Monday, January 31, 2005

January 2005 Early Indications II: The Battle for the Wire

In the beginning was the wire: a telephone connection to the home or business, owned and operated by AT&T with government regulation over what was determined, in 1907, to be a natural monopoly. The poles, cables, switches, and the voice service belonged to AT&T. What would later be named "content businesses" were the province of radio and later television networks, regional newspapers, and publishers like Time-Life and McGraw-Hill, none of which connected by wire to their customers. AT&T was more powerful than the leaders in any of these sectors until GE bought RCA in 1985 to acquire NBC.

In 1984, anti-trust authorities determined that while local service was still a monopoly, long-distance service should be opened and deregulated. Accordingly long-distance was provided by AT&T and competitors like MCI while local service was shifted to one of seven newly-created Regional Bell Operating Companies (RBOCs), which were geographically delimited and didn't have any notable competitors. Who you paid, and how much, depended on how far away you were calling and how long the connection lasted. Coincidental with, but hardly irrelevant to divestiture was the rise of the fax machine as a second source of traffic over the wires.

Ownership of the wire was thought to confer great advantage. During this period, according to Frances Cairncross in The Death of Distance, AT&T was told by a certain firm of strategy consultants that the total market for mobile phones would be 900,000 two decades later, so it exited that business only to expensively re-enter it in the 1990s by buying McCaw, that it later had to sell to raise cash. In part, this transaction allowed AT&T direct access to consumer markets. (By 2003, meanwhile, there were over 500 million cell phone users worldwide - McKinsey was wrong by a mile.)

Without direct connection to wireline customers, AT&T tried repeatedly, and often successfully, to add consumer presence to its already impressive corporate customer base. Over the years AT&T launched a brilliant credit card program that it later sold to Citicorp, as well as a potentially competitive Internet Service Provider that it soon effectively abandoned. AT&T also bought a computer company (NCR) that it mismanaged and later divested, and two very expensive cable TV properties that it - not surprisingly - later sold at a loss. AT&T also spun out its equipment-manufacturing group as Lucent Technologies as it struggled to find a business model for capitalizing on its place in the communications ecosystem. Arguably the most powerful company in the world at its peak when it employed over a million people, AT&T repeatedly stumbled as it tried to reinvent its business model in the aftermath of deregulation.

The U.S. Congress passed legislation that was supposed to sort out rules of communications competition in 1996, but that law's interpretation has been contentious, uneven, and protracted. In the meantime, several rapidly evolving technologies - including voice over IP, wireless data transmission, encryption, and search - have made many of the Act's provisions "quaint" and "obsolete," as they say in Washington. As of late last year, the rules for who could do what were not fully in place. Directly because of this confusion and delay, the U.S. now lags over a dozen other nations in the speed and quality of its high-speed connections, and many commentators assert that this policy faux pas has major implications for trade and national competitiveness.

The fight for the wire has created new competitors whose actions are often front-page news: FCC chairman Michael Powell is stepping down and leaves a jumble of regulations, litigation, and lobbyist complaints that will have to be sorted out by his successor, the courts, and perhaps Congress if it revisits the Telecom Act. SBC (itself an amalgamation of three of those seven RBOCs) finalized arrangements to acquire AT&T. Comcast and the other cable companies are striving to define their place in the new world, knocking heads with everyone from Fox to Disney to Verizon to AOL/TimeWarner. Google hired some key Firefox developers, renewing last year's speculation that Google could elbow its way into the browser wars. Cisco is installing tens of thousands of corporate VoIP phones at a shot. Apple just reported robust earnings, as did Yahoo.

Control of the wire remains highly contested and economically central, but the competitors and the areas of competition have both evolved. In the 1990s, long discussions were held over the relative merits of content and connectivity - the ExciteAtHome debacle remains memorialized in its infamous real estate in Silicon Valley. But rather than inhabiting a binary universe, the parties concerned with the wire now appear to be distinguished as at least four camps. Whether and how it's possible to make money in each of these are still open, and urgent, questions.

The first group of companies are the owners of the wire, the daughters of Ma Bell and their competitors. Thus far, it's proven to be next to impossible to make money here. Upgrading the physical network is expensive, and payback is uncertain, largely because of the Telecom Act. In a closely-watched pilot deployment, Verizon is attempting to invest well over $1000 per house in fiber-optic connections to customer premises, and they've had rapid uptake from gamers in particular who will pay a premium price for a fast raw connection. The company may not be able to afford to expand beyond the trials for any number of reasons: investor revolt, regulatory setbacks, technological innovation in wireless, or competitive pre-emption of its planned video services. Across the landscape, carriers face what David Isenberg (following equity analyst Roxane Googin) has called the paradox of the best network: you apparently can't make money delivering the kinds of fast, "featureless" connections that make possible rapid innovation and appealing - from the users' perspective - cost/performance curves.

Even though they own a different wire into the house, I wouldn't lump the cable operators into the first group. Instead, they own an application - television - that's popular and almost immune from either competition or pricing pressure: monthly bills keep climbing despite a slow economy and aggressive competition from satellite dish networks. Cable is lightly regulated, particularly in contrast to the RBOCs. That the cable operators also own a wire is gravy because they can add mid-band Internet connection and voice service, two additional popular applications, to the already-profitable television core. In response, it shouldn't be a surprise that Verizon hopes to deploy video over the aforementioned fiber, but that move would still leave them needing content.

Other application owners include Apple, whose iTunes service lets companies like SBC or Cablevision build infrastructure, bill customers, and do other unappealing heavy lifting: the hyper-profitable iPod hardware play bundles hardware, an application, and a content base, the latter two of which rely on the network for their existence but do not by themselves turn a profit. A final noteworthy application that has yet to drive revenue is instant messaging.

A content area looking for a predominant application is digital photography. Adobe and Google (Picasa), among others, have strong offerings here, and Kodak is trying to enrich its Ofoto property even as Flickr tries to make noise as an upstart by making the photo application more social. It's hard to see many areas with more promise, and I can't imagine that a leader won't emerge in the next 18 months.

The transition of voice service from its state circa 1980 to what it will be by 2010 is simply stunning. First, billing by duration of call and distance of connection is eroding rapidly. Second, the ability of voice service to be disconnected from a wall jack is changing people's habits and industry economics. Significantly, much as fax rose on top of the circuit-switched infrastructure by being modulated into audio, data in the form of text messages is now an increasing percentage of cellular traffic: what started life as a voice network became, with the help of generational changes and cultural shifts, a data service that utilized the existing infrastructure, from the 12-key dialpad onward. Meanwhile, Verizon's profitability last quarter depended disproportionately on its stake in the wireless business.

Sometimes the customer connection is not a wire at all, but companies in this space (pun intended) are still directly relevant to the fight for the wire. Sirius and XM, the two primary satellite radio providers, have clearly been affected by the iPod's success. It's fascinating to watch Apple ease into the car audio market, initially via a cable from their hardware to a BMW's factory stereo, at the same time that XM and Sirius try to go after both the indoor connected listener and the mobile exercise segment with the XM2go devices that compete directly with the iPod in its core market. Unlike Apple, however, XM and Sirius aren't shooting for hardware sales, but for ongoing subscription revenues.

Information, including metadata or information about information, comprises the third industry directly related to the wire. Whether it's Yahoo selling advertising space, Dow Jones or Hoovers selling investment data, or Google, Amazon, and Microsoft fighting for search supremacy as the gateway to billions of pages of content, many companies have staked their claim in the information market. AOL/TimeWarner, the TV networks, and others must often deal either with intermediaries (such as Cox) or with competitors because barriers to entry can be low. Comcast and other cable providers are integrating upstream, getting into the content business by buying interests in professional sports teams: good luck watching the Philadelphia 76ers regular-season basketball games on satellite TV, given that Comcast, which is 11.5% owned by Microsoft, owns the team. There are many other similar arrangements. The answer to the age-old question as to which layer in the cake commands the highest brand premium is of course "it depends," but clearly content appears to have an upper hand over transport.

Building - and profiting from - the equipment connected to the wires can be a frustrating proposition. Cisco has clearly prospered in this niche, but such strong firms as HP, Sony, and Ericsson have had at best mixed success. Samsung appears to be well positioned for future growth, in part because it can market to the world with different product mixes. We mentioned Apple earlier, but apart from that, it's hard to find any hardware producer except Dell who's maintained margins while increasing volume. Intel recently reorganized, in part to address precisely this question of how processing relates to communications.

In the business-to-business domain, the growth of Linux (another product of the Wire) has driven proprietary, expensive iron from HP and Sun into new competitive positions as commodity "white boxes" serve admirably and often disposably. Storage becomes a different business when networks are central to its architecture and deployment, and here again Cisco appears to be blurring the line further. IBM and HP want to sell computing as a service over the wire, but this market is still nascent.

What can we draw from this discussion? First of all, the companies best positioned for a new market often have the most to lose from repositioning their existing assets. AT&T and Kodak are important examples here: technological innovation made them great; greatness translated to size, bureaucracy, and turf wars; and their research labs had important new intellectual property to undergird the next generation that often went overlooked or underappreciated. But the classic innovator's dilemma, precisely as described by Clayton Christensen, froze them in agonizing ambivalence. Caught between trying to build fences of every possible description -- patents, regulation, branding, price-cutting, alliances -- around the existing franchise and seeing that change was accelerating, these companies' executives and the corporate culture prevented the weeding, pruning, hybridization, and seeding that could produce new growth.

Second, there appears to be a two-way street regarding bundling with adjoining layers in the model: Comcast and Apple use two or even three of the four to build an offering stronger (or more coercive) than any one piece standing alone. Conversely, when AT&T lost hegemony over the voice application and was relegated to more and more a pure transport role, it lost momentum, margin, and ultimately direction. Finally, innovation that disrupts existing revenue streams - in either direction - is seldom obvious to incumbents. Time-Life, owners of Sports Illustrated, didn't see ESPN coming, nor did Atlantic Records anticipate the impact of MTV until long after it was established. For SBC, whose chairman and CEO Ed Whitacre is staking his legacy on the AT&T deal, the potential for similarly disruptive change is certainly ample.

We began at the beginning; where are we now? The economics of networks can behave with extreme speed, it is true, but the most persistent lesson of the wire and its impact relates to the human propensity to see what we want to see: the light at the end of the tunnel, the turn in the road to profitability, the upstart competitor as an object of pity or derision. The distinctly American invention of the late-20th-century corporation has been exposed, in many of its variations, as nearly immune to intentional change, even in the face of massive market evidence pointing to substantial threats and opportunities. Forensic autopsies of AT&T will be fascinating to read, but they will merely confirm what's been evident for decades. In the end what doomed the company had far less to do with wires and protocols than attitudes and habits.