Service Desk survey

[viv121]

hi all,

Back with a question again. ITIL question indeed as I am looking for the best practice known to you. What should be the percentage of user population which should be surveyed ( considering we survey daily)to understand the normal mood of the user population. Gartner confused me here.
Viv

[UKVIKING]

Viv

The issue is whether the users want to respond

There is no guidance from ITIL

What was gartner tellign you that confused you

This gets into the mystic realm of lies, damn lies and statistics

Now me

I would ask 3 people. one may be a user

Then I can say 2/3 of ... or 1/3 of....

but hey .... statistics
_________________
John Hardesty
ITSM Manager's Certificate (Red Badge)

Change Management is POWER & CONTROL. /....evil laughter

[Diarmid]

Viv,

"...to understand the normal mood..."

This sounds very academic/sociometric or something. If it is then you need immense rigour in your survey and for that you need advice from scientists and psephologists (who may also be scientists for all I know).

However, if you are just looking for a measure of general satisfaction levels among your customers:

1. There can be no such thing as a best practice percentage because of some of my other points below.
2. You need to understand your users tolerance to being asked, as Viking pointed out.
3. You need to understand whether your users all do much the same kind of work or have very diverse profiles (if they do, then you need to isolate those in poll results).
4. You need to ensure that the users all understand your questions in the same way and that you understand your questions in that way too.
5. You need to guard against having a self-selecting group doing all the responding because they will not be properly representative.
6. You need to have customer validation of user satisfaction because users often want more than customers are willing to pay for. The relationship between customers and users needs to be thoroughly understood.
7. Finally, the reliability of your sample(s) can only be determined by a combination of statistical rules and independent verification.
8. And then you have to know what kind of thing you can do in response to these surveys. a) it's no use having low satisfaction if you don't know what the cause is; b) it's no use if you do not have the resources and authority to do something about it; c) you may need customer approval for some kinds of actions; d) user expectations may be affected by mere existence of the survey and even more by the nature of the questions asked of them.
_________________
"Method goes far to prevent trouble in business: for it makes the task easy, hinders confusion, saves abundance of time, and instructs those that have business depending, both what to do and what to hope."
William Penn 1644-1718

[viv121]

Thanks for the responses. I especially liked the tolerance levels response of yours. Answers my question.

[UKVIKING]

Viv

You did - for me - due diligence... You mentioned Gartner.. That means you a) know what they do or....... you saw the word in some IT magazine

I think the former so i did not rip your post as a dumb question.

It is a good question and as diarmid and I have stated there are ways...

I always find that if you contract a survey company to take a survey....the report that comes out from the company and a random walk style statistics report... usually will have some similarities

The hard part of doign surveys... is dont do them shortly you when you screw up a service
_________________
John Hardesty
ITSM Manager's Certificate (Red Badge)

Change Management is POWER & CONTROL. /....evil laughter