The digital subscriber line industry

The speed with which Internet usage grew in the late 1990s created a large and rapidly growing market of Web surfers who needed all kinds of things, among them faster Internet access. By 2000, 375 million people worldwide had Internet access, of which a third were in the USA. All indications were that Internet usage would continue to grow worldwide. In the USA, forecasts indicated that by 2004, 75 per cent of all households would have Internet access.

Not only did this market appear attractive at the macro-level, but towards the end of the decade a new technology called digital subscriber lines (DSL) made it possible to deliver data – including Internet data – over ordinary copper phone lines at dramatically higher speeds than the common 56k modems could deliver. At speeds of up to 1.6 million bytes per second – 10–28 times faster than 56k modems – this technology offered dramatic benefits, including:

For consumers, faster downloads, thereby saving the surfer’s time. Imagine – no more waiting for the file to load. Stock quotes, airline reservations, movie tickets in seconds . . .

For small and medium-sized businesses of every kind, a chance to join the Internet revolution.

For all, the ability to download audio and video, which was not very practical on slower 56k modems.

With benefits like these and a large and rapidly growing market, DSL technology looked like a terrific entrepreneurial opportunity. But how attractive was the DSL industry? Let’s take a look at the American DSL industry in 2000 from a five forces perspective.

Threat of entry

For an entrepreneur considering the DSL industry, a number of barriers to entry presented themselves. These barriers included logistical hassles, turf wars, technical knowledge and equipment costs. All difficult, to be sure, but great to have in place if one could somehow overcome them and enter. And enter they did. Armed with high-flying business plans that attracted buckets of venture capital, executives from telecom firms left their employers to set up DSL firms with names like Covad, Rhythms and NorthPoint Communications. With market growth like this, how could they go wrong?

By the end of 2000, Covad, Rhythms and NorthPoint Communications had won 442,000 subscribers, but they were finding one entry barrier far more difficult than they had imagined. To serve their customers, they needed to send their signals along the proverbial ‘last mile’ to the customers’ premises, the twisted pair of copper wires owned by the incumbent telephone companies, or regional Bell operating companies (RBOCs). But try relying on another company for such access when that same company is a staunch competitor.

It seemed that the RBOCs also saw the potential for DSL, and they were busily deploying it on their own. How likely were they to allow the newcomers access to their copper and other facilities where the gear of the DSL upstarts had to be co-located? In spite of regulatory efforts to encourage such access, the reality was quite different. As AT&T lobbyist Peter Jacoby said, ‘The basic problem, if you’re a DSL provider, is you’re relying on Baby Bell facilities to deliver your product’.

With little to gain by helping out the new DSL providers and, in some cases, a lot to lose, the RBOCs developed a reputation for being slow and stubborn. To compound the problem, once accessed, the copper wires were not always ready for DSL transmission. DSL was often incompatible with older copper phone lines. But carriers didn’t know which lines would be incompatible – and costly to convert – until they started work in each neighbourhood. These problems took additional time to resolve and led to delays of up to six months in getting new customers’ services switched on – not exactly what companies poised for high growth had in mind.

A barrier to entry that’s high enough to keep out competitors is good. A barrier so high that you can’t get over it is a problem. Access to the ‘last mile’ was such a barrier.

Supplier power

Siemens, Lucent, Nortel, Cisco, Nokia, Ericsson, Alcatel, NEC, Fujitsu – the DSL suppliers read like a laundry list of the most reputable high-technology firms in the world. While some of these suppliers offered highly differentiated products that could command high prices, much of what the DSL companies needed was available from multiple suppliers. The result was that few suppliers had significant influence over the DSL industry.
In general, supplier power was not a significant problem for the DSL industry.

Buyer power

DSL buyers came in all shapes and sizes. While DSL was an attractive option for residential customers, especially those who were heavy Internet users, it was also an appealing choice for small and medium-sized businesses. It permitted high-speed Internet access at a fraction of the cost of older T1 lines used by large businesses with heavy telecom usage. The good news for the DSL providers was that these kinds of buyers had little power. They knew the advantages of high-speed Internet access and were prepared to pay for it, especially if the service were moderately priced, which it was. Buyer power was not a problem either.

Threat of substitutes

DSL was one of several methods of connecting to the Internet at significantly faster speeds than when using a traditional 56k modem. The most significant substitute for DSL in residential markets was the cable modem, which, like DSL, offered comparable service at a reasonable monthly price. Unlike DSL, cable modems used coaxial cable, the same cable that carried television to most American homes. Cable broadband providers advertised connection speeds of 1–3 Mbps, 15 or more times faster than a 56k modem and faster, in some cases, than DSL.

In the residential market, cable operators got the jump on DSL due to a variety of factors, the most significant of which was the inability of telcos, particularly in the USA, to provide DSL service at distances of greater than two miles from the central office. Since cable companies existed long before the advent of broadband Internet connections, cable companies also had the benefit of an established customer base and access to millions of residential homes and neighbourhoods. By the autumn of 2001, there were approximately 5 million cable-modem subscribers.

There were, however, some shortcomings of cable modems. First, the personal computer had to be located near a TV cable, which meant additional cabling in most homes. Second, a cable employee had to hook up the system. And, because cable networks operated on a shared basis, some users were concerned about the safety and privacy of their Internet data and about slower service when the systems were busy. Nonetheless, with cable television already present in most American homes, cable was a powerful substitute for DSL, limiting the prices DSL providers could obtain, with a consequent effect on margins.

For the business market, cable was typically not a viable substitute, as cable often didn’t serve commercial areas. Alternative Internet connections, like fixed wireless, were comparably priced and provided similar speed to DSL, but they required line-of-sight transmission, restricting many businesses from its service. Sharing a T1 line with one’s neighbours was also a possibility, assuming there were neighbouring companies that wanted highspeed access. Thus, for the commercial market, the threat of substitutes was relatively low, and this is where most newcomers like Covad, Rhythms and NorthPoint Communications focused their efforts.

Competitive rivalry

Not so long ago, the American telecommunications landscape looked very different. With no competition and a virtual monopoly on telephone service, the Baby Bells and AT&T dominated the local and long-distance markets. All this changed as the trend towards deregulation took hold in the USA. In 1996, US Congress passed the Telecommunications Act, intended to increase competition in the telecommunication industry, encouraging service-provider start-ups and long-distance phone companies to enter local markets previously dominated by the RBOCs. Greater rivalry was the intent, as well as the result.

The act set off a stampede of competition. It opened lucrative local telephone markets to new players. At the same time, the Internet was booming, and technology prophets predicted that high-speed networks would be needed to satisfy growing demand for digital traffic. A gold rush ensued. Nearly 400 telecom companies raised $489 billion on the stock market and took on an additional $389 billion in debt.

Without question, however, the incumbent operators had a distinct competitive advantage. The long-distance companies, themselves products of earlier deregulation, and the Baby Bells had abundant resources. They owned most of their infrastructure – including the ‘last mile’ – and they could offer a comprehensive array of communications options. They also had direct contact with prospective DSL customers. NorthPoint Communications’ own research showed that fewer than one-third of consumers even knew there were alternatives to the Baby Bell.

The Baby Bells proved vigorous competitors indeed. In 2000, when Covad, NorthPoint Communications and Rhythms served 442,000 subscribers combined, SBC (a Baby Bell serving the southwestern USA) alone served 767,000 subscribers, Verizon had 540,000, Qwest had 255,000 and BellSouth had 215,000.42 Rivalry was intense!

Summary of industry attractiveness

Severe threat of substitutes in the residential market. Brutal rivalry for residential and commercial customers alike. To top it off, the entry barrier to the ‘last mile’ posed by the RBOCs was difficult to breach. Is this the kind of industry where you’d like to play?

The results were not pleasant, especially for users who had subscribed to the DSL services of the new entrants. Investors weren’t happy either:

NorthPoint Communications, a DSL leader with more than 100,000 customers nationwide, closed in March 2001, leaving its customers scrambling for online access.

Soon afterward, Winstar Communications and Teligent went into insolvency.

By June of 2001, over 20 major providers of DSLs had shut down, filed for bankruptcy or found themselves dangerously short of cash.

The market for DSL services was attractive, to be sure. Customers loved DSL, with its blazing speed that was always turned on. But unfavourable forces in the USA had rendered the industry so unattractive that many new entrants had failed quickly

Lessons learned from the DSL industry

We’ve just said that barriers to entry are good, to be loved and cherished. But one must be sure that there’s no barrier that’s insurmountable, as was the case for DSL firms’ lack of access to the ‘last mile’ of copper to the customers’ premises. It takes only one insurmountable barrier to put you in deep trouble.

A second lesson to be taken from the DSL story is that substitutes – in this case, cable modems – can place strict limits on the prices you can charge. Plastic containers cap the prices that aluminium or glass packaging companies can charge, and so on. You must look outside your own industry to see what other industries might bring to your market. The once stealthy but now steady advance of digital technology into the realm of traditional silver halide chemistry in the photography industry does not bode well for companies like Kodak, Agfa and Fuji, whose aging technologies may not last very much longer.

It is also worth repeating that large and fast-growing markets like the Internet population are not sufficient to ensure a successful venture. DSL entrepreneurs were enticed by the predictions that, by 2001, DSL-enabled phone lines would reach as many as 10 million American homes. It’s worth noting that research companies make money peddling rosy forecasts in all sorts of markets and industries. Take them with a grain of salt, and don’t forget to examine industry attractiveness too. If it’s easy to enter the game you hope to play, then you can be sure you’ll have company sooner or later. And if the industry is structurally unattractive, as was DSL, then most entrants will probably fail.