AT&T And Data Communications (Circa ’70s-’80s)

Data communications began in the 1960s, when businesses, including telephone companies, began to purchase and integrate computer hardware and software into operations, and then began to use communication links to transport data between computers. Many at AT&T recognized this in the middle 1960s, and AT&T first developed private-line modems at 1.2 kbps, a rate that climbed to 9.6 kbps in the 1970s. AT&T also offered private-line point-to-point analog links for rent, for point-to-point private networks for business and public service sectors.

In 1974, AT&T launched its first digital tariffed service, Dataphone Digital Service (DDS), offering point-to-point and point- to-multipoint polled digital transmission at 2.4 kbps. 4.8 kbps. 9.6 kbps. 56 kbps, and 1.536 Mbps. In 1975, Bell Laboratories began studies on how to deploy a circuit-switched multirate digital service, and in February 1976, a development organization was chartered to develop the necessary hardware and software.

The vision was overarching: Computers would connect via a standard telecom industry digital interface, via a wall jack with a plug, at say 2.4 kbps, with a special data communications number (just as all telephones have numbers). A supervisory signal would request that a path be set up through the network to another data communications number operating at the same 2.4 kbps. If the request for a path were successful, the two computers could then send bits back and forth, using software resident in each computer to control this interchange, over an alldigital, reliable and ubiquitous network.

Two-wire digital transport at 2.4-9.6 kbps was envisioned, at increasing price points. Higher transmission rates-e.g., 1.536 Mbps, for the first point-to-point digital transmission system deployed by the Bell System in 1962 (T1)-would require four copper wires, using one pair to transmit and one pair to receive signals, and would be priced higher. Over time, speeds would increase from tens of kilobits per second to tens of megabits per second, all circuit switched, and cost would be driven out of equipment to provide high bit rates for longer time durations for lower costs, such as had happened in voice telephony.

If this had been launched and supported by the Bell System, computer vendors and third parties would have developed hardware and software to take advantage of it; simply having as a standard connector the RJ-45, cousin to the RJ-11 voice telephone jack, would have saved hundreds of millions of dollars per year in communication equipment costs! Directory services, billing, provisioning, operations, traffic management, fault management, security management-all would have been addressed within the existing organizational structure of AT&T and its operating companies.

So data communications was not something that people in the Bell System were unaware of. What happened to keep AT&T from offering switched digital communication services connecting computers and terminals throughout the 1970s and into the 1980s? AT&T’s internal organizational dynamics.

In 1976, the switched DDS development team at Bell Labs ran into a new organization at AT&T: General Departments, a marketing group headed by Archie McGill. A former IBM VP, McGill attracted a very capable staff, many from IBM, to join him in transforming AT&T into a customer- or market-driven organization, in contrast to being an engineering- or product-driven organization (as AT&T then was).

The switched DDS development team’s vision was technically clear, so Archie demanded that a business case be prepared to justify the new service. Archie and his team argued that the economies for the new service, assuming different pricing scenarios and different subscriber penetration rates, showed that the offering at best was marginal measured on internal rate of return (IRR). This was due to two factors:

* The circuit switching technology (4ESS), had just been launched and was priced high early in its product lifecycle to recoup development costs.

* The price per port of the digital interface was high (this was before integrated circuitry entered into the picture to drive cost out of both the access and the switching system).

Having shot down the DDS plan, Archie and his staff embraced an alternative and marketing-driven vision of data communications. In this vision, one connected all terminals and all computers together, using minicomputers from Digital Equipment Corp.(DEC) running Bell Labs Unix system software with applications put together by Bell Laboratories development teams.

The business case assumed customers would store large amounts of their records on computers in the AT&T network-in the face of the reality that many businesses were moving to minicomputers and that PCs were just then emerging (Tandy TRS 80, Apple Computer Apple II). Businesses were doing this to gain control over their information, not to give it up to central data processing organizations.

In 1978, AT&T filed with the FCC a Notice for Advanced Communication Services (ACS), a new service that would connect all computers and all terminals. This was not a tariff: there were no rates or prices in this document, but it did alert the entire global computer industry that AT&T was up to something big. IBM, DEC and HP, among others, immediately reviewed this filing, and internal groups at each company subsequently concluded this was technically infeasible at any type of price point for service that customers would accept.

In the spring of 1979, Archie pushed for an internal review of ACS, wanting to launch this as a service as soon as possible. The internal review suggested that it was premature to launch a service and that additional work had to be done. Nevertheless, in mid-1982, AT&T decided to launch this new service through an FCC-mandated Fully Separated Subsidiary (an oxymoron, but what the hey) of AT&T named American Bell. On Jan. 1, 1983, American Bell was launched, with its flagship new data communications service-for a service that, it turned out:

* ignored the PC/LAN revolution and the resulting dispersal of intelligence to the edge of the network; and

* had a price that was far too high to gain primary demand penetration.

As a result, the business went nowhere and it was shut down in the mid to late ’80s.

Various sources estimate that AT&T spent in excess of $4 billion from 1976 to 1982 on all of these different data communications activitiesa

Copyright Business Communications Review Mar 2005