The WorkForce Management Guide
to:
Center Performance Data Integration
--Part Two--
Benjamin Funnell
WFM Consultant
Does
your contact center accurately report center performance
to calculate future staffing requirements? You may be surprised
what you uncover by looking under some unturned stones.
A CLOSE LOOK AT AHT
In the simplest of terms,
AHT is the sum of average talk time (ATT) and average after
call work (ACW). ATT is the time from the beginning of time
an agent begins working on the event with the customer to
the time the customer leaves. After the customer has left
the agent, further time may be necessary to complete the
work generated by the initial event known as ACW.
As simple as it seems, Average
handle time, AHT, is one of, if not the most frequently
misrepresented data elements in contact center performance
reporting. The reasons for the misrepresentation are the
lack of understanding the variables comprising AHT, the
effect of hold time, the inability to apply actual handle
time to the same interval as the event worked, inconsistent
agent workstate management, and an activity allocation issue
caused by multi-skilled agents. Each issue has the potential
of causing an extreme impact to the over-all ability to
report actual performance in addition to the forecasting
of staffing requirements.
AHT Integration Process
Options
The actual values reported
for AHT performance are dependant upon the location of the
source data and the integration process used to transfer
data from an ACD to WFM. There are primarily two methods
for integrating data from an ACD, real-time and the reporting
of performance by time-increment. Each of these methods
will report different AHT values.
AHT Interval collection
is the sending of the AHT value each half hour, or other
determined length of time, from the ACD and collected within
the WFM system. In the majority of ACD, the time used to
derive the AHT value is based upon the interval in which
an event ends. If an event starts at 11:58 and ends at 12:02,
the event was 4 minutes in length. All 4 minutes of work
will be applied to the 12:00 to 12:30 increment. The negative
effect of this process is that the AHT will be understated
for the 11:30 to 12:00 while 12:00 to 12:30 will be overstated.
The greatest effect caused by this integration method is
seen at the beginning and the end of the day because the
peak periods of the day will have both negative and positive
effects being applied. A positive point of using this process
is that the data is retrievable from the ACD as long as
it stores the data, data integration can occur at any time.
The alternative approach
to reporting AHT is through the real-time collection of
agent workstate time. This process consists of sending a
continuous datastream from the ACD to WFM containing all
changes to agent workstates. The total amount of time within
each workstate is then calculated and divided by the total
number of events handled over the same increment of time.
This process applies the work to the proper increment. A
negative point is that if the datastream is down for any
length of time, the data cannot be re-sent and is lost.
Note needs to be taken that not all ACD will support real
time data streaming.
The real-time process is
also dependant upon employees being properly defined within
the WFM application to associate the work time. All employees
that potentially work scheduled activities need defined.
If an agent, lead, supervisor, trainer, etc. works an activity
and is not defined, the time worked is lost resulting in
an incorrect reporting of AHT and ultimately incorrect forecasted
staffing requirements.
Average Talk Time
Average talk time (ATT),
the primary component of AHT, is straightforward but how
hold time is represented needs to be reviewed. Talk time
accuracy issues arise when hold time is not known to be
included. Work time can possibly be missing if hold time
is not counted or even overstated if double counted. If
hold time cannot be factored into the AHT, the work time
should be summed regularly and included as a shrinkage category.
After Call Work
After call work (ACW), the
follow-up work that is required from a previous event and
the second primary component of AHT, is the most common
element that causes inaccurate workload performance reporting.
Reporting ACW is difficult because it is dependant upon
an agent’s ability to place their phone or CTI into
the proper work state and due to the affects of multi-skilled
agents.
Determining what activities
are supposed to be included in ACW is vital to properly
reporting performance. Ultimately, all activities are defined
prior to the setup of data integration. Usually the guideline
for determining if an activity should be included in ACW
is if it is a direct result of an inbound activity. Activities
that are not a direct result of an inbound activity should
be defined as a shrinkage category and be managed and planned
independently. For example, training is an activity but
not associated to an event generated by a customer. The
activity of sending a fax to the plant to finish a customer
order, in most instances, is included in ACW.
Upon determining the different
activities to be included in ACW, the agents need to be
educated on what phone and CTI workstates represent what
activities. This education is ideally supported through
quality control monitoring and coaching.
Multi-skilled agents make
ACW even more complex. Centers with multi-skilled agents
using an ACW will not be able to associate the ACW time
to the proper activity. If agents are assigned to handle
two activities and place themselves into ACW, there is no
link to associate the work and activity. Some ACD and WFM
applications will report the ACW time to the most recent
inbound activity, while other ACD will only report the overall
ACW time.
A CLOSE LOOK AT SOME NUMBERS
Let us look at some efficiency
numbers to see the impact of not staffing properly due to
an inaccurate representation of center performance.
In this scenario, a center
has one activity and handles 1000 events during a day with
a talk time of 3 minutes. However, the after call work does
not include all of the after call work activities and reports
45 seconds instead of 1 minute. The overall cost is that
the activity will be understaffed by approximately 6%.
