The Fallacy of Email

Lets get into a discussion of some of the interesting new tools available, including CRM and the new theories of multichannel access for customer contact. But it seems appropriate to pause here for a moment and discuss, since I’m on the subject of the changes wrought by the Internet, the strange nature of email in call centers.

With an astonishing speed, the way we discuss the relationship between the call center and the Internet has changed. It’s not going to be about “click-to-call” buttons on Web pages. And it’s not going to be Internet telephony or IP networks handling calls.

Those things will happen, but not so soon as the vendors of those technologies would have you believe. No, what’s really happening is email, and its cousin the Instant Message, and its other cousin the live-text chat.

This makes a lot of sense, when you think about it. The call center was built on the premise that customers initiate contact, and that whatever they need can either be handled in real time by an agent, or handed off to an automated system that can extract the information needed and deliver it to the customer.

Email is just a phone call that happens in slow motion. It’s trackable, contains its own audit trail, and can be parsed by computers looking for keywords. And most important, it’s ubiquitous. Everyone has it. Not as many people as have telephones, certainly, but an explosive number.

If you run a call center, and you believe that the Internet is coming your way through a call-me button, then you have to make certain assumptions. You have to plan for how the webpage is going to connect to the call routing engine, for one. That’s a complicated piece of business; you have to make certain assumptions about queuing that may or may not be accurate because people surfing the Web are not the same kind of randomness as people picking up the phone.

And you’re going to have to deal with the complex CTI interactions between the forms that they fill out online and the data residing back at the host. Because it’s silly to collect information and then not coordinate it back at the agent’s desk. And even sillier not to take advantage of the connection to do screen sharing between the agent and the customer.

But once you’ve begun to travel down this road, you’ve introduced a lot of new costs, and managerial headaches. So many, in fact, that I’ve been able to find very few call centers that decided it was a road worth travelling — at this stage.

Many people believe that this kind of Web/call center integration is only a few years away, and I can see that logic. But it’s not a here-today kind of technology, for most people.

But while this debate was going on, email kept coming. It’s flooding into companies in much the same way telephone calls did in the early days of call centers. It’s often undirected, (what intelligent Webmaster, after all, is going to answer 10,000 emails about why a product doesn’t work?) and uncategorized. It’s a mess. Just as the days when call centers were about triaging the incoming calls and finding someone — anyone — to handle the traffic.

The response to this phenomenon has been very interesting. First, a set of tools was developed to tag and track incoming email for companies. The first thing you have to do, obviously, is identify what each one is about and try to match it up to known data about customers. After that you apply some rule to decide what the appropriate response should be.

Lots of little companies entered the fray, many (if not most) from the Internet side, not the call center or telecom industry.

Then, as things developed, these small companies aligned themselves with the voice switching firms, and others that provided call center infrastructure. In a strange paradox, the call center has grown so important to corporate well-being, and has taken on the role of “customer contact center,” that it is a natural place to push all these unsorted emails, whether or not the call center is set up to handle them.

The fallacy of email, however, is that handling an email is not the same as handling a call. It’s similar, but that’s about all you can say. Expecting a call center to do double duty as an email processing center is as hazardous as expecting a single agent to handle both inbound and outbound calls. Most centers don’t do that, because they know that the skills required are different.

Some centers begin this process by assigning some agents to email processing part time. They can route email for response to a particular rep or group during preset low points in the call traffic, for example.

But this makes it hard to assess the agent’s overall productivity, because he’s doing different kinds of work that require different measurements of success.

And it prevents you from setting the same kind of consistent service level targets for email that you would for calls. You can’t very well expect to respond to all emails within 24 hours (a common target) if you don’t know at any given moment how many people will be available to handle emails, or how many you’re going to receive.

If you truly believe that an email from a customer is as important — as valuable — as a call, then you have to treat it equally. And you can’t rely on handling it during the call center’s off-time.

What’s true for email is also true for some of the other email-like interactions. Text chat is where the agent and the customer are connected through a text window and type back and forth at each other. Some products that facilitate this are proprietary, and others use one of the open Instant Messaging tools like those from Yahoo or AOL.

True story: last year I was watching a demo of one of the new text-chat systems. It was a fine piece of technology. The system worked as advertised, it seemed easy to use, and looked good from a customer point of view. It contained software to parse what the customer was saying and present the agent with a choice of responses to save time.

The exec at the company showing it off had six windows up on the screen, meaning that the agent was engaged in six simultaneous customer conversations. I thought this was just a bit of showing off, until he said that they were selling that point to potential customers, pointing out that they could have each agent do six times as much work.

Think about this: in a world where 35% turnover is not unusual, does anyone expect a call center rep to carry on six customer conversations at once? When not on the phone? You’d be crazy to pitch that to reps.

I’d be surprised if they didn’t balk at two.

What this argues for is a separation of function. Email, chat and other textual communication should be in the hands of people specially trained for the job. There are nuances to it. There are special skills that should be identified as key to this kind of work. Training should be dedicated to it.

Maybe there are people who can do both, but I don’t think you’ll find enough of them to fill all the call centers.

I think people who choose to go into the customer contact field — which is a better way to think of call center repping — should be tested for the kinds of interactions they excel at, and should be put on either email or chat or phones, depending on what suits them. Maybe some of each, to create a backup in case of emergency.

The fallacy of email is that although it has a lot of the same qualities as a call, it’s a different kind of interaction. People have different expectations for what’s going to be in the response. And there is no consensus on when they should expect that response.

The explosion of tools, many of them wonderful, from companies like Mustang, FaceTime, Kana, eGain, Brightware, WebLine and many others speak to the health of call centers. It means they are grappling for solutions to a growing problem. You could see it as a problem of riches: many customers using more channels means more information gathered about their likes and dislikes, more opportunities to make sales.

The problem is not with the tools. It’s with the assumption that the person handling the calls is the right person for other kinds of connections, as well. Let’s experiment with dedicated email groups. Don’t build “email centers” just yet, but separate the groups who are best able to handle emails. And don’t jump to morph your call center into a “customer contact center” unless you’re really willing to deal with the management responsibilities that come along with it.

The Connections in Practice | The Web & Call Centers

A book is not the perfect place to talk about such a dynamic and fast-changing subject as the relationship between call centers and the Internet. In any case, one way to see the impact of the changes is through the flurry of products that have come out recently, and the changing relationships between companies as they cross from one product category to another.

For example, a Web-browser-based publishing tool makes collecting and sharing customer data within the call center, and throughout the enterprise, as easy as accessing a webpage. Envision Telephony’s SoundByte Enterprise Web Desktop lets centers publish customer contact information in a browser page format or “Web desktop.”

The new format simplifies the transfer of customer information from the call center to the enterprise. Departments outside of the call center can use Web Desktop to get customized access to data necessary to make more strategic, timely business decisions. The kinds of data at work here are primarily digital call recordings and call center performance reports.

Executive management, marketing and product development, for instance, can track customer response to promotions, monitor service quality and query customers through this system. Web Desktop is part of Envision’s SoundByte Enterprise call center suite, an automated call monitoring system that collects and publishes information about customer contact with a company’s agents.

Core elements of Web Desktop include call evaluations and graphical comparisons of individual versus group performance. Supervisors can also add training tools, demonstrate productivity reports, publish department specific issues, and highlight morale topics. For example, supervisors can feature “agent of the month”, incentive programs, and other events.

VocalTec’s Surf&Call Center lets online shoppers make toll free calls directly from a website into any telephone or formal call center. A Pro version goes on to let online shoppers speak over the Internet to call center agents as well as jointly surf webpages and fill out online forms in a secure environment.

Surf&Call Center uses the same standards-based IP telephony network infrastructure that carriers can use to target a variety of audiences with services such as phone-to-phone, fax-to-fax, PC-to-phone, and Internet Phone Call Waiting.

The SDK also allows system integrators to customize webpages and agent desktops to support Surf&Call Center’s voice and joint data collaboration capabilities.

As previously stated, the market for Web-enabled call center solutions is expected to grow at a compound annual growth rate of more than 110% through 2004, according to Frost & Sullivan stats. Interestingly, this product is aimed squarely at the carrier market and the call centers those carriers spawn.

Primus’s WebPack is a set of tools for creating knowledge bases and publishing them on the Internet, complete with natural language search capabilities. The gimmick is that you’re supposed to be able to complete said knowledge base within a single workweek.

The advantage to putting the knowledge out in front of the call center is simple: customers will use it to try to solve their own problems before coming to an agent for help. This cuts down on call volume.

(Though I suspect that the duration of the remaining calls would be somewhat higher, because those calls are likely to be the thorny ones that can’t be worked out using an automated system.)

One early user of the WebPack system reported that for 65% of its customers, the self-service option directly solved their problem.

Another Internet front-end product takes a somewhat different approach. FaceTime’s Message Exchange service tries to meld all the newer modes of Internet communication — instant messaging, for example — into a logical package that can be parsed by a call center.

Consumers engaged in online shopping at a FaceTime-enabled website instantly receive live and/or email-based sales and customer service assistance by connecting to the company’s call center. The FME service suite routes, queues, and reports on inbound email and instant messaging interactions, helping sites prioritize customer requests and manage service levels.

Most interesting is Instant Messaging Management, enabling ecommerce sites to manage real-time, live text-based communication between a CSR and a customer. FaceTime accepts, routes, and queues these requests to the appropriate CSR, and provides supervisory functions and management reporting. (A lot of the actual interaction can be automated.)

The other main component, Email Management, manages inbound or forms-based requests from consumers.

The FaceTime Message Exchange features a real-time monitor that shows supervisor and managers current queue and agent status. It can also monitor customer specified service level targets and alert the supervisor when these targets have been exceeded.

FaceTime is an application service provider (ASP) for Internet customer service. In other words, FaceTime is an online service, not a product. That’s good, because it means you can be deployed in days, scale up or down and avoid implementation hassles with the IT department.