|
Still alive, if only in spirit, Procrustes would appreciate contemporary
notions of interactivity, or rather what passes for interactivity, whether
by way of the phone or the Web. Although Procrustes didn't run a contact
center or manage a Web site, this character from Greek legend did
entertain guests within his roadside villa. The guests, according to the
legend, were promised a pleasant meal and a night's rest in a remarkable
bed, which, they were told, would match their length exactly. Before long,
Procrustes' unlucky guests would discover their host's true intentions: he
meant to adjust the length of his guests, and not that of his bed. Taller
guests would have their legs chopped off, and shorter guests would suffer
the rack.
Strange hospitality, I'm sure we'd agree. But while we're most likely
horror-struck by the plight of Procrustes' guests, most of us are unmoved
by the plight of customers who are commonly required to adjust their own
means of expression to accommodate the limitations of a service interface.
Thus the strange hospitality of legend meets the strange convenience of
contemporary commerce.
THE GREAT TRADEOFF
How is it that we tolerate interface limitations so well that we hardly
know we're doing so? For one thing, we don't know what we don't know. For
another, we recognize, consciously or not, a tradeoff we're willing to
accept. The tradeoff, in this case, is the inconvenience of communicating
by way of mouse clicks or keypad presses, for the convenience of
collecting information in a format that is easy to track and quantify.
Also, a stream of clicks or a sequence of keypad presses may be channeled
through a more or less rigidly designed Web site or IVR menu -- so much
easier than channeling, tracking, or quantifying a real dialogue, a
conversation of the sort ordinary human beings negotiate all the time.
But, in the end, is the tradeoff acceptable? Before answering this
question, we should consider all that may be lost when we insist on a
reductive approach to communications. That is, communications reduced to
clicks and presses. What nuances might be lost? And what confidences might
be withheld?
Imagine a poker game reduced to the sparse inputs and outputs that
would be convenient for an ordinary computer. The computer would follow
the rules flawlessly, and execute sensible bids by keeping track of all
the cards on display. But what if the computer couldn't detect a bluff, if
it failed to recognize and account for every pause and gesture, every
change in its opponent's tone of voice? What sort of game would that be?
And what about other sorts of interactions involving human beings? What
negotiations, what demonstrations, what transactions? All of these
interactions could suffer, including those dedicated to developing
enriching commercial relationships.
Such interactions are often managed within contact centers, thanks to
the presence of a contact center agent, a sort of interface, if you will,
between a customer and all the technology of commerce. This is automation?
A human-mediated interface introduces the risk of inconsistency and error;
moreover, it's expensive.
What sort of interface might be better? For the moment, let's leave
aside practicality. Let's describe the ideal. I'd say the ultimate
interface would be a holographic avatar, something that had all the
appearance of being human, but was actually entirely computer generated.
The avatar could assume the most effective appearance in any given
situation. Also, the avatar wouldn't miss any nuance of a customer's
communication, and yet the avatar would meditate the dialogue according to
a given set of rules, which could be continuously refined, depending on
the data collected and the results of data analysis.
ADDING MORE BEDS
How far from this ideal is an IVR menu or a browser window? So far that we
might well find the comparison humbling. We have been making progress,
however, thanks to early efforts in customer relationship management
(CRM). With CRM, we try to impose a guiding intelligence in all customer
interactions. Also, we try to impose this intelligence consistently,
across all the different communications channels a customer might use.
But before we congratulate ourselves, we might return to the example of
Procrustes. In an alternate version of the Procrustes legend, the
monster's guests were given a choice of two beds, one of which was quite
small, and the other quite large. Thus, the guest could choose between
being stretched or being ... truncated. In a sense, CRM does the same
thing. It adds more beds -- or, if you will, more communications channels,
which might include e-mail, chat, click-to-talk voice over IP,
conventional voice, Web browser, video kiosk, point of sale, etc.
All this is no mean feat. CRM has taken great pains to minimize the
lack of coordination between communications channels. Thus, despite a
proliferation of channels, a corporation interacting with customers may
impose a consistent set of policies guiding all of its interactions.
CRM can become incredibly elaborate, spanning not only multiple
communications channels, but multiple contact centers and multiple data
repositories. An integration nightmare? Yes, often it is. But wait ... it
gets even worse!
THE CRM CYCLE
If it is to achieve its avowed purpose, CRM must ultimately institute a
continuously self-refreshing cycle, from customer interactions to data
repositories to analytical engines back to customer interactions. CRM must
impose a consistent means of collecting and manipulating data, so that the
manipulations may yield hints as to what action would be best at any given
time with any given customer.
It is at this point that we may recognize CRM's relationship to
enterprise resource planning (ERP). ERP represents a collection of
processes in which data is collected, stored, and analyzed, so that
knowledge may be generated, knowledge that may guide decision making. ERP
has its roots in Materials Requirement Planning (MRP), which was developed
to mediate the procurement of natural resources and production processes.
ERP takes the logic of MRP further, into less physical and more
intellectual activities, including the mediation of human and financial
resources. And now, with CRM, ERP would apply the logic of automation to
even more complex systems, that is, systems that would mediate long-term
commercial relationships between suppliers and customers, as well as
between partners.
WHY ENCOMPASS ERP?
We've been talking about the limitations of a reductive or Procrustean
approach to communications, and about how these limitations may frustrate
customers, stretching their patience to the limit. (Incidentally,
Procrustes means "he who stretches.") We may also consider how
corporations that provide or deploy CRM solutions may also feel stretched,
given prevailing business trends.
These trends point towards the virtual corporation, an entity that will
severely test nascent ERP and CRM capabilities. The virtual corporation
presents an alternative to the vertically-integrated corporation. The
virtual corporation is, in essence, a group of smaller corporations that
work together, forming a continuous value chain. The virtual corporation
would attempt to get the best of both worlds. First, it would realize the
benefits of granting autonomy to each value-chain constituent. Second, it
would preserve the benefits of order and predictability, which heretofore
required the rigid hierarchies of vertical organization.
The virtual model should appeal to engineers. It applies the
time-honored engineering technique of breaking up a problem into
sub-problems, each of which may be attacted independently. The virtual
model should appeal to market purists, who could forsee the more difficult
challenges being isolated within one or another horizontal layer, so that
they might more easily succumb to the efforts of competitors within a
given layer.
The challenge, however, is imposing a sense of coordination among the
constituents of a virtually-integrated value chain. To achieve such
coordination, to permit a group of peers to act as though they formed a
unified entity, will require no little sharing of information, no little
communication. It is here where the virtual corporation and its
constituents will feel constrained by a reductive, Procrustean approach to
communications. When the boundaries blur between customers, suppliers, and
partners, the constraints that were once felt only by customers will be
felt by all.
NATURAL COMMUNICATIONS
With so many depending on communications, does it make sense to
constrain it, if we can avoid doing so? Consider the case of the dot-coms.
To what extent may we attribute their difficulties to a constrained sense
of communications, to a habitual reductiveness which would ignore all the
potential richness of communications, and instead embrace a sequence of
mouse clicks? If dot-coms and others must compete for the hearts and minds
of customers and partners, doesn't it make sense to make communicating
with these customers and partners as natural as possible?
But if we are to promote the idea of natural communications, we must
confront a familiar challenge: how to mediate such communications
effectively, how to collect data from such communications that will drive
the cycle of continuous improvement, to and from the data repository.
This challenge is being approached from several directions. For
example, CRM is improving in several dimensions, encompassing more
communications channels, spanning more switching and routing platforms,
embracing increasingly sophisticated data manipulation schemes:
- At the level of communications channels, CRM applications present
customers and agents a consistent interface across what might
otherwise appear a motley collection of point solutions. Examples of
companies active here include Altitude,
E.piphany, Nortel
Networks/Clarify, PeopleSoft/Vantive,
and Siebel Systems.
- At the level of switching and routing platforms, infrastructure is
increasingly able to support CRM applications, including those relying
on seamlessly negotiating the boundaries between traditional
circuit-switched networks and next-generation, IP-based networks.
Examples of companies emphasizing this aspect of CRM include Alcatel/Genesys,
Aspect, Apropos,
CELLIT Technologies, CosmoCom,
Interactive Intelligence, Telephony@Work,
and Quintus.
- In the data manipulation dimension, CRM is emphasizing data mining,
knowledge development, and customer analytics. Examples of companies
emphasizing this aspect of CRM include Athene
Software, Applix, Baan,
Broadbase, Hyperion,
IBM, Paragren,
thinkAnalytics, and
Xchange.
To date, the importance of less constrained, more natural
communications is better reflected at the application and infrastructure
levels, as evidenced by Web-enabled call centers, voice-enabled Web sites,
click-and-mortar schemes, and converged networks generally. With CRM in
these dimensions accommodating richer communications, the remaining
dimension -- that of data manipulation -- will feel increasing pressure to
accommodate richer communications as well. Already, the familiar names in
the CRM application space cite various forms of data manipulation within
their software suites. Also, companies specializing in data mining,
knowledge development, and analytics are forming partnerships with CRM
solutions providers.
ENTER SPEECH REC
And now, if I may indulge in a little speculation, I'd say it won't be
long before we hear about increasingly ambitious CRM schemes involving
speech recognition, specifically, natural language speech recognition.
With this technology, natural spoken language may be parsed, yielding
keywords which may guide further interaction, particularly within a system
taking advantage of artificial intelligence (AI) technology. Chopping up
utterances, while stretching AI to make the most out of a collection of
keywords, may itself seem Procrustean, but it's a start. If sufficiently
refined, such approaches, in combination with text-to-speech, could
convince a caller that he or she were talking to a live (and very
knowledgeable) operator. We'd be that much closer to the ideal avatar
which we discussed earlier.
But we needn't let such speculation carry us away. We have more than
enough immediate challenges. Chief among them is the challenge of
deflating overblown talk about interactivity. Such talk is all too common,
leading otherwise sensible people to ignore the Web's current limitations.
To keep your senses, just apply the example of Procrustes. Remember, he
promised a perfect fit between guest and bed, but by cutting or stretching
the guest to fit the bed. Similarly, the Web promises wonderful
interactivity between humans and data structures, but by limiting the
means of interactivity, the means of communications, to mouse clicks.
Thus, the human element accommodates the data structure, and not the other
way around.
Surely, we can insist on something better. Otherwise, we may well ask
who is serving whom, notwithstanding all this talk about better customer
service. Let's be sure customer service is worthy of the name, for it is
through superior customer service that Internet-age creations such as the
virtual corporation actually work, exercising knowledge about customers,
partners, and suppliers -- not to mention competitors -- so that
value-chain constituents may create and defend advantageous positions for
themselves. Let's encourage CRM and ERP to accommodate market constituents
to the fullest extent possible, and that includes accommodating natural
communications.
[ Return
To The February 2001 Table Of Contents ]
|