+

QUEUEING SYSTEMS
A CASE STUDY ON QUEUEING IN CALL CENTERS

Queueing Theory- Definition

Queueing theory is the mathematical study of waiting lines, or queues . The theory enables mathematical analysis of several related processes, including arriving at the (back of the) queue, waiting in the queue (essentially a storage process), and being served at the front of the queue. The theory permits the derivation and calculation of several performance measures including the average waiting time in the queue or the system, the expected number waiting or receiving service, and the probability of encountering the system in certain states, such as empty, full, having an available server or having to wait a certain time to be served.

Communication Delays
Processing Delay This is the delay between the time of receipt of a packet for transmission to the point of putting it into the transmission queue. On the receive end, it is the delay between the time of reception of a packet in the receive queue to the point of actual processing of the message. This delay depends on the CPU speed and CPU load in the system.

Queuing Delay

This is the delay between the point of entry of a packet in the transmit queue to the actual point of transmission of the message. This delay depends on the load on the communication link.

Transmission Delay

This is the delay between the transmission of first bit of the packet to the transmission of the last bit.

This delay depends on the speed of the communication link.

+ Propagation Delay

This is the delay between the point of transmission of the last bit of the packet to the point of reception of last bit of the packet at the other end.
This delay depends on the physical characteristics of the communication link.

Retransmission Delay

This is the delay that results when a packet is lost and has to be retransmitted. This delay depends on the error rate on the link and the protocol used for retransmissions.

Characteristics of a Queueing System

Arrival Process

The probability density distribution that determines the customer arrivals in the system.
In a messaging system, this refers to the message arrival probability distribution.

Service Process

The probability density distribution that determines the customer service times in the system. In a messaging system, this refers to the message transmission time distribution. Since message transmission is directly proportional to the length of the message, this parameter indirectly refers to the message length distribution.

+ Number of Servers

Number of servers available to service the customers. In a messaging system, this refers to the number of links between the source and destination nodes.

M/M/1 Results
First we define p, the traffic intensity (sometimes called occupancy). It is defined as the average arrival rate (lambda) divided by the average service rate (mu). For a stable system the average service rate should always be higher than the average arrival rate. (Otherwise the queues would rapidly race towards infinity). Thus p should always be less than one. Also note that we are talking about average rates here, instantaneous arrival rate may exceed the service rate. Over a longer time period, the service rate should always exceed arrival rate.

Mean number of customers in the system (N) can be found using the above e equation:

we can see from the equation that as p approaches 1 number of customers would become very large. This can be easily justified intuitively. p will approach 1 when the average arrival rate starts approaching the average service rate. In this situation, the server would always be busy hence leading to a queue build up (large N). Lastly we obtain the total waiting time (including the service time):

Again we see that as mean arrival rate (lambda) approaches mean service rate (mu), the waiting time becomes very large.

What is a Call Center

A call center constitutes a set of resources (typically personnel, computers and telecommunication equipment), which enable the delivery of services via the telephone. The working environment of a large call center (Figure 1) could be envisioned as an endless room with numerous open-space cubicles, in which people with earphones sit in front of computer terminals, providing tele-services to unseen customers. Most call centers also support Interactive Voice Response (IVR) units, also called Voice Response Units (VRUs), which are the industrial versions of answering machines, including the possibilities of interactions. But more generally, a current trend is the extension of the call center into a contact center. The latter is a call center in which the traditional telephone service is enhanced by some additional multi-media customercontact channels, commonly VRU, email, fax, Internet or chat(in that order of prevalence).

Call Centers as Queueing Systems

Call centers can be viewed, naturally and usefully, as queueing systems. In a queueing model of a call center, the customers are callers, servers (resources) are telephone agents (operators) or communication equipment, and telequeues consist of callers that await service by a system resource. The simplest and most-widely used such model is the M/M/s queue, also known in call center circles as Erlang C. For most applications, however, Erlang C is an oversimplication: for example, it assumes out busy signals, customers impatience and services spanned over multiple visits.

Comparison of Queueing Options

Comparing traditional and virtual queuing timelines shows the difference in the customer experience. In this first example, the customer waits in a traditional queue for 12 minutes. When he's finally connected with an agent, he talks for 3 minutes - but some of that time is spent complaining about his time spent in the queue. Note that many customers in this situation would abandon the queue before reaching an agent, and retry the call later, resulting in additional telecom costs for the contact center and skewed call center metrics.

In the second example, the customer is treated by a virtual queuing system. She listens to a greeting that informs her of her EWT and offers her the option of receiving a callback rather than waiting in a queue.

Example 1

She prefers to remain in the queue, so her call enters the queue and she is connected with an agent when her turn arrives. It's unlikely that she will waste time complaining because she was informed of her estimated wait and presented with options for managing her time. This is indicated in the example with Saved Talk Time. She may also be less likely to abandon the call because she was informed and made a conscious choice to remain in the queue.

The third example shows a customer who is treated by the virtual queuing system and chooses to receive a callback in the same amount of time as if he waits in queue. After entering his phone number and speaking his name, the customer hangs up the phone and a virtual placeholder reserves his spot in the queue.

This "virtual queue time" saves inbound telecommunications charges (because the customer is not on the line) and frees up the customer's valuable time. When the placeholder is near the front of the queue, the system calls the customer back, greets him, and puts him at the front of the queue, where he is next to be answered by an agent. Since the customer has had a positive experience, he may be less likely to complain about a long wait.

IMPACT
Virtual queuing impacts the call center metrics in many ways. Queue time is normally measured as Average Speed-toAnswer (ASA). When callers are offered the option to receive a First in First out callback, the callers acceptance rates are typically 45% to 55%. Therefore, about half of the calls that would normally queue for 5 to 10 minutes will now only accrue a speed-to-answer (ASA) of approximately 10 seconds. Likewise, these callbacks with a shorter ASA will score within the Service Level objective. Since callers cannot abandon while in a virtual queue, the overall number of abandoned calls will decrease. The impact on customer satisfaction is positive, but tends to be more difficult to measure objectively. Virtual queuing can result in better customer experiences and improved contact center operations.

Data analysis and forecasting

Any modeling study of call centers must necessarily start with a careful data analysis. For example, the simplest queueing model of a call center requires the estimation of calling rate and mean service (holding) times. Moreover, the performance of call centers in peak hours is extremely sensitive to changes in its underlying parameters. It follows that an extremely accurate estimation/forecasting of parameters is a prerequisite for a consistent service level and an ecient operation