901 Explorer Boulevard
Huntsville, Alabama 35806-2807
Fax: (205) 971-8699
Web site: http://www.adtran.com
Sales: $250.1 million (1996)
Stock Exchanges: NASDAQ
SICs: 3661 Telephone & Telegraph Apparatus; 3825 Electric Measuring Instruments & Meters
In 1970, a 29-year-old electronics entrepreneur named Mark Smith founded Universal Data Systems (UDS) in Huntsville, Alabama, to make moderately priced data communications equipment for the growing U.S. computer industry. UDS’s principal product was an early incarnation of today’s commonplace computer modem, a device for connecting computers over phone lines so they can transmit data to each other. The first modems connected mainframe computers over dedicated phone lines using a standard called “serial synchronous data transmission,” which required that the sending computer be perfectly synchronized with the receiving computer.
When the minicomputer began to introduce diversity into the computer industry in the 1970s, however, AT&T’s Bell Laboratories created a new standard, “asynchronous serial data transmission” (ASDT), that enabled computer transmitters and receivers operating at different speeds to successfully transmit data to each other. Because ASDT would make possible the connection of different connector designs, electrical requirements, and transmission conditions, AT&T also developed a series of ASDT specifications that forced order on the world of minicomputer data communications. Moreover, because AT&T created the specs, industry users were forced to rent AT&T’s ASDT data set, giving it effective control of the fledgling modem industry.
Of Modems and Monopoly: 1978-84
Then in the late 1970s the acoustic coupler was invented, suddenly making it possible for a computer modem to be connected to the phone line by simply inserting an ordinary telephone handset into the acoustic coupler’s twin cups, which were shaped out of foam rubber to receive the handset and eliminate signal decay. The acoustic coupler converted the sound waves sent over the phone line into a stream of electrical signals the modem and computer could understand. More importantly, it opened the door for a modem industry that did not require the AT&T ASTD standard. Companies like Carterphone, Datec, Digicom, IBM, U.S. Robotics, and Universal Data Systems rose to create a new generation of initially expensive acoustic coupler modems, which transmitted data at 300 bits per second (bps)—at the time believed to be the fastest reliable data transmission speed possible over the phone network.
By 1978 UDS and its competitors began offering inexpensive acoustic coupler modems for the data communications market growing up around the mainframe computer centers of the nation’s universities and scientific organizations (the focal points of the Internet’s forerunner, ARPANET). That same year, however, electronics giant Motorola acquired UDS, and Smith became a Motorola vice-president. AT&T was meanwhile defending its ASTD modem turf, suing acoustic-coupler maker Carterphone for daring to ignore its modem specifications. The case went all the way to Supreme Court, where in 1984 the nine Justices ruled that AT&T had been running a monopoly in the U.S. local and long distance service and equipment markets.
As a consequence of the Court’s trust-busting decision, the Federal Communications Commission (FCC) was granted regulatory control over the standards for telephone and computer communications AT&T once ruled, and Ma Bell herself was forced to spin off her local phone service businesses into seven regional bell operating companies (RBOCs), which were in turn prohibited from offering long distance phone service or manufacturing telecommunications equipment. With this landmark decision to inject competition into the telecommunications equipment industry, Smith saw an opportunity to step in and supply the telecommunications network equipment that the seven RBOCs—and over 1,300 independent U.S. phone companies—would now have buy on the open market.
Enter Adtran: 1985-91
With John Jurenko, Lonnie McMillian, and four other founders, Smith left Motorola for good in 1985 and formed Huntsville-based Adtran Inc. (for “advanced transmission”), which opened for business in January 1986. Despite his roots in the modem industry, Smith saw the historic AT&T divestiture as opening a wider window of opportunity to manufacture an entire range of high-speed digital infrastructure products to optimize the performance of the nation’s copper-wire phone network. “We founded this company,” Smith later told Investor’s Business Daily, “to take advantage of the trend toward complete digital systems ... the technology of the future.” Specifically, Adtran would focus on the manufacture of digital transmission telecommunications equipment for phone companies’ “Central Offices” (the local telephone switching and distribution facility) and their “Local Loops” (the portion of the phone network connecting the Central Office to individual homes and offices).
In the mid-1990s, there were roughly 30,000 such Central Offices in North America, which provided their subscribers with access to a specific portion of the phone network’s bandwidth either on a switched basis (known as “switched access”) or on an exclusive basis (“private line”). From the central switching office, phone lines ran to the Local (or home) Loop, which numbered some 160 million in the United States alone in the mid-1990s. Local Loops consist of a circuit assembly housing a channel unit and U-Basic Rate Transmission Extender that plugs into a channel bank or shelf of the Central Office. Each Central Office contained several hundred to several thousand circuit assemblies, one for each telco subscriber. Seizing on this huge market, Adtran began designing and manufacturing speed-optimizing versions of these basic Central Office/Local Loop components, which would rely exclusively on digital rather than analog technology to wring the maximum performance out of conventional copper-wire phone lines. Adtran was soon developing a complete line of data transmission, repeater, extension, and termination products, from dataports and channel and data service units to digital repeaters and extenders (used to boost a phone signal in order to extend the distance of the transmission).
The heart of Adtran’s first digital telco equipment line in the 1980s were its “Digital Data Service” (DDS) products, based on the pioneering DDS technology introduced by AT&T in the late 1970s and which were ideal for multiple-location data transmission operations such as linking automated teller machines to banks. Huge advances were being made in data transmission technology, and the unwieldy wire coils and unstable capacitors of early modems, for example, were being replaced by integrated circuits, quartz crystals, single-chip circuits, improved data compression methods, and cheaper manufacturing techniques. From its all-digital DDS technology, Adtran began developing “digital signal zero” data ports for interconnecting two channels in a telco’s Central Office; OCU dataports (a kind of Central Office channel assembly) and related products for delivering DDS to telco subscribers; mid-span DDS repeaters for extending the Local Loop’s service range; and Digital Data Station Termination equipment to monitor and detect problems in each individual circuit.
By 1991, Adtran had already gained leading market share in its target market, Central Office and Local Loop telecommunications equipment for U.S. télcos. To solidify its hold on these core telco customers—which five years later still accounted for 60 percent of its sales—Smith adopted a three-pronged strategy of new product development, customer service, and product excellence. But to reduce Adtran’s vulnerability to the changes in the téleos’ demand for equipment, he also sought out a new market: the corporate end-user served by the Local Loop, in a segment known as “customer premises equipment” (CPE). At the customer’s premises, the phone signal transmitted by the Central Office is received by terminating equipment where it is converted for use in such products as LAN interconnection gear, video conferencing equipment, PBXs, and personal computers. Adtran’s first CPE products were data service units, unveiled in 1991, and by 1996 it was manufacturing some 100 different types of these products alone. Within four years Adtran’s corporate-user sales had grown to account for 20 percent of its sales.
OEMs and ISDN: 1992-94
With sales to corporate customers bolstering its telco revenues, Smith turned to a third target customer: original equipment manufacturers (OEMs). Adtran would now customize its standard Central Office equipment for manufacturers of telco and corporate-user equipment, producing “private-label” versions of its products under the nameplates of manufacturers of video delivery equipment and LAN interconnect equipment as well as information services providers. With its OEM customers underwriting the cost of modifying its products, Adtran collabo-ratively tested and refined the altered equipment with the OEMs, who often agreed to purchase a minimum number of products, which ranged from dataports, data service units, and other corporate-user products. By 1994, sales to OEMs accounted for 26 percent of sales and Smith took Adtran public in a $38 million offering.
With its three-tiered customer base—téleos, corporate end-users, and OEMs—now identified, Adtran began developing a nationwide sales force and network of equipment resellers and pumped prodigious sums into new product development. Smith’s strategy was to assign his engineers the designing of new products with an eye toward their later redesign. When new technologies emerged or competitors’ products shifted the playing field, Adtran would be able to enhance its next generation of products quickly and cheaply, allowing the savings to pass on to customers who in turn would reward the company with greater market share. The company’s new-product life cycle was reduced to a brisk 18 months.
While AT&T’s DDS technology was the backbone of Adtran’s first digital products, as early as 1989 it had begun making devices for two new emerging data protocols—ISDN and HDSL. ISDN, short for Integrated Services Digital Network, used the telco’s copper-wire phone lines to deliver up to 128,000 bps of text, video, audio, or fax data over a single phone line while simultaneously holding a phone conversation. Realizing the technology’s potential, Adtran developed ISDN products for extending and terminating ISDN service from PBXs and Central Office switches to multiple locations. In 1993 it unveiled its ISDN Service Unit product family, the first of its kind on the market, which among other features allowed telco’s to support ISDN calls without upgrading to ISDN-capable switches at all their Central Offices. By 1996 Adtran was the leading supplier of ISDN equipment to the RBOCs, with more than one million circuits established.
Although U.S. téleos could use their existing networks to provide roughly 80 percent of their customers with ISDN service, infrastructural limits prevented them from extending it to the remaining 20 percent without new technology. In the early 1990s the Midwestern RBOC, Ameritech, selected Adtran to develop a “Basic Rate ISDN dual loop extender” to push ISDN’s range from the conventional limit of 18,000 feet to 29,000 without the use of expensive repeaters. In early 1996, Adtran demonstrated this technology to other U.S. téleos as Total Reach, and by the end of the year GTE, Pacific Bell, and BellSouth had signed on. Between 1995 and 1996 alone, Adtran’s 50-odd ISDN products experienced a 41 percent increase in revenues. In 1996, Adtran introduced ISDN terminal adapters that automatically programmed ISDN modems with the correct phone line configuration information, greatly simplifying the ISDN installation process.
HDSL, Tl, and POET: 1995-96
Adtran’s 1989 introduction of ISDN devices was coupled with its first foray into another new speed-multiplying protocol named HDSL, for High bit-rate Digital Subscriber Line. HDSL emerged as a way of delivering a new data transmission service named Tl that was even faster than ISDN. Where ISDN lines could provide customers with two voice or data lines at up to 128,000 bps, Tl offered up to 24 individual voice or data channels capable of streaming a blistering 1,544,000 bps. HDSL permitted the telco’s old-fashioned copper phone lines to carry this deluge of data for more than two miles without the use of expensive signal repeaters. Adtran’s first HDSL product family was introduced in late 1993, well before widespread corporate use of the Internet necessitated the need for a reliable medium for quickly sending oceans of data. When the corporate world began to demand high-speed, high-volume data for its internal “intranets” in the mid-1990s, téleos turned to producers like Adtran to come up with a way to more efficiently carry Tl services over their copper phone lines, which had never been designed to transmit so much information.
The solution appeared to be HDSL2, a Tl delivery technology that could transport Tl data within a telco’s Local Loop over a single pair of copper phone wires instead of the less efficient double pair required by HDSL. Adtran’s HDSL2 solution was named POET, and it promised to reduce signal interference and extend Tl’s effective transmission distance—thus making Tl service cheaper for Corporate America. Adtran licensed the POET technology to Siemens and adapted it for other emerging network data services such as Frame Relay (an offshoot of Adtran’s original data protocol, DDS), Fractional Tl (a Tl variant), and Primary Rate ISDN (an ISDN variant).
“The Right Place at the Right Time”: 1997
By mid-1997, tests had revealed that Adtran’s POET would not meet the performance specifications of all phone line regions, and such competitors as PairGain and Level One moved to establish another industry standard. Nevertheless, by early 1995 Adtran could claim roughly 15 percent of the $60 million HDSL-based product market in the United States. By 1996, hundreds of thousands of copper Tl lines were operating across the country, and HDSL’s share of the installation of these Tl services had grown faster than anyone had expected. The company responded by sinking larger sums into HDSL engineering and technology (for example, developing its own custom HDSL integrated circuits) with the end strategy of controlling as much of the HDSL market as it enjoyed in DDS and ISDN products.
Adtran’s runup since its 1994 initial public offering had been eruptive. By July 1996 its stock had peaked at $72 a share, up from some $15 in less than two years. A $50 million project to quadruple its Huntsville plant’s production capacity had been announced, by the end of the year sales had broken past the $250 million mark, 125 new products had been launched in
1996 alone, and over a million Adtran-based Local Loops were operating worldwide. The company’s investment in engineering and technology resulted in the release of 125 new products during the year. Moreover, only five percent of U.S. Local Loops were digital, suggesting that Adtran’s engineers had plenty of room to run. “Our growth had been fueled by the convergence to digital even though the percentage is still small,” Smith told Investor’s Business Daily. “What we’re doing now is taking advantage of those trends. It’s nice to be lucky enough to be in the right place at the right time. In terms of whether [Adtran’s success] is sustainable, that depends on how good of a job we do compared to everybody else in the business.”
Unfortunately, perhaps for Adtran “everybody else” included such telecommunications powerhouses as Lucent Technologies, Motorola, and Paradyne. Adtran had enviable market share but its head start on the competition was not insurmountable. Moreover, in 1996 two of Adtran’s customers—GTE and Sprint—accounted for 26 percent of its total sales. Worse yet, in 1996 the seven RBOCs claimed 36 percent of its sales, and the Telecommunications Act of 1996 once again allowed them to manufacture telecommunications equipment—Adtran’s customers could now become its competitors. With téleos still accounting for 60 percent of its sales (corporate end-users and OEMs claimed 28 and 12 percent, respectively), this posed no small threat. To stay ahead, Adtran had to keep rolling out the small, lightweight, sturdy products that had fueled its growth. In 1997, for example, it released its innovative Adtran Total Access System (Atlas) 800, a combination digital cross-connect system, ISDN access switch, Tl bandwidth manager, and remote access multiplexer rolled into one. It also had to penetrate the international market it had all but ignored until the mid-1990s. Although its international sales accounted for only 10 percent of total revenues in 1997 it was taking steps to corral overseas customers to the potential tune of 30 to 40 percent of total sales. Adtran had reached the corporate limelight by being the supplier of digital speed solutions to the copper analog telecom world; to stay there it would have to seize just such opportunities.
Breskin, Ira, “Speed Need,” Investor’s Business Daily, March 22, 1996, p. A4.
Elstrom, Peter, “Adtran Inc.,” Investor’s Business Daily, August 10, 1995, p. A5.
Finnerty, Brian, “Data Race,” Investor’s Business Daily, October 31, 1994, p. A6.
Fioravante, Janice, “Adtran Inc.,” Investor’s Business Daily, March 29, 1995, p. A6.
Robert, Don, et al, “No Longer Out of Reach,” Telephony, October 30, 1995, p. 36.
Veit, Stan, “What Ever Happened to ... 300-bps Modems?” Computer Shopper, July 1996.
—Paul S. Bodine