SD Metrics Definitions

Overall Client Satisfaction
with Call Center

Question 7 (Overall) from the weekly per-ticket client sat survey,

The 7 components of client service in the survey do register client satisfaction with each specific behavior, and the Overall question gets the gestalt of the client's take away from the experience. 

CSS HQ runs the weekly survey process.

Surveys report;
QR

This is not a weighted average of the sub-components, it is the answer to question 7.  A weighted average approach is interesting to some and would be a development project.

Tickets per Topic

Count of tickets in a large subset of Service Center queues, including Help Desk::Accounts. 

Tickets are assigned a one-word keyword "topic" based on algorithms and some manual post-processing in the topics tool stored in the topics folder on hd-bs-02's Reporting drive.  75% of tickets are tagged by algorithm; the remainder need to be hinted by a person reading the text used by the tool.

Rob runs the Topics Brio extract which feeds the Topics excel engine.

QR

Reported as a side effect of the ticket Topics extraction.
A parallel process within the Call Center annotates tickets by topic using a very different set of topics.  The daily tabulation takes a SD consultant quite a while to do each day.  Work to do involves reconciling these work streams.

Unique IDs per Cohort

Counts of unique email IDs reported for the group of queues tabulated in Tickets Per Topic. 

Portrays the mix of user types drawing on the Call Center in particular.  Overrides perceived wisdom about who a service is "for" by showing who is in fact interacting with it. 

Rob runs the Topics Brio extract which feeds the Topics excel engine. 

QR

Reported as a side effect of the ticket Topics extraction.
 Data Warehouse feed annotates tickets with "Current Employee", "Current Student", "Non-Current Employee" and "Non-Current Student"; these are the categories we'll use in lieu of any "academic"/"admin" breakdown, unless that level of detail becomes necessary.  It is a tool-building exercise to be more specific than what the feed puts in.

Calls Offered

ACD counts of Calls Offered

Measures Gate 1 (Call Center) traffic, what we normally think of as the help desk call center.  Calls Offered relates to Consultant Hours to drive performance variables like Wait on Hold and Abandon Rate.

Rob collates ACD data into the ACD excel engine

ACD weekly metrics.
QR

Gate 2 is Repair Services; Gate 3 is being installed for Telephone.  There's an opportunity there for more development if desired.

Tickets by Method

Breakdown within the Method field of how a ticket is created.

Tickets are set to "Email" method if they are created that way.  It is the email corollary to Calls Offered; all Voice method tickets were generated by a phone call; all Email method tickets were generated by client email. 

Rob runs the Topics Brio extract which feeds the Topics excel engine.  

Reported as a side effect of the ticket Topics extraction.
The followup to a ticket is usually a mix of calls and emails.  Developing a portrait of that would be a project involving analysis of the transaction detail records in the warehouse.

Consultant Hours Per Day

Daily reading of the total number of minutes (or hours) logged by Call Center consultants in the ACD. 

The relationship between call volume, handle time agent availability and other call center variables is well studied by HDI and others.  Abandon rate and hold times as well as client perception of availability relate to the number of agents  available.  Below a floor of about 35 hours per day, call center service noticeably deteriorates.  The sweet spot is about 40 hours per day, judging by historical patterns in our call center environment.

Rob collates ACD data into the ACD excel engine

ACD weekly metrics.
QR

Wait on Hold

Average Wait on Hold, counting the seconds

Clients typically would wait this long before a consultant answers the phone.  Balances with client impatience to drive the abandon rate.  Higher agent availability per call volume would lower the wait, assuming handle time is the same.

Rob collates ACD data into the ACD excel engine

ACD weekly metrics.
QR

Abandon Rate

% of callers who hang up before an agent reaches them

Standard call center metric.  10% is a typical goal level. 

Rob collates ACD data into the ACD excel engine 

ACD weekly metrics.
QR

1st day resolution

% of Call Center tickets that are resolved that day.

Data warehouse feed calculates a time to 1st resolution figure for every ticket.  First call resolution is often tracked in call centers, but it doesn't take into account email transactions (which are inherently delayed and separate reply from the query) and the MIT preference among many to submit the problem with a call and then get a call back, rather than spend time on hold.  So tickets that are resolved on the same day is a proxy for solving it "fast enough".

Rob runs the Topics Brio extract which feeds the Topics excel engine. 

QR

Reported as a side effect of the ticket Topics extraction.

New

Repair Center Client Satisfaction

Question 7 (Overall) from the weekly per-ticket client sat survey,

Like the Call Center metric, but understanding that the Repair Center is a very different business and its clients will respond differently.

CSS HQ runs the weekly survey process. 

Surveys report;
QR

New

Repair Center Tickets by SubQueue

Traffic volume in "Hardware" and "Software" subqueues

Knowing the relative volume of each helps with staffing allocation.

Rob

QR

Reported as a side effect of the ticket Topics extraction.
Other Subqueues not tabulated here are "Printers" (which is now right-sourced to KSL) and "Walk-Ins" which don't get past the front desk into the Repair Center

Telephone Moves/Adds/Changes

Traffic volume in MAC requests.

The "how busy are we" measure in CSR interactions.  This is not the same as tickets, as some single tickets can generate many tens of MACs.

Jana T sends a table

QR

How does the shift from Telephone to VoIP affect MAC volume?  The two are aggregated here.

AudioBridge Reservations

The number of requests for this conference calling feature.

How much is this service used by the Community? 

Nancy H sends a table

QR

Would growing use of WebEx affect this?

% using NameConnector

This feature automates part of the "Operator" call for a number

Level of acceptance of this technology by the Community reflected in the % of operator calls fielded this way; preserves operator availability for calls that need more human interaction. 

Nancy H sends a table

QR

VoIPHelp Tickets Created

The telephone call center metric about client demand

Ideally, the more widespread VoIP is the less tickets created per voip installation, as the new technology ought to require less service calls. 

Jana T sends a table

QR

% VoIPHelp Solved in Tier 1

The Telephone equivalent of 1st Day Resolution

Clients are served better if their concern can be handled without escalation to the busy next tier. 

Jana T sends a table

QR

Also a measure of growing CSR skill in handling VoIP related issues, a team-learning effect.