The recent flurry of merger announcements has led some pundits to proclaim that the telecom industry is transforming back into its old Ma Bell structure. And on the surface, it’s easy to see why.
Within 12 months, Bell Atlantic, NYNEX, GTE, AT&T, BellSouth, MCI, Pacific Telesis, Ameritech, and Southwestern Bell will be jam-packed into just two companies, which, when combined, will account for approximately 80% of all U.S. access lines. On the supplier side, Alcatel, Lucent, Newbridge, DSC, Ascend, and Cascade will have collapsed into one. Most recently, Nokia and Siemens have announced their intention to merge their network equipment operations into a new entity. Moreover, few expect any CLEC to provide meaningful competition to AT&T or Verizon, and venture capitalists have as much interest in startup equipment vendors as they do sock puppet dot-coms.
Yet in spite of all the consolidation, innovation is alive and well in the telecom industry. One main reason hardware products have advanced regardless of diminished competition is that by shipping their products to multiple customers, merchant silicon vendors have been able to reach higher production volumes and lower unit costs. Seven years ago, when startups seemed to rule the world, a double-density ADSL card was considered state-of-the-art, but today a 48-port ADSL2+ card can be purchased from a reseller for a comparable price. The price/performance gains of Ethernet line cards have been even more impressive. These are just two advances that occurred while RBOCs were merging with each other, CLECs were filing for bankruptcy, and startup equipment vendors were being sold for little more than the cash they had in the bank.
This trend of communications semiconductors and networking hardware separating into discrete companies has been essential to pushing down the price per bit of networking cards. As the fixed development costs of proprietary ASICs increases with each reduction in process geometry, it has become very difficult for a startup to reach break-even manufacturing levels without being able to sell its designs to someone else. As a result, new data networking features that were once developed by box makers like Extreme, Foundry, and Bay Networks are increasingly dependent on communications chip suppliers like Broadcom, Marvell, and Atheros.
In addition to the chip sector supplying new features into the industry, the major phone companies have been pushed forward by the fact that their customers can also be their biggest competitors. While AT&T and Verizon are going after each other more than ever now that they both own competitive long-distance divisions, their greatest competitor besides each other is not the cable company, Level 3, or some other CLEC — it’s the business that chooses to manage its own network.
Consumer groups have blitzed the press wires with fearful cries that these mammoth telcos will have little reason to advance new technologies. However, the telcos themselves are no longer the source of technology innovation. Business customers themselves are free to choose which hardware they want on their internal connections. If the phone companies do not provide the services they demand, corporations will just keep their lucrative traffic on their own networks. This is why many enterprise protocols, such as Ethernet and IP, are ultimately embraced by slow-moving carriers.
Another factor preventing the reemergence of Ma Bell is the global market for equipment. The old Western Electric-AT&T monopoly was largely shielded from what happened in other countries. Other industrialized nations had their own equipment suppliers, which mostly sold their products to the in-country PTT. Today’s worldwide market for hardware means that AT&T and Verizon are using many of the same products as France Telecom and Deutsche Telekom, and cannot dictate product requirements the way AT&T used to with Western Electric.
The international hardware market has already taken a toll on the RBOCs’ OSMINE process, which is increasingly becoming outdated and outmoded for two particular reasons: one, because few PTTs force vendors to go through such a strenuous software integration process, and two, because element management systems have evolved to the point that little benefit is achieved by running everyone’s systems through Telcordia before implementing them in a live network.
With merchant silicon condensing upgrade cycles, global requirements taking many product development decisions out of the Bells’ control, and more software-based enterprise equipment sneaking into telco networks, the major phone companies are increasingly expecting to toss out a lot of their hardware soon after it has been installed. The staid class 5 switch, add/drop multiplexer, and crossconnect markets used to be built on tightly integrated manufacturing processes, slow upgrade cycles, and limited competition among suppliers. The pace of change in these hardware markets dictated RBOC purchasing procedures. The fact that there are now fewer Baby Bells than there were 10 years ago is having little impact on the telcos’ need to swap out low-density, low-rate cards constantly. At BellSouth, for example, 83% of the DSL lines installed in the first quarter of 2006 were provisioned at 3 Mbits/sec or higher, an accomplishment made possible by junking 1.5-Mbit/sec ADSL cards long before their useful life was up.
While the communications industry of 2006 does not look like the open marketplace Congress hoped for when it passed the Telecom Act of 1996, it continues to advance in new and exciting directions, thanks to its ever-closer ties to the computing and information technology sectors. Product development cycles, once dictated by a bureaucratic monopoly, are now being pushed forward by advances at Silicon Valley and Asian fabs, and the RBOCs no longer control the locus or the rate of innovation in the industry as they once did. These factors create an interesting new world that should continue to encourage innovation on many levels.
David Gross is a senior telecom analyst with BIA Financial Network (www.bia.com). Rick Ducey is executive vice president of strategic consulting for BIA Financial Network. He also serves as president of SpectraRep, BIAfn’s wireless multicast data delivery division. He can be contacted at 800-331-5086 or firstname.lastname@example.org.