Define Measure Acheive. Repeat

Archive

Posts Tagged ‘SLO’

Do you measure your incident backlog?

May 5th, 2010

Incident backlog provides an informative KPI that you should consider adding to your reporting repertoire. The KPI should measure the number of incidents outstanding that have missed an SLO or SLA. The KPI should be trended over time and should either be stable or decreasing.

In the example chart below, we can see the backlog rising over the last three months.

Incident Backlog

In this example we see the backlog increase by 50% over 3 months and should be investigated. To determine how urgent the issue is the first thing to explore is to breakdown the  backlog by incident priority.

Open Incidents

It would be highly unlikely that a backlog with high priority issues would persist over a period of time and the chart above now shows that a majority of the back log are priority 3 and 4. This is fairly common and is often systemically ignored. However, it is worthwhile to examine further. 

The reasons for the rising trend could include:

- Second line support personnel are not closing tickets in your ITSM tool.

- Resourcing level may not reflect the current volume of tickets being received on the Service Desk.

- The organizations Change and/or Release practice is causing unscheduled spikes in incident volume

- Are their particular workgroups with particularly high backlogs that could indicate a bottleneck?

  • Share/Bookmark

Charles Cyna Uncategorized , , , , , , , , , , , , , ,

Anatomy of a KPI – Mean Time to Restore Service (MTRS)

March 3rd, 2010

Mean Time to Restore Service is an important KPI that most help desks measure (or at least should). MTRS tells us about the average customer experience a user has when a service interruption is identified.

To calculate the MTRS you take the total amount of time of open incidents divided by the total number of incidents logged in a given time period (normally a month). I would recommend that the KPI only show the top 2 tiers of classification as performance on lower classification would probably reduce the usefulness of the KPI based on how most organization service their lower priority issues.

Now the usefulness of KPI is just that, ‘an indicator’. If it is going in the wrong direction (i.e. up) there is no reason to panic – the most important thing is to identify whether there really is an issue and if so then be in a position to address it as soon as possible.

practice-indicators1

The first thing to look at in regard to MTRS is to see whether Incident volume has spiked. When incident volume changes unexpectedly, the help desk doesn’t have a chance to change resourcing so the average time to restore service will often rise.

analyze-your-mtrs1

The second thing Read more…

  • Share/Bookmark

Charles Cyna Uncategorized , , , , , , , , , ,