Forbes recently published an interview with Greg Gianforte, CEO of RightNow Technologies (Tools That Transform Customer Service, Jan 22). There are two interesting things discussed in the article. One is just what Gianforte’s company does and its implications for managing call centers and other customer service organizations. The second is some assertions he has about how large call centers are managing their workforce. Let’s start with what RightNow does. Here is part of Gianforte’s pitch:
We’re finding that consumers want choice in how they contact a company. All the avenues have to be open, whether it’s chat or e-mail or phone. In fact, consumers are differentiating less. They might start an interaction via e-mail and then escalate to a phone. So companies need systems in place.
So one way of understanding what they are trying to do is that they trying to provide an integrated platform for remotely dealing with customers. Whether customer call in or chat, the agent on the other end will have the resources necessary to answer questions or complete transactions. On top of that, they are monitoring social network sites. Thus, it is not just that a firm can respond when someone proactively calls, the firm can be proactive on its end to reach out when a customer expresses frustration or complains.
This cloud monitor capability that we’ve just introduced allows companies to actually put a listening ear out in the primary networks and connect it with the workflow back in their business, even to the point of measuring how angry someone who’s sending a tweet might be. …
So what companies really need is an early warning system to know what’s going on out there in these social networks and then a positive way to at least tap their customers on the shoulder and say, “Can I help?”
As I have noted before, this is one the interesting twists in service management over the last few years. The standard prescription is that companies need “listening posts” to gather information about customer experiences — particularly negative ones. However, with enough tweeters out there, listening can be replaced with reading and it becomes a question of tracking the flow. This then is the other cool thing RightNow sells:
We support 33 different languages today in our base system, and have a natural language processing capability that can read text and determine the emotional level of that text. … We detect emotional content about 75% of the time in terms of giving a valid positive or a valid negative indication. That means in three-quarters of the cases we can actually put this issue on the desktop of someone in the organization who actually can do something about it. In the prior situation we didn’t have that at all.
That is, we can automate at least the initial pass of what tweets or posts need to be followed up. That seems immensely helpful. A few years ago, I and several other faculty were asked to meet with a senior Korean telecom executive who was visiting the Kellogg School. He posed a question to us about how he could use contacts with his call centers to determine when something was going wrong. He gave the example of an internet virus that had hit his firm’s customers as well as many other users worldwide. It ultimately reached the point that it was on the evening news and everyone was talking about it. However, when they went back to look at the data, they realized that they had seen a steady increase in customer service calls days before this became an international crisis. A system that automatically tracks comments, posts, and call center interactions seems what the Korean executive needs.
Final point on that Gianforte makes is the potential to integrate customer service and sales:
[W]e’re seeing a collapsing of these two organizations. Most of the executives we work with that are running responsibility for customer care are being asked to generate revenue now. … [I]f I sell something to you based on a real need that you have, that is the highest form of service. We’re not talking about schlocky up-sell or cross-sell. We’re talking about understanding base needs and fulfilling them.
So I am a little more dubious on this point. From a business point of view, I get it. From a customer point of view, I am inherently dubious of firms trying to sell to me at every instant. I hate, for example, being put through a customer retention process when I am trying to cancel a service. I am not convinced that linking sales in service is particularly easy; if agents know that management prioritizes revenue, the service process can easily suffer.
The second point that Gianforte discusses is how call centers are making use of agents working from home:
We had a summit with a dozen executives from the retail industry at our corporate headquarters late last year. The topic we spent more time on than any other was work-at-home agents. Some of these firms had used an outsourcer who had a strong work-at-home agent program. A number were kind of rolling their own work-at-home agent programs. But they found that when employees are given the ability to work from their home environment, attrition dropped. The employees were more loyal. The customer satisfaction level from the end-customer was higher because they had more knowledgeable agents.
The argument is that companies offer working at home as a perk for agents who have proven their competency working in the regular call center. On the downside, the employee has to provide his or her own computer and internet connection but on the upside they gain flexible hours, less commuting, and a more relaxed dress code. This all sounds like a win-win proposition. But it also does not resonate with a recent BusinessWeek cover story (The Disposable Worker, Jan 7). That article opened with a profile of a work-at-home call center agent working for LiveOps, a call center outsourcing firm. Here is the kicker relative to the rosie picture Gianforte paints. LiveOps pays by the minute that the agent is engaged. They may be willing work, sitting at their terminal, but if there are no calls, no one gets paid. For the firm, this is an attractive. Labor, like computing power, can be a true variable cost — staffing is scaled up and down with demand. From the worker’s perspective, it means a lot more risk than traditional employment. To some extent, someone in the transaction has to bear that risk, but it is doubtful that it is socially efficient to leave it with workers.