RE: Statistical Analysis

["Alan Meekings" <alan@landmarkconsulting.co.uk>]

Hi Alistair,

I've actually had a lot of experience in measuring performance in call
centres and contact centres, dating back to the earliest manifestations of
Automatic Call Distribution technology.

I even published a number of articles in the International Journal of Call
Centre Management, the last two which may be interest you are:

(1) "New Measures for a New Era: the Vital Link between Measurement and
Culture in Call Centres", January-March 2000.

(2) "Unlocking the Potential of Natural Language Speech Recognition",
June-July 2002.

What's particularly interesting about call centres - at least, from a
performance measurement perspective - is that: (a) almost everything in call
centres is measurable; and (b) given that staff may be answering over a
hundred calls a day, the statistically significant impact of any systemic
changes made can be seen in hours or days, rather than weeks or months.
Hence, it is arguably one of the most exciting areas of organisational
activity for the application of new (Deming-based) thinking on performance
measurement.

If you'd like to contact me off-list, to let me know a bit more about your
specific circumstance, I may be able to help.

Meanwhile, three points are worth making in general terms:

(1) The measures themselves must be genuinely systemic.

(2) All process-based data must presented in SPC format. The software for
SPC charting is now so inexpensive as to be affordable by all. I personally
favour WinChart (http://www.winchart.net)- it's intuitively easy to use, and
also has a new 'performance dashboard' feature that greatly aids the
management review process.

(3) The key to performance measurement lies not in measures themselves, nor
how they're determined, but how they're actually used in practice. If you
can, grab the review process in your call centre, and then do everything
possible to make it work properly. Tackling the review process effectively
'pulls through' insight and improvement, rather than 'pushing it at' the
organisation.

Regards as ever,
 
Alan
 
Email: alan@landmarkconsulting.co.uk
Mobile: 07785 258 741





(1)	Optimise the Value of Measurement. To optimise the usefulness of
measurement: (a) it is essential to include measures that matter to
citizens, customers and consumers; (b) the coverage of measures needs to be
sufficiently comprehensive to provide a genuinely ‘systemic picture’ of
performance; and (c) data needs to be captured sufficiently frequently to
track the ‘heartbeat of performance’.
(2)	Present Data Visually to Aid Intelligent Interpretation. To derive
the maximum value from data: (a) process-based data needs to be presented in
a time-series format, ideally showing the limits of predictable variation;
and (b) large volumes of data need to be presented in a ‘performance
dashboard’ format to highlight insights economically worthy of attention.
(3)	Stucture the Review Process Properly. To structure the review
process properly: (a) there needs to be a clear definition of who needs to
come together to review what, where, when, why and how; and (b) the
frequency of review is critical. Typically, the faster the cycle of review,
the faster is the rate of progress.
(4)	Ensure that Review Meetings Deliver Value. Review meetings will
deliver value provided that: (a) the measures themselves are useful [see (1)
above]; (b) the data is presented in a visual format that aids intelligent
interpretation [see (2) above]; (c) the person accountable makes the review
process matter; (d) no arbitrary numerical targets are set (goals are fine,
as long as the culture promotes root-cause analysis, problem-solving and
innovation, but arbitrary numerical targets are not); and (e) people need to
be held accountable for outcomes.
(5)	Connect the Review Process between Levels. Connecting the review
process between levels means: (a) promoting problem-solving at the local
level; (b) putting in place a mechanism for best practice transfer where
appropriate (particularly in large organisations); (c) creating a mechanism
for ideas (and outline business cases, where appropriate) to be submitted
upwards for incorporation into properly planned and resourced improvement
initiatives, sponsored at top level, and for feedback to flow back to those
individuals who offered them up; and (d) allowing issues to be addressed at
the most appropriate level of review from a systemic perspective.
(6)	Tackle Implementation Progressively. Implementation will succeed and
take root if: (a) the approach adopted is more like doing a jigsaw (i.e.
tackling the easiest pieces first) than moving forward on a broad front; (b)
the initial measures adopted are clearly positioned as ‘provisional’ (i.e.
open to revision in the light of subsequent experience); and (c) the person
at the top of the organisation pays serious attention to grabbing the review
process and ensuring that it works properly – up, down and across the
organisation.

-----Original Message-----
From: clauson@deming.ces.clemson.edu [mailto:clauson@deming.ces.clemson.edu]
On Behalf Of alastair.cox@contact-energy.co.nz
Sent: 17 May 2004 04:08
To: den.list@deming.ces.clemson.edu
Subject: Statistical Analysis

Hi all
I have been tasked with doing some statistical Analysis of the performance
of the people who answer the phones in our Call Centres.  The main problem I
have is that I know next to nothing about Stats.  I am going to be using
excel to analyse and produce charts into a power point presentation.  The
person who did this prior to me used Pivot tables which seem to be
reasonably straight forward, sort of.  Any clues advice or the like?
Cheers

Alastair Cox

Message posting through the Clemson CQI Web Server.