Taske Call Center Management Tools Traffic Analyzer Instructions Manual

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-11-Using the Traffic Analyzer program
Creating a Traffic Analysis
The Calculation Wizard prompts for all of the information needed to create a
graph of the traffic analysis calculation.
1. Click the Calculation Wizard button.
2. The Statistic Selection dialog box appears. Choose the class of
algorithm for the traffic calculation.  The choices are:
Erlang C ModelErlang C modeling assumes calls placed by a source that are
blocked will never hang up, but will queue forever waiting for a
server to come free. This model follows the characteristics of an
incoming ACD caller to an ACD group or a caller into a Pilot Group.
Erlang B ModelErlang B modeling assumes that calls placed by a source that are
initially blocked will not reattempt the call and that the call will not
queue up waiting for a server to come free. When the call is initially
blocked, the caller hangs up and NEVER re-attempts to originate
again. This distribution typically models outgoing trunk groups.
Poisson ModelPoisson modeling assumes that calls placed by a source that are
initially blocked will hang up and reattempt the call at random
intervals until the call successfully connects OR until it has waited
the length of time the call would have lasted if not blocked. The
model assumes that calls will never queue.  This distribution is not
widely used in North America, but is provided for completeness.
The Erlang C formula is used to model ACD Queues and Agent Groups.
Erlang B and Poisson are formulae used to model Trunking and Blocking. 
						

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-12-Choose the class of algorithm (Erlang B, Erlang C, or Poisson) to use for
the traffic calculation, and click Next.
3. Depending on which traffic model was selected, one of three dialog
boxes open.  The dialog boxes are illustrated below.  Each dialog box
contains a list of formulae, which corresponds to the traffic model
selected.
Tip  To get detailed information on any dialog box in the Wizard, click its Helpbutton.For Erlang C, follow the directions below.  For Erlang B, skip to the next
section of this manual “Erlang B”, on page 18. For Poisson, skip to the
section of this manual titled “Poisson” found on page 21. 
						

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-13-Erlang C
Selecting Erlang C opens the dialog box illustrated below.
The Erlang C Selection allows the following calculations to be performed.
The explanation of each of the parameters shown above together with an
example indicates how these calculations are used to manage Call Center
traffic.
Probability of Blockage (ERLC_GOS formula)
This calculation relates the GOS to the number of Agents and the
call load (or the number of calls per hour and the average duration
of these).  The calculator accepts  the number of calls per hour, the
average call duration, and the number of available Agents as
parameters.  The GOS that results from the call load being
presented to a given number of Agents, may be determined.
For example, the Probability of Blockage calculation can be applied
to estimate the expected call load and to know how many Agents
are scheduled to work. Then, you can calculate the GOS that
Callers are expected to receive.  Although the GOS is generally
stated as the probability of Blockage, the GOS in an Erlang C
model actually represents the probability that a Caller will
experience a delay.  Thus, a GOS of P.10, or 10%, means that a
Caller has a 10% chance of being delayed. 
						

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-14-Probability of X or Less Agents Busy (ERLC_X_CHAN formula)
This calculation determines the probability that X Agents are
simultaneously busy when a given call load is presented to a limited
number of available Agents.  The calculator accepts the number of
calls per hour, the average call duration, and the number of
available Agents as parameters. It then calculates the probability of
a specified number of Agents being busy at any point in time.
This calculation is used to determine the probability of having no
Agents that are busy (i.e. All Agents idle).  Agent utilization, or the
proportion of time that an Agent is busy compared to the total work
time of the Agent, depends on the call load and the number of
Agents available.
For a certain call load and a limited number of Agents, the
calculation reveals that an Agent(s) would be idle for a significant
proportion of the time. Then, you might consider reducing the
number of Agents for the shift.  However, there is a close
relationship between Agent utilization and the TSF experienced by
Callers.  Particularly for a small Agent group, you may find that a
relatively low Agent utilization is the price that must be paid to
ensure a high level of service to Callers.
This calculation determines the probability that between X and Y
Agents are simultaneously busy when a given call load is presented
to a given number of available Agents.  The calculator accepts as
parameters, the number of calls per hour, the average call duration,
and the number of available Agents, and calculates the probability
that greater than X and fewer than Y Agents are busy at any point
in time.
This calculation determines the probability that between X and Y
Agents are simultaneously busy when a given call load is presented
to a given number of available Agents.  The calculator accepts as
parameters, the number of calls per hour, the average call duration,
and the number of available Agents, and calculates the probability
that greater than X and fewer than Y Agents are busy at any point
in time.
This calculation determines the probability that four or more Agents
in a group of eight are busy at any known point in time.  If you
supervise a Queue with eight Agents in the morning work shift and
four Agents in the afternoon work shift, you can calculate the affect
of reducing the number of Agents to four for the morning work shift. 
						

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-15-Probability of X and Y Agents Busy (ERLC_INTEG formula)
This calculation determines the probability that between X and Y
Agents will be simultaneously busy when a current call load is
presented to a limited number of available Agents.  The calculator
accepts the number of calls per hour, the average call duration, and
the number of available Agents as parameters. Then, it calculates
the probability that no fewer than X, and no more than Y Agents are
busy at any point in time.
You can use this calculation to determine the traffic on the least
occupied Agents in an Agent Group.  For example, you can
calculate the probability of Agents 8 to 10 (for a group of 10 Agents)
being busy when presented with 250 calls per hour, and having an
average duration of 70 seconds. This calculator calculates the
probability of Agents 8 to 10 being busy and determines whether
the individual Agent traffic justifies the cost.
Traffic that can be Carried (ERLC_TRAF formula)
This calculation determines the traffic or call load that can be
handled by a limited number of Agents within a given GOS.  The
calculator accepts the number of available Agents, the required
GOS, and the average call duration as parameters. Then, it
calculates the number of calls per hour that can be handled.
For example, you can determine the call load, with an average call
duration of 70 seconds, that can be handled by four Agents, while
maintaining the GOS within the company’s target of P.05.  This
calculation can demonstrate that the number of calls per hour that
can be handled is less than expected. To remedy this, you can
either increase the number of Agents to maintain the GOS within
target, or the GOS must be allowed to decline to enable the call
load to be handled by the available Agents.
Agents Required to Achieve a GOS (ERLC_TKS formula)
 This calculation determines the number of servers (or Agents in
this case) required to handle a certain call load within an
acceptable GOS.  The calculator accepts the number of calls per
hour, the average call duration of those calls, and the required GOS
as parameters. Then, it calculates the number of Agents required to
handle the calls.
For example, you can determine how many Agents are required to
handle 300 calls per hour at an average call duration of 80
seconds, while maintaining the GOS rating at P.20.  If this
calculation shows that the number of Agents required is more than 
						

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-16-the number you expect to have available, then the GOS will be
worse than the target GOS of P.20.
Probability of Delay (ERLC_DELAYT formula)
This calculation determines the probability of a Caller having to wait
a minimum of XX seconds to be answered, when a limited number
of Agents are presented with a certain call load.  The calculator
accepts the number of seconds for the threshold time, the number
of calls per hour, the average call duration of those calls, and the
number of available Agents, as parameters. Then, it calculates the
probability of Callers having to wait for at least the threshold time.
For example, you can determine the probability of  Callers that will
have to wait more than 20 seconds for an Agent to respond, when
30 calls per hour with an average single call duration of 70
seconds, are presented to 3 Agents.  If the Call Centers level of
service target is that 80% of calls be answered within 15 seconds,
then 20 seconds might be considered a very poor response.  The
Supervisor can use this probability to know the number of  Callers
that receive this poor level of service.
Average Queue Depth (ERLC_AVQ formula)
This calculation determines the average number of Callers waiting
in the Queue at any point in time, when there are a limited number
of Agents being presented with a certain call load.  The calculator
accepts the number of calls per hour, the average call duration of
those calls, and the number of available Agents as parameters.
Then, it calculates the average number of Callers who are delayed
and waiting in the ACD Queue at any point in time.
Note  Because this calculation produces a value for the averagenumber of Callers, the result may not be a whole number.  For
example, in the first second there may be one Caller waiting in
Queue.  For the next three seconds there may be no Callers
waiting; the average probability is then 1+3/4 = 1.25 Callers in
Queue.This calculation determines the average number of Callers waiting
in Queue, when 70 calls per hour with an average call duration of
70 seconds, are presented to 4 Agents.
Use this calculation to determine the amount of Recorded
Announcement Devices (RAD) that are required to provide Queue
delay announcements to Callers. For example, if the calculation
indicates that the average number of Callers in Queue is 0.03, thenenough RAD equipment to deliver an announcement to one Caller 
						

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-17-at a time is required.  If the calculation for a larger Queue were to
indicate that the average number of Callers in Queue is 3.8, then it
would probably be wise to provide enough RAD equipment to
deliver announcements to 5 Callers at a time; sometimes there are
more Callers in Queue than the average.Average Delay (ERLC_AVGDLY formula)
This calculation determines the average time that a Caller waits in
the ACD Queue before being answered by an Agent, when a
limited number of Agents is presented with a certain call load.  The
calculator accepts the number of calls per hour, the average call
duration of those calls, and the number of available Agents as
parameters. Then, it calculates the average delay experienced by
the Callers.
For example, you can determine the average waiting time for
Callers when 7 Agents are presented with 70 calls per hour with an
average call duration of 70 seconds.  In this case, the average wait
is insignificant. If different parameters are applied and the wait time
is considered to be too long, then additional Agents should be
made available, or the average call duration should be decreased.
Agents Required (ERLC_AGENTS formula)
This calculation determines the number of Agents required to
handle a given call load within a stated TSF.  The calculator
accepts the number of calls per hour, the average call duration of
the calls, and the required TSF percentage and time as
parameters. Then, it calculates the number of Agents required to
handle the call load within the TSF requirements.
For example, you can determine the number of Agents required to
handle 150 calls per hour, with an average duration of 80 seconds,
and within a TSF of 80% of calls answered or abandoned within 20
seconds.  This is a key calculation in the Forecasting and
Scheduling process.
Choose the type of statistic from the choices above for the calculation
to represent, and click Next >.4. Enter the calculation parameters in the Wizard dialogs as they appear.
After each dialog, click Next >.
5. In the final dialog in the path, click Finish > to create the analysis
graph. 
						

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-18-For Erlang C, follow the directions above.  For Erlang B, follow the
directions below.  For Poisson, skip to the section of this manual titled
“Poisson” found on page 21.
Erlang B
Selecting Erlang B opens the dialog box illustrated below.
Probability of Blockage (ERLB_GOS formula)
This calculation relates the Grade Of Service (GOS) to the number
of  Trunks and to the call load (or the number of calls per hour and
the average duration of these).  The calculator accepts the number
of calls per hour, the average call duration, and the number of
available Trunks as parameters Then, it calculates the GOS that
results from the designated call load being presented to the
designated number of Trunks.
For example, you can determine the probability of a Caller receiving
a busy signal when trying to access an outgoing Trunk under the
following conditions: it is expected that the number of outgoing calls
is 150 per hour; the average call duration is 70 seconds; and the
number of Trunks is 3.  In an Erlang B model, the GOS directly
represents the level of Blocking.  This calculation results in a P.34
chance of receiving a busy signal.  This means that a Caller trying
to access an outgoing Trunk has a 34%, or approximately a 1 in 3
chance of receiving a busy signal.  This is a very poor level of
service.  The GOS would typically be in the range of P.01 to P.10.
Probability of X or Less Trunks Busy (ERLB_X_CHAN formula)
This calculation determines the probability that exactly X number of
Trunks will be simultaneously busy when a stated call load is
presented to a given number of available Trunks.  The calculator 
						

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-19-accepts the number of calls per hour, the average call duration, and
the number of available Trunks as parameters. Then, it calculates
the probability that exactly the specified number of Trunks are busy
at any point in time.
For example, you can determine the probability of getting an all
Trunks busy condition in a group of 4 Trunks under the following
conditions: there are 250 calls per hour, with an average duration of
70 seconds.  The calculation results in a probability of about 0.39,
meaning that there is a 39% chance of all 4 Trunks in the group
being busy at the same time.  This can also be expressed as
precisely 4 Trunks being busy for 39% of the time.
Probability of X and Y Trunks Busy (ERLB_INTEG formula)
This calculation determines the probability that between X and Y
Trunks will be simultaneously busy when a current call load is
presented to a limited number of available Trunks.  The calculator
accepts the number of calls per hour, the average call duration, and
the number of available Trunks as parameters. Then, it calculates
the probability that no fewer than X, and no more than Y Trunks are
busy at any point in time.
You can use this calculation to determine the traffic on the least
occupied Trunks in a Trunk Group.  For example, you can calculate
the probability of Trunks 8 to 10 (for a group of 10 Trunks) being
busy when presented with 250 calls per hour, and having an
average duration of 70 seconds. This calculator calculates the
probability of Trunks 8 to 10 being busy and determines if the
individual Trunk traffic justifies the cost.
Traffic that can be Carried (ERLB_TRAF formula)
This calculation determines the traffic (call load) that can be
handled by a limited number of Trunks within a certain GOS.  The
calculator accepts the number of Trunks, the GOS required to be
met, and the average call duration as parameters. Then, it
calculates the number of calls per hour that can be handled.
Alternatively, the number of calls can be used.  In this case, the
calculator determines the average call duration needed in order to
meet the GOS.
For example, you can determine the call load, with an average call
duration of 70 seconds, that can be handled by three Trunks while
maintaining the GOS within the companys target of P.03.  If this
calculation shows that the number of calls per hour that can be
handled is less than expected, then the number of Trunks must be
increased to maintain the target GOS.  Otherwise, the level of 
						

							Traffic Analyzer                      TASKE Call Center Management Tools Version 7.0-20-Blocking must be allowed to worsen to enable the call load to be
handled by the Trunks that are available.
Alternatively, the same calculation might be performed with a
known number of calls per hour.  The calculator then determines
the average call duration that can be accepted while still remaining
within the target GOS.
Trunks Required to Achieve a GOS (ERLB_TKS formula)
This calculation determines the number of Trunks required to
handle a certain call load within a stated GOS.  The calculator
accepts the number of calls per hour, the average call duration of
those calls, and the required GOS as parameters. Then, it
calculates the number of Trunks required to handle the call load
within the required GOS.
For example, you can determine the number of Trunks required to
handle 150 calls per hour, with an average duration of 70 seconds,
within a GOS of P.02.  This is a key calculation in the Forecasting
and Scheduling process.
Choose the type of statistic from the choices above for the calculation
to represent, and click Next >.4. Enter the calculation parameters in the Wizard dialogs as they appear.
After each dialog, click Next >.
5. In the final dialog in the path, click Finish > to create the analysis
graph.