Organizing and handling large volumes of contacts, or those from specialized or niche customers, is challenging. Do you centralize agents in one large call center or a mix of small call centers? Should some or all agents telework?

Teleworking is a separate decision because it involves being comfortable with, and selecting, agents to work off-site. Teleworking also entails separate technology. (See also the April and July 2002 issues.)

Deciding between one, a few, or many call centers is analogous to a railroad dispatcher who must send loads down the track. Do you couple the loads on one long train or in short, separate trains for each class of products?

While there are distinct advantages and disadvantages to multi-site operations, the technology to enable them is improving and becoming more cost-effective. This includes functional CTI middleware, ACD management software and packetized voice, i.e., voice-over-IP (VoIP).

John Cosgrove, vice president of solutions strategy for systems integrator Expanets (Denver, CO), has seen companies swing between having one large and multiple small call centers. Prompting many of these moves were company and management preferences or changing labor market conditions. Cosgrove expects these trend swings to continue.

"Call center managers' choices of technology and methods to enable multiple locations and teleworking are increasing," he says. "That gives them flexibility in their site strategies. But there is no right or wrong choice. It is up to each company to determine what is best for them."

MAKING THE MOST OF LOCATIONS

Many companies that already have multiple call centers are integrating them to act as virtual call centers, with excellent results.

Too often, companies had set up separate centers to handle specific functions, such as customer service, collections, sales, or different product lines. Sometimes there are centers for each region of the country. Such centers are often outgrowths of local or regional offices.

Consequently, companies risk wasting resources, annoying customers and losing revenue when customers go elsewhere. They may have customers on hold for 20 minutes in one center and idle agents checking the stock market in the other centers.

Todd Carter, solutions engineer with Norstan (Minnetonka, MN), says that if you have separate centers for sales and customer service, you can boost revenue by integrating the centers. He advises that call centers establish sales as a priority by training customer service reps to handle sales.

"Then when the sales center reaches the queue limit, you can dynamically expand your call routing to offer those sales calls to the customer service agents regardless of their location," he says.

Gerry Dupuis, director of consulting practice for Softel (Toronto, Canada), worked with an unnamed high-end financial services firm that had six geographically dispersed call centers. All had the same functions: customer service, qualifying inbound leads and selling policies to qualified customers.

The call centers had the same hardware and software, including switchgear. There was untapped network routing software on the switches. Yet the switches were not connected. When the client linked call centers, its costs dropped. The firm needed fewer agents and IS/IT staff; it found these employees work elsewhere in their organization.

"We had to educate [the firm] on the benefits of creating a virtual call center," recounts Dupuis. "When they implemented our advice, they had fewer abandons. Their call completion rate rose to 98% from 94%. Revenue climbed by 14% because customers weren't hanging up and going elsewhere."

Creating a virtual call center, however, can be a big challenge for multiple call centers with different switching, routing and IVR platforms. Often, the platforms do not communicate with each other because they are built to different standards.

Many companies face this problem when they acquire another firm or decide they need call centers to support the acquired firm's products. Or they need to keep the people, the space and the technology, while reorienting the assets to sell or support products and services.

If the call center technologies don't talk to each other, you can't have necessary features such as skills-based routing, explains Bernard Drost, vice president of technology for Akibia (Westborough, MA).

Consultants say that opening a new center with standards-based switches, software and routing technologies, is almost always cheaper and less burdensome than integrating acquired technologies with yours.

If the ACD directs the call to somebody who can't take the call within the required time, or if the agent answers but cannot resolve the matter, the agent transfers the call or returns it to the queue.

The queue might then transfer the call to the second call center's main number. That forces the customer to navigate the IVR menu again because only the call could be transferred, not the data associated with the call.

"That gets annoying to customers very quickly," says Drost. "You get a lot of abandoned calls after the second or third transfer."

TECHNOLOGY OPTIONS

There will be better technology integration in the future, say consultants.

Most companies, for example, will have a standards-based XML (extensible markup language) interface to port data. A few years down the road, voice over IP (VoIP) may become sufficiently reliable, with acceptable voice quality to replace circuit-switched voice. (More about that later.)

Until all systems can talk to each other, there are no right answers to provide effective skills-based routing across call centers and platforms. There are two main routing options: carrier-based and customer premises-based routing, which is similar to leasing versus buying a car. Each has pros and cons.

With carrier-based routing, you avoid the high up-front capital costs but you pay more in operating costs, especially line charges. You also do not control the technology, as you do with premises-based routing.

If you prefer premises-based routing, you have to determine what to do with older, but costly-to-replace legacy switches and IVR systems with several years' life left.

According to Norstan's Carter, new PBX and ACD platforms can cost from $300,000 to $2 million each, while a 24-port IVR can run about $100,000. Yet switches installed as long ago as the mid-1980s can still provide good solutions under the right circumstances.

That could mean deploying CTI middleware, like Aspect's (San Jose, CA) Enterprise Contact Server; Cisco's (San Jose, CA) Intelligent Contact Management (ICM); and Genesys' (San Francisco, CA) Universal Queue2.

These middleware products integrate different PBXs and ACDs to provide dynamic contact routing across locations. The middleware "normalizes" the proprietary events and services from each platform to provide a virtual queue, enterprise reporting, and the ability to pass associated call data among locations for screen pops.

Organizing and handling large volumes of contacts, or those from specialized or niche customers, is challenging. Do you centralize agents in one large call center or a mix of small call centers? Should some or all agents telework?

Teleworking is a separate decision because it involves being comfortable with, and selecting, agents to work off-site. Teleworking also entails separate technology. (See also the April and July 2002 issues.)

Deciding between one, a few, or many call centers is analogous to a railroad dispatcher who must send loads down the track. Do you couple the loads on one long train or in short, separate trains for each class of products?

While there are distinct advantages and disadvantages to multi-site operations, the technology to enable them is improving and becoming more cost-effective. This includes functional CTI middleware, ACD management software and packetized voice, i.e., voice-over-IP (VoIP).

John Cosgrove, vice president of solutions strategy for systems integrator Expanets (Denver, CO), has seen companies swing between having one large and multiple small call centers. Prompting many of these moves were company and management preferences or changing labor market conditions. Cosgrove expects these trend swings to continue.

"Call center managers' choices of technology and methods to enable multiple locations and teleworking are increasing," he says. "That gives them flexibility in their site strategies. But there is no right or wrong choice. It is up to each company to determine what is best for them."

MAKING THE MOST OF LOCATIONS

Many companies that already have multiple call centers are integrating them to act as virtual call centers, with excellent results.

Too often, companies had set up separate centers to handle specific functions, such as customer service, collections, sales, or different product lines. Sometimes there are centers for each region of the country. Such centers are often outgrowths of local or regional offices.

Consequently, companies risk wasting resources, annoying customers and losing revenue when customers go elsewhere. They may have customers on hold for 20 minutes in one center and idle agents checking the stock market in the other centers.

Todd Carter, solutions engineer with Norstan (Minnetonka, MN), says that if you have separate centers for sales and customer service, you can boost revenue by integrating the centers. He advises that call centers establish sales as a priority by training customer service reps to handle sales.

"Then when the sales center reaches the queue limit, you can dynamically expand your call routing to offer those sales calls to the customer service agents regardless of their location," he says.

Gerry Dupuis, director of consulting practice for Softel (Toronto, Canada), worked with an unnamed high-end financial services firm that had six geographically dispersed call centers. All had the same functions: customer service, qualifying inbound leads and selling policies to qualified customers.

The call centers had the same hardware and software, including switchgear. There was untapped network routing software on the switches. Yet the switches were not connected. When the client linked call centers, its costs dropped. The firm needed fewer agents and IS/IT staff; it found these employees work elsewhere in their organization.

"We had to educate the firm on the benefits of creating a virtual call center," recounts Dupuis. "When they implemented our advice, they had fewer abandons. Their call completion rate rose to 98% from 94%. Revenue climbed by 14% because customers weren't hanging up and going elsewhere."

Creating a virtual call center, however, can be a big challenge for multiple call centers with different switching, routing and IVR platforms. Often, the platforms do not communicate with each other because they are built to different standards.

Many companies face this problem when they acquire another firm or decide they need call centers to support the acquired firm's products. Or they need to keep the people, the space and the technology, while reorienting the assets to sell or support products and services.

If the call center technologies don't talk to each other, you can't have necessary features such as skills-based routing, explains Bernard Drost, vice president of technology for Akibia (Westborough, MA).

Consultants say that opening a new center with standards-based switches, software and routing technologies, is almost always cheaper and less burdensome than integrating acquired technologies with yours.

If the ACD directs the call to somebody who can't take the call within the required time, or if the agent answers but cannot resolve the matter, the agent transfers the call or returns it to the queue.

The queue might then transfer the call to the second call center's main number. That forces the customer to navigate the IVR menu again because only the call could be transferred, not the data associated with the call.

"That gets annoying to customers very quickly," says Drost. "You get a lot of abandoned calls after the second or third transfer."

TECHNOLOGY OPTIONS

There will be better technology integration in the future, say consultants.

Most companies, for example, will have a standards-based XML (extensible markup language) interface to port data. A few years down the road, voice over IP (VoIP) may become sufficiently reliable, with acceptable voice quality to replace circuit-switched voice. (More about that later.)

Until all systems can talk to each other, there are no right answers to provide effective skills-based routing across call centers and platforms. There are two main routing options: carrier-based and customer premises-based routing, which is similar to leasing versus buying a car. Each has pros and cons.

With carrier-based routing, you avoid the high up-front capital costs but you pay more in operating costs, especially line charges. You also do not control the technology, as you do with premises-based routing.

If you prefer premises-based routing, you have to determine what to do with older, but costly-to-replace legacy switches and IVR systems with several years' life left.

According to Norstan's Carter, new PBX and ACD platforms can cost from $300,000 to $2 million each, while a 24-port IVR can run about $100,000. Yet switches installed as long ago as the mid-1980s can still provide good solutions under the right circumstances.

That could mean deploying CTI middleware, like Aspect's (San Jose, CA) Enterprise Contact Server; Cisco's (San Jose, CA) Intelligent Contact Management (ICM); and Genesys' (San Francisco, CA) Universal Queue2.

These middleware products integrate different PBXs and ACDs to provide dynamic contact routing across locations. The middleware "normalizes" the proprietary events and services from each platform to provide a virtual queue, enterprise reporting, and the ability to pass associated call data among locations for screen pops.