Benchmarking Customer Service

To 'Benchmark' something simply means to measure or judge it against a known standard. This is an extremely simple definition for a process that, when applied to a customer contact operation (call and/or web center), is anything but simple. Here for your benefit are benchmarks to use as guidelines and ways of structuring your thinking:

Every customer contact operation functions through the complex inter-relationship of a dizzying range of continuously fluctuating variables. If you think that was hard to say you should know that it's even harder to manage. The foundation of a well-constructed call center management program includes clearly defined and realistic goals supported by a team that understands how to drive their performance metrics from current levels to the goal.

Benchmarking allows you and your team to know where you customer contact operation stands today. This creates an understanding of how your operation rates as compared to other call centers and also provides a baseline measurement against which future changes can be compared. A handful of the Key Performance Indicators that most call centers track are provided below for your review. Remember that a thorough BPSI benchmarking audit will consider the impact of these variables on your organization and customize them to meet your needs.

Customer Service Benchmarks:

. Total Calls: An internal metric for all calls presented to the center including   blocked, abandoned, and handled. This number may include all inbound and   outbound calls and should also separately report total inbound and total   outbound calls.

. Calls Offered: Total number of calls sent by the phone company to the call   centers'. This count is normally presented as a total and then broken down   by line and generally comes from the phone company in regular reports.

. Blocked: The number of calls that were offered but not accepted by the call   center phone system, or a specific line, due to technical or capacity issues.   This number usually comes from the phone company in regular reports and   should be reported on daily and tracked both weekly and monthly. The goal   is always ZERO.

. Answered: At the call center level this number will represent the total   number of calls accepted into the call center phone system. If calls blocked   equals zero then the number of calls accepted should equal the number of   calls offered. The goal is to answer (accept) all calls offered.

. Abandoned in IVR: A call that terminates prior to having had the   opportunity to complete routing to or listening to a complete FAQ recording.   The goal is to have zero calls abandoned in the IVR system.

. Calls Placed in Queue: An internal metric which is simply the number of   calls placed in the queue (hold) to wait for a live agent. This number is   generated by the call center phone system.

. Average Answer Delay: The average length of time (in seconds) a caller   must spend waiting before the ACD can find an available agent to take the   call. This number is not the equivalent of Average Speed of Answer as it   includes only those calls that actually experience a wait. Also known as   average time of delay. An average queue time of less than 30 seconds is   considered exceptional.

. Average Speed of Answer (ASA): Equal to the total time in queue   divided by the total number of calls answered. This data is available from   the ACD. This number is generated by the call center phone system. 20   seconds or less is a common ASA goal for world-class operations.

. Abandoned: The number of calls that are terminated by the caller while   waiting for a live agent. The goal is zero.

. Abandon Time: An internal metric for the average time (seconds) a caller   waited before abandoning a call.

. Calls per Hour: The average number of calls that an agent handles per   hour, and is equal to the total calls handled during on working shift divided   by the total time logged into the telephone system. This number is   generated by the call center phone system.

. % Service level: (Calls answered in less that X seconds) / (Offered calls   *100). A typical representation of this goal is 80/30 where the objective is to   answer 80% of the calls within 30 seconds. World class operations typically   push service level expectations to 90/20 or better.

Team / Individual Metrics:

. Adherence To Schedule: A measure of whether agent's are "in their   seats" as scheduled. Adherence is calculated as a percentage equal to   (actual time an agent is logged into the system ready to answer the   telephone) divided by (the total time the AGENT is scheduled to be ready to   answer the telephone) times 100. The data for the percent adherence is   taken from the ACD and should be reported on daily and tracked both   weekly and monthly. Typically the goal is 80% or better.

. Post Call Process (a.k.a. After Call Work Time): This is the time after a   call is completed that the AGENT needs to complete administrative work   related to the call. The data for after call work time is taken from the ACD   and should be calculated by individual and group daily, weekly and monthly.   Based on the workflow design it is expected that after call work will average   less than 120 seconds.

. Average Handle Time: An internal metric that is the sum of talk time and   after call work time. With and avg. talk time of <180 sec, wrap time of <120   sec, and hold time of <30 sec the average handle time should consistently   be less than 330 seconds.

. Average Dials per Hour: (outbound only) In operations without an   auto- dialer this is the count of agent initiated outbound calls.

. Average Contacts per Hour: (outbound only) This may be used to   represent the total number of live contacts that an agent made during   outbound calling, or it may be refined to indicate only the number of   Decision Maker Contacts per Hour.

. Average Sales per Hour: (sales programs only) a count of the average   sales closed by an agent. May be represented as a discrete count of actual   deals closed or may be stated in terms or a dollar amount per hour against   an appropriate dollar-based goal.

. Hold: This is the number of seconds that an agent keeps callers on hold and   should ideally be zero.

. Occupancy Rate: Occupancy rate is equal to (Talk time + hold time)/(Talk   time + Hold time + Idle time) times 100. The data to make this calculation is   available from the ACD. In an operation with a 75% utilization target that is   tightly managing idle time it is possible to achieve 80% or greater Utilization   numbers.

. Utilization: Agent total talk time divided by available time. Utilization target   is generally 75% or higher.

. Available Time: The amount of time that the agent was logged into the   phone system and available to take calls.

. Logged In Time: The total amount of time that an agent was logged into   the phone system regardless of status.

. Not Ready Time: The amount of time that an agent was logged into the   phone system but not available to take calls. The BA is considered to be idle   from a phone system perspective.

. Average Talk Time: Average length, in seconds that the agent spends on   a call. Initial target is <180 seconds for agent calls.

. Total Talk Time: An internal metric for the total number of seconds the   caller was connected to an agent.

. Calls Escalated: Number of calls passed from an agent to other staff for   resolution. This number is generated by the call center phone system.   Patterns will emerge over time allowing for a reasonable range to be set for   this number.

. Calls Transferred: An internal metric that is the percent of total calls   transferred from the original agent to an external (to the call center)   number. This number is generated by the call center phone system.   Patterns will emerge over time allowing for a reasonable range to be set for   this number.

Call Quality Metrics:

Call quality numbers are based on key call criteria that have been specifically selected as a result of applicable laws, organizational guidelines, and/or operational objectives. The framework for a quality monitoring program will vary between operations with the one common component being that all world-class operations have robust quality monitoring program in place.