Service mgt tool Categories for Incident Management

[Gazza]

Hello,

I'm currently reviewing my companies Incident Management tool and it has over 3500 incident categories available to choose from. This is a lot for any service Desk analyst to sort through every time they have to log an incident.

Now I know most will say the tool needs to be data partitioned but this is something we cannot do on the tool we use.

Does anyone have a list of Generic Incident categories that would be useful to implement and easy for incident trending?

[rjp]

This 'difficulty' is more common than you might expect.

Incident classification issues come up quite a bit on this forum - and I have posted quite a few times on the topic. Quite a few of those address your question in varying ways. So I will keep this reply to a few key points. If you want you can browse my posts by clicking on the profile button below and using the 'all posts' link on my profile. I'd also suggest searching or browsing the froum for other people's past comments.

At 3.5K kinds of incidents you are going to be pressed to get accurate, fast, and consistent incident classification.

First up: Are you classifying incidents or your infrastructure? If your categories read like a schema for your CMDB - every type of component listed - you are effectively classifying 'errors'.

Incidents are disruptions to services - they are only formal 'break-fix' events after you have engaged probelm management.

Your classification should be sufficient to identify who is going to do the resoultion, what SLA based response times you are committed to, and which services have been impacted. Services, not things are the key to effective and efficient incident classification schemas.

I work in a very large ICT environment with most everything you would expect to find, and then some - and we have a quite serviceable incident management process with about 115 categories of incident classification - including those that take care of service requests and general queries.

Editorial Aside: Consultants and toolset vendors often assume customers want component based incident classification: Or see this as 'low hanging fruit' because they know their customers are receptive, and it's an easy setup option. But they should know better and advise against it - once you start down that road... well you end up with just the situation you described. Meanwhile you are not 'picking the rest of the tree' and it's value is rotting on the branches.