|
|
Gregory R. Tatchell
Stentor Resource Center, Inc.
4535 Canada Way
Burnaby, B.C., Canada, V5G 1J0
604-654-7486, Email address: GRTatchel@Stentor.ca
The existing telephony customer interface is based on the input
of commands to the telephone system via the touch tone pad of
a standard telephone set. The services activated by these commands
are hard to learn, easy to forget and confusing because of this
restrictive user interface. A solution to ease of use problems
(and the 30 others associated with the existing touch-tone telephony
interface) requires a new interface. The success of the graphical
user interface in the PC industry suggests the following question
for the telephone industry: "Is it possible to achieve
the same quantum increase in ease-of-use with the customer interface
in the telephony industry as was achieved in the PC industry?”
This and four associated mini-papers explore both the problems
that exist and solutions that can be considered.
Phone-based interfaces, Intelligent Agents, Personal Agents, Voice
Recognition.
To fully understand telephony customer interface problems and
limitations we looked at the codes, rings, tones and announcements
that a customer encounters during the activation, deactivation,
confirmation, notification and in-session prompting of various
network services. This was done for 5 service classes across 4
access environments and summarized in a matrix. The matrix represents
the actions (codes, rings, tones and announcements) the subscriber
(caller and called parties) must take and/or hear before and during
a call. Analysis of the matrix revealed 31 problems with the existing
interface, from a customers perspective. These 31 problems will
be used as the measuring stick against which to gauge the success
of potential solutions to the existing interface problem.
A strategic context for addressing telephony customer interface
problems is provided by looking at the financial trends of the
Canadian telephone companies in the last 25 years. Since the 1970's,
toll revenues have always exceeded local revenues. Since
the mid-80's, however, because of competitive pressures on the
toll side, and increasing revenues on the local side, the gap
has progressively narrowed. In 1994, an event of strategic significance
occurred concerning the relationship between these two revenue
classes in Canada; for the first time in over 25 years, as shown
in Figure 1, local revenue exceeded toll revenue in parts of Canada.
Further, the local revenue line is comprised of both wireline
and wireless revenues. As a component of total revenues, wireless
is still very small. Significantly, though, the sale of wireless
lines in Canada in 1994 exceeded the sale of wireline lines. It
is in the context of this double cross-over (of local over toll,
and of wireless over wireline) that we contend deliberations on
the telephony customer interface should be considered.
A second perspective on the customer interface is its technical
evolution over the last 50 years. Figure 2 illustrates the evolution
of the telephony customer interface since the end of the Second
World War. It has two objectives; one, to show how the interface
has evolved over fifty years, and, two, to illustrate that the
customer interface has come full circle in that time.
The wall phone was the first of four generations of customer interfaces
that have graced the public telephone network in the last 50 years.
The rotary dial and touch-tone phones represent second and third
generation, while a variety of different and evolving CPE's (ADSI
sets, Voice Activated Dialing sets, PC's, etc.) represent the
fourth.
Communication using the first generation wall phone was natural and personalized; natural in that all interactions (control and conversation) were with the spoken word; and personalized in the sense that the operator often knew members of small communities quite well, much like answering services do today. We concur with AT&T's position that the use of intelligent agents is, perhaps, the best way to economically return to this personalized, intelligent service in the future.
The intelligent agent implies a fresh, new, fifth generation customer interface for the public telephone network. We believe the first phase of implementation of the customer interface should have three design characteristics:
Note that these three were also characteristics of the first generation
customer interface, albeit with a different implementation. With
the introduction of the fifth generation intelligent agent, the
customer interface will have come back full circle upon itself:
the wall phone serves as a metaphor to remind us of that fact.
The three design characteristics represent an initial implementation
over what will be multiple generations of an evolving telephony
intelligent agent. Intelligent agent design will build upon two
essential and evolving characteristics; degree of agency, and
degree of intelligence. Figure 3 provides a long term context
in which to view the evolution of the intelligent agent (personal
agent) in a telephony customer interface context.
The 'Agency' aspect is represented on the vertical, while
the 'Intelligence' aspect is represented on the horizontal.
Time frames are provided along the top horizontal axis to indicate
when increasing levels of agency and intelligence are likely to
be viable in the telephony customer interface context. The diagonal
line integrates the three and hypothesizes a telephony intelligent
agent evolution path.
Three critical success factors exist for intelligent agent design in the telephone customer interface:
Three associated mini-papers follow this one and address these
critical success factors in turn.
We believe that an affirmative answer can be given to the question
posed at the beginning of this paper; i.e. "Is it possible
to achieve the same quantum increase in ease-of-use with the customer
interface in the telephony industry as was achieved in the PC
industry?”. It will be achieved by augmenting the existing
touch-tone interface with a new, fifth generation, intelligent
agent, telephony customer interface. It will be developed by properly
identifying communications needs of consumers and designing an
interface that ensures natural, easy-to-use, personalized communications.
As it is implemented, it will eclipse touch-tone and dial-tone,
much as touch-tone eclipsed (although not totally replaced) rotary
dialing.