Sunday, April 29, 2012

Ever increasing priorities

Inspired by a recent Martin Widlake post, I also have something to say about urgency.


In most problem/incident management systems there is some 'priority' level. And all the ones I've met, count it the wrong way, by starting with "1" as the highest priority.


The reason this is wrong is that, as anyone who has worked on the receiving end of such a system, there is always something that will trump the current highest priority.


"1" should refer to the lowest priority item. A typo in the help text (or the use of the phrase 'most unique'). The items that will never actually get fixed, but you record in the system just to let everyone know you are aware of the issue.


Then you work your way up to level 2 where the work might one day get done by some new guy as part of a learning exercise, but where it isn't something that is affecting the business in any material way. Actually a lot of issues at this level are trivial to fix, so if someone needs to 'get the numbers down' this is the place to target. 


Level 3 should be the 'regular' work. In theory. In practice, you'll find most stuff at this level is actually level 2 items that have been pushed up because someone wants to see some progress. But also the hard stuff tends to sit here a long time as people argue over solutions.


My level 4 would correspond to what most systems class as urgent. The business expects all issues at this level to be actively being worked on by someone. If you have 20 issues at this level and 5 team members, this won't actually be the case though.


This is where corporate politics gets involved, and who has the biggest influence. And this is why I recommend counting from the bottom up. Because as people start pushing their issues up, there is unlimited room for the priority levels to be increased. Levels 5 and 6 will quickly come into play. After a few months, someone will find a need for level 7. When people find out there is a level 7, the issues at levels 5 and 6 will be pushed up the rankings.


Don't settle for single digits though. You'll hit 9 within a year and then you'll need a high priority issue to make enough room for a level 10.