More on Incident Backlogs - handling issues

This is a follow-up to this post about Incident backlogs.

Firstly, I omitted something very important: contract/agency staff as a possible option (thank you to commentator Robert for pointing that out). Out of all the things I suggested, this is probably the one that has the greatest potential for reducing the backlog – provided you have, or can get, the additional budget you need.

In this post, I’m going to discuss some of the problems that might arise if, like Linda, you are actively attempting to reduce a large Incident backlog.

I’ll start with contract staff, great if you can get the right people – but if you recruit the wrong person it will make the job of the Helpdesk/Service Desk Manager harder. My advice to Linda would be: make sure the spread of technical skills is right (naturally), but look for someone with a positive attitude who can manage themselves without close supervision. Always check references and qualifications carefully (I have experienced a number of fake CVs in my time) and don’t be afraid to set a small technical test. It will be an opportunity to see how the person copes, but if doing this make sure your candidate knows before the interview what you plan to do.

Referring back to my original post, and Linda’s new role, there is potential problem with my suggestions in terms of quality. If we make the focus one of closing Incidents, then we run the risk of overdoing this, and Incidents that are not resolved (even though we think they are) are closed. Although this makes are statistics look better for a short while, it does the business we are trying to serve no good whatsoever.

Therefore we need to take steps to prevent this. Firstly, make sure you communicate to your staff the need for quality and explain the potential problem to them, and remind them from time to time about this as the backlog is reduced.

Secondly, consider introducing a two-stage completion process for Incidents if you don’t already have one. This is where an Incident that an engineer believes to be resolved is set to a Pending Complete stage, and then someone else contacts the customer to ensure that the Incident is actually resolved – and then and only then sets the Incident to Complete. This will increase your workload of course, but has the advantage of stopping premature closure issues much sooner.

Thirdly, monitor the number of Incidents re-opened after closure on a weekly basis.