May 2003

Meade: The ideal candidate profile for an inbound agent is as follows: the candidate must be at least 18 years of age; must pass a criminal background check and UA; can type 20 words per minute; has the ability to handle repetitive tasks; possesses a clear speaking voice; has solid listening skills; has high patience and the ability to keep calm during stressful situations; has the ability to treat like circumstances with like consequences; has prior customer service work/likes interaction over the phone; has the ability to find the root cause of the problem; has a solid attendance/punctuality record; can sit for long periods of time; and has the ability to follow directions and continue without deviation until told differently.

Kuwahara: Today's inbound agents must operate in complex environments. They not only need to efficiently handle customers, but they also interact with multiple computer systems and resources. Beyond the obvious requirement for strong verbal communication skills, inbound agents must be proficient at:

' Developing customer rapport, which is essential to building a positive relationship with the customer. During interviews we look for individuals who are able to hold interest in a conversation, exhibit good listening skills, repeat what is said and verify understanding. We also build upon these basic skills in our training programs. 

' Discovering customer needs, which is at the heart of being able to identify and present options to customers that they are not aware of. This is often a driver of improved customer satisfaction.

' Problem solving and responding, which is vital to first-call resolution, a satisfaction factor important to customers. These skills are the building blocks of customer service.

' Learning and applying procedures, so that the agent is able to adhere to business and risk-management policies.

' Multi-tasking, where agents need to be able to talk while accurately gathering information.

Additionally, virtually all inbound agents must have a sales orientation. Even pure customer service programs are in the business of renewing relationships with their customers at every interaction. 

CIS: How do you and a client determine which services should be promoted through upselling/cross-selling techniques?

Kuwahara: Convergys clients generally have an existing portfolio of products and/or services that they are already trying to upsell or cross-sell to their customers. Their challenge, however, is knowing which product, or bundle of products and services, to offer to which customers, and when. The other challenge is being able to optimize conversion rate, average handle time (which influences cost/conversion), product/service penetration and retention, and customer satisfaction. All of these variables need to be considered relative to each other when considering the appropriate upsell or cross-sell offer.

Convergys Knowledge Management Services uses an approach called affinity modeling. Typically, we first gather data from a variety of sources, such as: past product/service conversions by customer, customer geo-demographics, customer satisfaction data, current customer product profile and usage history. 

These data are then analyzed to determine which variables are the most important in terms of predicting the most likely product the customer will buy next, and which variables are most likely to be correlated. The next step is to build a product segmentation or affinity model, which combines a host of different statistical techniques. These models are then scored and tested against a segment of the customer base for validation purposes, before they are actually tested in a live-agent environment.

Meade: StarTek's methodology for determining which products or services to offer customers when an inbound call is placed follows a four-step process as follows:

1. Multi-Channel Customer Dialog
' Immediately get customers who have called in or established a chat session in a suitable frame of mind, via scripted dialog.

2. Segmentation Of Customers
' Quickly determine the status of this particular customer (i.e. high-/low-value customer, determine current product/service mix, identify prior buying patterns. This is done through a series of database queries.

3. Offer Presentment
' Probe, assess and persuade through tailored, scripted questions to reach the right solution (product or service).

' Deliver a sales offer that has a very high probability of a favorable outcome.

4. Negotiate And Close Deal
' Possible offer of discount, additional product or service at no charge, bundle solutions

' Enter order into order management application.

CIS: What technologies assist the agent with understanding the customer and determining the right products and approach to upsell and cross-sell?

Meade: StarTek has established a number of business partners who offer the best-of- breed solutions required based on our customers' requirements.

Kuwahara: The suite of technologies that Convergys uses to accomplish this can range in complexity. The simplest application that we can deploy for a client is a Convergys-proprietary ANI screen-pop application. This application enables us to match an inbound caller, based on their telephone number, with a third-party name and address database. Appended to that database are the modeled scores for the products and services that were created as a result of the analytics I described earlier, attached to each record. When the caller calls in, the application will decide based on score values to invoke a customized script for that particular caller. The script is based on the customer's current product profile and products or services they are deemed most likely to buy. After the main inquiry is resolved, the agent will then offer the customer a sequence of offers tailored to that caller's profile.

More complex technologies entail the use of intelligent routing and queuing equipment, such as premise-based CTI. In addition, the database that sits behind the desktop application (such as Siebel) can contain more detailed customer history and preference data, if we are marketing to existing customers. A real-time decision engine, such as HNC or SAS, could be used to introduce offers based on more complex decision rules and customer profile characteristics maintained within the operational database. The bottom line is the agent actually does not need to be very educated relative to the use of any of these technologies. Most of the technologies are programmed and operated behind the scenes.

CIS: Are there any particular challenges for your offshore centers?

Kuwahara: There are two unique challenges associated with the offshore centers. These challenges center on the communication skills of the agents and the cultural differences. The communications issues are best described by first understanding that many American concepts, words and colloquialisms do not exist in many offshore countries. Americans have developed a language of their own. Translating this language into comparable language for offshore agents often begins with setting context. For example, to an agent in India, a hardware store is a small establishment with a limited number of items, while in the U.S. we refer to a large home improvement warehouse as a hardware store. 'Living in the boonies' is not a phrase that translates readily so that many offshore people can understand it.

The cultural differences manifest themselves in many ways. Americans are very time-sensitive. That is not a behavior that is shared by many of the offshore cultures. Conflict adversity is another cultural difference. Many offshore cultures choose not to deal with conflict. These are just a couple of examples. There are many that can go undetected until an issue arises.

[ Return To The May 2003 Table Of Contents ]