Newsgroups: comp.admin.policy
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!think.com!sdd.hp.com!elroy.jpl.nasa.gov!decwrl!hal.com!halaus!ables
From: ables@hal.com (King Ables)
Subject: Re: User Satisfaction ?
Message-ID: <1991May21.190130.1155@hal.com>
Organization: HaL Computer Systems, Inc.
References: <1991May21.163045.25115@cs.utk.edu>
Date: Tue, 21 May 1991 19:01:30 GMT

>>Do any of you have metrics for user satisfaction?  
> 
> How about response-time statistics for problem calls?
> How about system availability?  E.g., a table showing the
> uptime/downtime for each system.

Both are good metrics of how well you're servicing your users,
but neither will tell you beans about user satisfaction.  Users
can be very dissatisfied in an environment that's up 99% of the
time if the 1% was the hour before a big presentation.  And a
*big* part of user satisfaction comes from the attitude they
perceive you have about helping them get their problem solved.
If they feel like you're doing it only because you have to, and
not because you really care about it, the satisfaction level will
be lower, no matter how well you do it.  You're heart has to at
least *look* like it's in it.  And that's hard to measure, at least
objectively.

> A comprehensive direct user survey can be dome occasionally, but most

Yes, if you can get people to take it seriously, this is one of the
best tools... like he said, do it infrequently and make a big deal about it.

> problems.  Then you could show how, e.g., queries about remote
> printing have dropped X% since you published a handout on the topic.

This is a good way to use the problem call stats!

If the presentation is really meant to estimate user satisfaction,
it's going to be real hard to back up.  You can list all the things
that have been suggested and conclude that the users "should be satisfied,"
but that's as far as you can go.  If management wants to know if people
are satisfied, maybe they should ask them themselves.  ;-)

If you don't have some kind of e-mail address or newsgroup for local
suggestions, you might want to add that.  I think people are usually
willing to suggest improvements when it's easy enough.  How much traffic
showed up there would be an indicator of how comfortable people are with
the way things currently work (assuming you could be sure people would
really use it if they felt the need).

Just the rantings of a former user-support-type, for what it's worth.
-king
