Re: Customer?

[Eugene Taurman <ilx@execpc.com>]

James just a few reactions to you example which you already know but
reactions anyway.

>Over two years ago  XYZ  Associates conducted a customer research survey
for our company. One of their
>major findings was that while our customers liked our
>products (quality and selection) they said our delivery was
>lousy.
>
>The following actions took place:

>2. An on-time shipment was operationally defined (customer
>commit date vs. ship date) and standardized throughout
>the company. On-time=0 days late/3 days early.
 
Always avoid operational definitions (specs). Simply measure deviation from
perfect with a run or X bar and R chart. It easier and sends a better
message. Also people will meet spec and not actually improve the process.

>3. I was asked to set up a process behavior chart to track
>the overall percent of on-time shipments monthly (separate
>XmR charts were set up for 8 different product lines).  
Perhaps measure by product line is ok but only if they have different
processes. Chances are they require a mix of the same processes.

<<snip>>
Apr-93.13%
>May-93.16%
>Jun-92.11% (UCL=95.5, LCL=90.4, Xbar=92.3)

People use what management measures and acts on to decide what is important.
Were processes permanently changed or was this achieved by bringing
attention to the issue. Forcing and issue can make it better but it won't
stay changed unless the underlying system is changed. The data surely looks
like the system has changed. 

Can you identify the reasons the delivery has changed. If not we may have
the same old system working under duress.
Did cost go down and quality improve at the same time Unless they did there
was likely no basic process change. Only duress.

Eugene Taurman
interLinx Consulting
414-242-3345  e-mail ilx@execpc.com
fax 781-459-825
http://www.execpc.com/~ilx 
======================================================================