by Glen Taylor
I’ve had many paying jobs over the past sixty years. During that time, I’ve worked for and worked with some amazing people. However, the high point of those sixty years were the twenty-one years I worked for Bell Labs. Of those years, the thirteen that I worked in Conversant R&D in Columbus, OH, were the apex.
I had two short careers prior to joining the Labs. I spent three and a half years in the Navy. And, I spent seven years as an academic culminating with one year as an Assistant Professor at Texas A&M University. In 1979, I joined Bell Labs in Holmdel, NJ, as a human factors specialist. My academic background included a PhD in Experimental Psychology, and during graduate school, I’d co-enrolled in a Master’s program in Computer Science. In late 1978, I had decided to change direction once again. At the time, my impression of Bell Labs was similar to the popular image. I was aware that Bell Lab scientists had won Nobel Prizes and invented such things as the transistor and the laser. When I was encouraged to apply, I had some trepidation. I knew my accomplishments weren’t of that caliber.
What I didn’t know, but would soon discover, was that a research and development (R&D) institution like Bell Labs resembles the fabled iceberg. That extremely visible and impressive research component is just the visible tip of a much larger iceberg. The nationally known research area was perhaps 10% of Bell Labs with the remaining 90% “invisible below the surface.” Those were the development areas that were responsible for creating the hardware and software products used by the AT&T long distance business, by the local telephone companies, and by consumers. To be clear, the most qualified candidate possible was sought for every job, but not every Bell Labs employee was just waiting his or her turn to receive a Nobel Prize.
For established scientists and engineers, their prior work played a major role in the admission process. However, many of the young engineers that Bell Labs hired were straight out of school. Historically, Bell Labs hired PhDs, although for hardware and software jobs the engineering and computer science graduates might only need a Master’s degree. Some very bright Bachelors graduates might be hired and possibly sent to get their Master’s degree at company expense; a program called One Year on Campus or OYOC (pronounced “oh” “yock”). As a “soft science” candidate, my required entry credential was a doctorate. I was a bit more than a recent graduate having done two years postdoctoral research and taught for a year. I had published ten refereed journal articles and had a research monograph in publication. Those were helpful, but it was the combination of my computer science and psychology research backgrounds that seemed to attract attention. I would interview with thirteen different Bell Labs supervisory groups in five locations across three states. The job I selected as my first choice hired me, and I started in Holmdel in June 1979.
Holmdel was an interesting place. The building was a huge, rectangular solid whose exterior walls were a smoked glass. The interior was four multi-story towers containing a warren of cross aisles connecting to an exterior walkway along the glass outer wall and an interior walkway overlooking an atrium area on the ground floor. The exterior walls were the only windows. Every building looked exactly like every other building. It was very easy to get lost.
Bell Labs had a somewhat schizophrenic relationship with human factors or human engineering. On the one hand, the company well understood the need to develop products and services that were easy to use and satisfying to its customers. It usually employed experimental psychologists for this work because we were trained in empirical observation and research design. We generally were invested in understanding how people learned and how they performed various tasks. I and my peers who hired in as human factors specialists were led to believe that we’d employ those skills doing applied research. We’d design test scenarios where user “subjects” would interact with the user interfaces we were evaluating and engage in tasks using the systems under observation. That would allow us to determine what was easy, what was hard, what led to errors, what led to correct behavior. My initial work in Holmdel actually did some of this, and we published that work in the Bell System Technical Journal. However, that honeymoon was soon over.
Although the company understood at some level what was required for “good” human factors work, it was unwilling to spend the time and resources. Engineers, mathematicians, and computer scientists are all trained that there are well known methods and procedures for most work. Yes, research goes on in every area, but development is the application of learned principles to the creation of products and services. When a software designer came to me with his or her interface design, the question was “How should I have the user do X?” My response would be, “I don’t know the best way to do X, but give me the time to conduct a small study and I can tell you.” That created a fundamental tension. As pressure to develop code grew and deadlines loomed, the software developers would soon come in with their designs and simply say, “Here’s the design. I need you to ‘bless it’.” I believe nearly every one of us who was hired into human factors jobs in that late 1970s & early 1980s period would soon go elsewhere. Some left Bell Labs. Most of us migrated to other “jobs.” I became a C programmer for a year or so.
My Holmdel project was high pressure. We were building the first product that AT&T could not deliver as a regulated entity. We would have to be spun off into an unregulated subsidiary. On July 1, 1982 my project, about a thousand people across several functional AT&T organizations, became the nucleus of a new company, American Bell. When the pressure became unbearable and morale was at its lowest point, I escaped NJ to Indianapolis, IN, and returned to a human factor’s role.
Indianapolis was the AT&T Consumer Products division. It designed telephones for home and business. It had a “lab home.” This was a building whose primary rooms were designed and furnished like a consumer’s home. The “home” rooms such as the kitchen, living room, dining room, and bed rooms were surrounded by other spaces that contained equipment for observation and testing. Indy was working on “smart house” concepts including security systems, videotex terminals that would connect to a standard television and telephone line, etc. This lab made actual experimental observation possible. I would, once again, have the opportunity to do some amount of applied research, but it was still clear that the company had only a limited appetite. I made the transition that was probably the most common for human factors psychologists, and I became a systems engineer.
It would be as a systems engineer that I would transfer to Columbus, OH, in January 1987. There was a fairly large exodus from Indy at the end of 1986, and five of us would end up in Columbus working on the Conversant project: Skip Tourville, Bill Erwin, Dennis Noonan, John Siroky, and me. Skip and I were systems engineers although from much different backgrounds. Skip was an engineer by training. I would be joining my colleague Steve Riederer as an ex-“human factors” systems engineer. Although there were individual differences in our personalities, those of us with the psychology backgrounds tended to maintain a fundamental perspective looking out for the end user that not all of our development or even systems engineering colleagues had.
I quickly felt at home in the Conversant organization. I’d worked in an emerging venture my last year in Indianapolis so the venture “vibe” was very comfortable. The entrepreneurial spirit of Conversant in 1987 was infectious. Having all of the core business functions co-located within easy walking distance and having the opportunity to see all aspects of the business were eye opening. Instead of feeling like a very small piece of the hundreds of thousands of employees that was AT&T when I joined, I was an integral piece of a business of less than a hundred people (or so it seemed).
My focus almost from my arrival in Columbus was directed toward positioning (“selling”) Conversant into the AT&T and local telephone company networks. Other colleagues would be more focused on other customers. It would be several years before I would actually recognize my work as a form of “sales” work. Once that clicked, I began to realize how much value there was interacting with customers and how much I enjoyed it. I would transition around 1996 from my network projects to more of a focus on the Conversant product itself. I had the privilege of using all of that exposure to our customers to gain a deeper understanding of what Conversant needed to do. I could then translate those into the requirements documents that would guide our release developments. I was deeply involved in Conversant V6 which was the “international” release to many countries around the globe. Although I didn’t get the opportunity to globe trot, I had the extreme pleasure of working directly with Linda Dotts as the product manager and George Hsieh as the development manager. The three of us formed the nuclear “core team” responsible for getting that release out the door. I would continue to be the primary systems engineer for the V7 and V8 releases.
All good things must come to an end. That’s the old aphorism. The year 2000 saw the end of my time in Bell Labs. AT&T had already broken up leaving Bell Labs and Western Electric to form Lucent. In 2000, Lucent would spin off the “enterprise” business – PBXes, contact center, and contact center adjuncts such as Intuity Audix and Conversant. Lucent would retain its “network” business. At that point, Conversant development would transition to Denver where the bulk of the “enterprise” R&D work was done. Most of my Columbus R&D colleagues would be reabsorbed into a Network Systems organization that remained in Lucent. I was given the choice of transferring to Denver to go with Conversant R&D or remaining with the Columbus Network Systems R&D. I chose neither path. I wanted to stay with the Conversant product and the IVR business and make the transition to the new Avaya, but I wasn’t prepared to move my family to Denver. I left R&D and moved into a technical sales specialist role. I would remain in that role by various titles for nearly a decade.
Leave a Reply