In
both cases, the applications are obvious, and the contribution is cost-saving
automation, more responsive customer service, or broader
availability—particularly when mobile—of the service.
A focus on familiar
applications makes sense in the early stages of a new technology. First, buyers
understand the application and can estimate the benefits quantitatively. Second,
callers are familiar with the service and have an easier time dealing with a new
user interface than if everything were new.
Unfortunately, the
familiarity of the applications makes advanced speech technology seem merely an
enhancement rather than a fundamental new opportunity. By contrast, the Internet
was viewed as a critical new technology for companies to address at the highest
level. (The enthusiasm rose to extremes, but nevertheless reflected a real
opportunity.)
The Web did more than
extend existing applications. For example, information that might otherwise be
delivered as a paper document is much more accessible as multiple Web pages with
links. Finding information on specific products or services is fairly easy with
Web-search engines. Many customers use the Web as a primary source of product
comparisons and sources, and many buy online. Does the telephone voice user
interface have similar applications that constitute truly new opportunities?
Conversational marketing
This editorial suggests one
category of such applications. The telephone with speech recognition should be
viewed as a new media, one which addresses a number of marketing issues that
experts have highlighted in research and books. I will refer to this opportunity
as conversational marketing. Conversational marketing is a systematic
attempt to make the most effective use of every customer call by involving the
customer in a dialog which advances the organization’s objectives.
Most current call-center
applications fit into this framework by improving the customer experience
relative to touch-tone or a long wait for an agent, but don’t generally attempt
to integrate the applications with other company marketing. Today’s applications
are generally targeted more at shortening the customer interaction for cost
reduction, rather than making the most of the call.
The problem
A number of marketing
writers and researchers have noted important changes in the marketing
environment:
• Dilution of marketing channels: There are
more ways to reach the customers, including many more television channels,
targeted magazines, and the Web. Customers are deluged with marketing messages.
Seth Godin and Don Peppers, in Permission Marketing, 1999, refer to “the
attention crisis in America…It’s just physically impossible for you to pay
attention to everything that marketers want you to.”
• Resistance to “interrupt” marketing:
Customers are less responsive to classical advertising that interrupts what they
are doing to present a message—e.g., TV or magazine ads and Web banner ads or
pop-up ads. Digital Video Recorders, which can fast-forward through an
advertisement, compound this problem.
• Tailored marketing: The desire to make
marketing more relevant to the consumer’s general or short-term interests.
• The need for marketing messages to entertain: As
consumers get more jaded, the marketing message must compete for their
attention. Yet, at times, the marketing message is lost in the entertainment.
• The need to retain customers as well as get new
ones: Studies have show that companies can boost profits 25%-85% by
retaining 5% more customers (“Zero Defections: Quality Comes to Services,”
Frederick F. Reichheld and W. Earl Sasser, Jr., Harvard Business Review,
September–October 1990). Part of marketing should be keeping customers who are
having problems, as reflected in many call-center and technical-support calls.
The power
of dialog
One way in which marketing
experts have summarized new marketing trends is by saying that companies need to
hold a “dialog” or “conversation” with their customers. For example:
• “One-sided bombardment is replaced by dialog.” (The
New Marketing Era, Paul Postma, McGraw-Hill, 1999)
• “Markets are conversations… Companies can now
communicate with their markets directly. If they blow it, it could be their last
chance.” (Christopher Locke, The Cluetrain Manifesto, Perseus Publishing,
2001)
Interestingly enough, these references to “dialog” or “conversation” never
include speech technology. At one extreme, they refer to very slow dialog:
targeted direct marketing by mail, and the customer replying to an offer. A
faster dialog is customer interaction with Web sites; but, in this case,
experience shows that a customer must be sold within about three pages of
interaction, or their interest drops rapidly.
Obviously, speech
recognition systems support dialog, and the nature of speech is that it can
support many turns back and forth within a conversation without losing the
caller. Customer information garnered from multiple calls by one customer
(identified, e.g., by an account number) can be accumulated and reflected in the
dialog. Even if the caller’s interests are unknown to the system prior to the
call, speech dialog can support finding out the caller’s interests and
preferences even within one call.
Callers can even relate to
an automated “persona,” as speech vendors have been calling the implied
personality of the automated system. Some companies have given their automated
systems an identity, e.g., Bell Canada’s “Emily” and Amtrak’s “Julie.”
Dialog with customer
service representatives (CSRs) can be effective precisely because dialog can be
so efficient at determining the customer’s needs quickly. Unfortunately, much of
the time agents spend in dialog is getting information that doesn’t require a
person—information that is easily gathered by a speech recognition or speaker
verification system, such as account numbers, validating identity, or the
product the customer is inquiring about. Because they spend so much time
simulating a speech recognition system, there is a high agent turnover and
boredom can often lead to uneven customer service. Speech recognition which
preserves the more interesting parts of a conversation for agents contributes
directly to agents being more effective in conversational marketing and makes it
more economically feasible for them to do so.
Making the
most of a call
The cost, however, of
handling every call with an agent tends to lead to an objective of keeping the
call short. Yet, a call to a company or information service is an unparalleled
opportunity to further the company’s relationship with the caller. When a caller
phones an automated system, they are usually interested in a product or need
help with a product. They are volunteering their time (a criterion of
“permission marketing,” to use a current term). The customer should have an
experience with the company that (1) is of interest at that time to the
customer, based on what they say; (2) solidifies the relationship with the
customer by entertaining, informing, or helping them; and (3) extends the
relationship, potentially making a sale during the call or getting permission
for further contact (e.g., by sending information through email).
A well-designed speech
recognition system contributes to conversational marketing in almost all cases
where it replaces an unfriendly touch-tone interface. At this level,
conversational marketing is a matter of good voice user interface (VUI), persona
design, and system design. Fortunately, we are seeing more of this enlightened
style.
The
telephone as media
Good design alone, however,
barely addresses the full opportunity. Exploiting the full potential of
conversational marketing requires new applications designed for that purpose.
Those applications involve significant creativity, including the involvement of
people who specialize in marketing (such as marketing departments and ad
agencies). Ideally, marketers should view the telephone as a distinct media,
like print, TV, or the Web. Every call should be considered a marketing event,
and marketing in other media can easily be integrated with telephone marketing,
as is done when a toll-free number is included in an ad.
Achieving the full
potential of conversational marketing will take time. First, some good examples
are required in order for marketing executives to understand the expanding
potential of the telephone. Those first few examples will probably require a
careful collaboration of marketing and VUI experts, along with careful
integration with supporting database systems. The long-term growth of
conversational marketing will probably require the development of tools that
don’t require an expert on dialog design or programming.
The full power of
conversational marketing using speech technology requires the integration of
many functions within an organization—the call center, Web management,
information technology, technical support, and marketing. That won’t happen
easily without top management recognizing the need for a corporate telephone
strategy, just as many corporations have a Web strategy.
Currently, most telephone
speech applications replicate what is being done less efficiently and less
flexibly by other means. Conversational marketing takes telephone speech
recognition into a more unique realm. It unifies existing applications and puts
them into a more strategic context. Conversational marketing has the potential
to make the telephone a central part of a company’s marketing plan and merits
attention at the top ranks of companies.
Copyright
2003 William Meisel