frequent service outages, consistently blamed local phone companies for glitches, and generally delivered poor service.
Covad spokeswoman Martha Sessums dismissed the complaints from current and former customers, saying "there will always be crybaby boobies who are unhappy with any company.
"I'd be happy to provide you with many, many customers who are extremely happy with our service," she said.
Undoubtedly, there are many customers who are happy with Covad's service. The company's number of installed DSL connections improved to 319,000, up
nearly sixfold from 57,000 lines at the end of fiscal 1999. And there's plenty more growth where that came from, analysts say. It's just a matter of who is going to be left standing and how big the market will become.
According to Jupiter Research, the number of homes with broadband access will jump from 4.8 million in 2000 to roughly 28.8 million in 2005.
But telecommunications researcher Atlantic-ACM recently lowered its projections for DSL growth from 26.8 million lines in 2004 to 20.5 million lines, a result of service providers curtailing their expansion plans.
It's this opportunity that convinced Charles Hoffman, who previously served as chief executive at Toronto-based Rogers Wireless Communications, to accept the top post at Covad seven months after Knowling resigned.
"I think the revenue growth is going fine, but there's a definite distribution challenge there," Hoffman said recently. "My biggest priority will be to convince the employees to hang in there while we sort out all the back-office issues, especially collecting fees and converting customers from those companies who have gone broke."
For Hoffman to meet his goal, however, Covad will need to secure additional financing to stay afloat. "It's a profitable model if done properly," he said. "The technology works. People love it once it gets installed."
Hoffman will get no argument from his predecessor.
"I really hope Covad (survives)," said Knowling, who is now chief executive at Internet Access Technologies, a Houston software development company. "The technology's great and there are a lot of great people there. The sad part of the story is that was my vision, my dream, and I didn't get it done."