off the topic, users satisfaction survey for network services.

CCIE_2011

Hi all ....
I understand this is not a cert topic, but i have no where else but here


Here in my company we are thinking to develop a satisfaction survey questionnaire for network services only.

Have any one done such thing before ?
Do u think that it will be hard for the user to differentiate between network services and other factors ?

thanks in advance.

Kaminsky

How would a standard user know the difference between the network being slow, the server being slow or their own PC being slow ? High probability is that they won't and you'll just be sending out another junk email nobody pays attention too. May look good for the IT bosses looking for cudos amongst their peers but that will be it. Otherwise, you are on a hiding to nothing.

If you measure the traffic across the backbone and subnetworks and those figures are not constantly high then that is your real measure of your performance. From there you can investigate the types of traffic and squirrel down to what may really be causing bottlenecks.

If the end user sniffs a chance at a fast PC or the company paying out for a faster doohickey, they will say yes to the PC every time.

Forsaken_GA

For most users, network services are pretty transparent. You'd have to be careful how to scope such a request.

To give you an example, a company I used to work for was a reseller. We used a piece of crap inventory app called Navision. The app wasn't installed directly on our computers, it was a Citrix connection to another server.

During the peak hours when everyone was at their desk, it crawled. Everyone complained about how slow the network was. Except for us guys over in the Cisco side of the business. The issue was never the network itself, it was the fact that the SQL server on the backend simply had a hard time keeping up with all the transactions. Once the guys responsible for it moved the SQL backend to a clustered solution, the problem magically disappeared.

Another company I worked for was a call center environment. One day, they decided to add monitoring software to our computers, they wanted to be able to see what we were doing on our computer screens while at the same time recording our calls. (In other words, they wanted to see which Everquest sites I was browsing while I fixed people who managed to break their Internet Explorer...).

Well, after they rolled this out, all of a sudden, performance throughout the entire call center went to absolute crap. No one could use anything really well. Everyone blamed the monitoring software for slowing down the computers. It turned out we were all on a 10mbit backbone throughout the call center. When you're monitoring real time video and voice traffic from several techs, that tends to be performance effecting. They upgraded the backbone to 100mbit, and it was better, but still a little sluggish. So they dual homed all of our work boxes, cabled another backbone, and sent the capture traffic out over it, and everything performed fine again.

The point of both of these stories is that users have a hard time from distinguishing the difference between a network problem, a local problem, or an infrastructure problem. If you're trying to gauge network issues in a survey, you'll have to ask some pretty targeted questions.

redwarrior

Ditto on what everyone else has said...

If your users are anything like my users, then they think anything in the world that is slow, or even just seems to be slow or is slower than they'd like, is network "slowness." Have one part of one application that's "slow," then it's "network slowness." It stinks and I think taking a survey on it would be pretty much useless where I work and would only serve to increase the ire of everyone who supports the network.

CCIE_2011

I really appreciate all your comments .... now I really have a good vision and some guide lines.
I what i am thinking now is to prepare some satisfaction survey for specific IT employees(e.g. servers section) and may be another one for all users for services such as internet and maybe cabling section response.


I'd like to thank everybody once again for you comment

Sepiraph

Do you really want to get back survey answers complaining that the 'internet is down' when their email doesn't work?

mamono

If the survey is targeted toward a specific recipient, then I don't see what the problem is. As recommended earlier, the questions are going to be very specific per type of recipient. That should be perfectly fine, but keep in mind to keep the diction/questions/survey at a tech savy level that corresponds with that type of recipient.

mikearama

Sepiraph wrote: »

Do you really want to get back survey answers complaining that the 'internet is down' when their email doesn't work?

Hey, I didn't know we both worked for the same company!

Classic:
Dilbert - Internet is Full - Joost