Good practice of requesting incident via Service Catalog

[mitzaka]

Hi SNC,

A dilemma here - what's a better approach if I want to give people the ability to submit incidents via the Service Catalog. From a tool's best practice and ease of use perspective:

1. Have a catalog item with variables to capture questions and a workflow behind, based on Requested Item table, which generates a related Catalog Task to the assigned group. Then decide if that is really an incident and convert the Catalog Task to an Incident.

2. Have a record producer with variables to capture questions. Use an incident template and directly create incident, which the first line of support could convert to Service Request (RITM) if the submission is not really an incident.

I guess that's an implementation question of design. We currently have option 1 in place, but I am wondering if we should simplify it. Are there any downsides to option 2?

Thanks in advance and will be glad to hear how some of you have this implemented.

[Kalaiarasan Pus]

no .. we are not using any templates for incident .... we have some of the fields being auto populated based on current user..

for the second point on the icon, not really sure.. will check and get back to you ....

[Anurag Tripathi]

I like the idea mentioned in option 2,

Based on the Record producer you can easily get a lot of information before actually creating the incident, automatically applying the template at creation of incident also sounds good.

Help desk will be happy . I don't see any downside also.

[randrews]

i think it depends do you have an SLA with your customers for incidents... if so that sla probably is supposed to start when they open the initial ticket.. not when someone reviews it and decides Yep that really is an incident not a request...

if that is true.. i suspect you are pretty much stuck going with scenario 1.. of create an incident right off <starting the sla> and then if required convert it to a request...

[mitzaka]

We are currently running case 2, where we create tasks and convert them to incidents, and we have SLAs for both the request and then the incident. But you are right, I think it's complicated this way, so this is why I am considering change to version 1.

Just seems so much more transparent.