The Web & Call Centers

By Keith Dawson of the Call Center News Service

Not that many years ago, the idea of a connection between the internet and call centers was theoretical, and not very well thought out. The most common way of thinking about the issue was to imagine self-support applications: literally, e-mail-based posting of customer support cases, and the ability of a customer to search a database of problems and solutions by themselves.

Just a few years ago, it was possible to talk about the different ways of delivering customer service without mentioning online support at all. When you did mention it, it was in the context of CompuServe forums or company-sponsored bulletin board systems. The Internet explosion that's brought us Web pages was still a year in the future. That's how fast things have changed.

A call center is not a place. It is a set of functions. It is the process of selling to people who are not in the room with you. And of serving their many varied needs. A call center's primary function is to create and keep customers.

Right now, that function requires a physical presence, a location -- an actual set of seats filled with people to help customers. Agents that have access to stores of data about the customers and the company, and the points of intersection. These people act as gatekeepers for the two-directional flow of information -- as intermediaries and interpreters.

What if those functions could be accomplished without people? Or with a vastly fewer number of people, so that those who are left are true experts who add value to the transaction, and who don't merely funnel that transaction along.

This is the goal experts are reaching toward when they tout the Web as a useful adjunct to the call center -- the promise of a workforce deployed in exactly the way that is most useful and efficient. Customers who solve their own problems, who in essence sell themselves, and a specially trained cadre of agents dedicated to doing what only humans can do.

To some extent, the call center industry has been flirting with this notion for years, with varying degrees of success. First there was fax, then fax-on-demand. Want information about our company at 2 am? Rather than pay to staff off-hours, make information available for retrieval by the customer himself. It's fast, cheap and gets generally high satisfaction marks from the customers.

Then there's IVR. It's great for routing calls to agents, shortening call times, getting people into the right queue, etc. But it also allows people to self-serve for simple database lookups like an account balance, an order confirmation or shipping status. Or to diagnose a technical problem.


Again, the trend is toward providing an automated response to customer interactions. The reasons are clear:

* Automated responses are cheaper than agent-provided ones.

* They are always the same for all callers. Two people who call for directions from the airport to your office won't get different routes from different reps.

* Automation is always available, even when you're closed.

Of course, there is a downside. Some people miss the personal touch. And any problem or question not planned for in the rules-based structures of your system requires human intervention anyway.

The Web is the third advance in self-serve automation. It has many of the advantages of fax -- it's a dynamic, easy-to-maintain format. It is available to huge numbers of potential customers. And it surpasses fax or IVR in one critical area: it has enormous multimedia capabilities (sound, graphics, video, and more). It is rich in the one quality IVR lacks -- the ability to control applications that require visual presentation or extensive keyboard output, or both.

Customers can order products. They can download software. They can read catalogs.

Clearly there is a need for alternate points of entry into the call center. You can think of the call center as the focal point of a "customer contact zone" where a lot of interactions take place. Ultimately, as more kinds of call center applications are developed that bypass the agent, a given customer will have more choices for entering the zone and concluding the interaction. Some points of entry may be better for making a sale -- document retrieval by fax-back, for example. Others are better for customer support, like IVR.

The Web offers an amalgamation of these techniques. Where the Web first began to take hold was with help desks. These smaller centers, already strained by escalating call volumes, were in the vanguard of agent-enhancement technology. Problem resolution software frees technical experts from the drudgery of answering repetitive questions, letting them get to the business of solving more complex problems. It puts them in a position to add value to the customer transaction, rather than merely pipeline a piece of existing information to the customer.

But what about other call center systems, like ACDs? If a caller has a choice of how to get into that customer contact zone, where will the switch fit in?

The answer has turned out to be twofold.

First, there is the omnipresent e-mail. (Itself a revolution in communications technology; if it weren't for the fact that the Web has pictures and sounds, we would be marveling at the amazing transformation e-mail has wrought in our society.) In call centers just the last two years has brought something that's sometimes called the "Internet ACD" or "E-mail ACD." Bad terms, but they do describe fairly well what's going on.

What they do is perform the same types of targeted routing that the telephony switch does for calls. They take large volumes of e-mail as it comes in, parses out some form of meaning (who is the sender, what is the subject, etc.), determines whether it can be answered with an automated response, and if not, sends it along to someone who can handle it in an effective way. Routing tables show who can handle what, and how often.

These systems also track and audit the response to those e-mails. So you can set an organizational parameter, for example, that all e-mails have to be "handled" within a certain timeframe.

The other major thrust has been a tentative exploration of one of the Web's major attractions, which is live chat. The idea being that a customer visiting your Web page would click on a button to initiate a text chat window, having a semi-real-time conversation with a rep back at the call center through a text typing session.

There are several advantages to this. One is that a single rep can handle multiple chat sessions at once, because many of the responses can be automated. (One vendor showed me a system where a rep could handle six at once -- a recipe for burnout and turnover if ever there was one.) The other advantage is that it is relatively simple technology connecting it to a rep's desk, in contrast to the much more complicated Web callback systems, that attempt to connect a Web surfer to a call center through an actual telephony connection.

As you would expect, vendors are closely watching Internet technologies, looking for ways to integrate their switches with Web-enabled applications. As one manufacturer suggested, it is when the consumer can (and does) cherrypick from a combination of entry points -- fax, e-mail, Web or voice call -- and expects to switch from one mode to another during a single "interaction" that advanced ACDs will need to be tightly integrated with the Internet.

What will happen is that human interaction will be reserved for where the agents can add the most value.


2000 Keith Dawson. All rights reserved. Keith Dawson is the editor of the Call Center News Service, a newsletter and website devoted to exploring the issues involved in customer contact. It is at



Back to Top of the Page, iCC Overview Page, iCall Centre or Home Page.