What do you consider to be acceptable?

Discuss and debate ITIL Change Management issues
Post Reply
User avatar
changeborg
Senior Itiler
Senior Itiler
Posts: 42
Joined: Tue Jul 14, 2009 8:00 pm
Location: United States

Tue May 15, 2012 10:22 am

Our organization is having a hard time coming to a consensus on what an acceptable percentage is for both failed and emergency changes. While we all strive for zero, we know in a realisitc world that it is simply not possible. I've tried to get some information from Gartner and other similar organizations without any success.

So I'm curious what you consider to be an acceptable percentage in both areas? We are a global organization that processes an average of 1000 changes per month. So far this year, our percentage of Emergency changes during the last 3-4 months is averaging about 6% and failed is below 2.5% even during our worst months.


User avatar
LuLoo
Itiler
Itiler
Posts: 5
Joined: Mon Aug 23, 2010 8:00 pm

Tue May 15, 2012 5:21 pm

I asked a similar question some time ago....modules.php?name=Forums&file=viewtopic&t=5135.

The answer to most of these questions often begins with "it depends". If your emergency changes are preventing or reacting to Major incidents then you need to look at the cause of the incidents and get problem management to work on the causes.
If you analyse the failures you can recommend improvements to stop repetition or similar failures.
Sometimes, as long as you're capturing cause, trending and reporting there isn't much more change management can do unless your change management function sits in a very influential part of the IT department.

On the face of it your figures don't look too bad but 15 emergency changes a week sounds a lot.
User avatar
changeborg
Senior Itiler
Senior Itiler
Posts: 42
Joined: Tue Jul 14, 2009 8:00 pm
Location: United States

Tue May 15, 2012 5:45 pm

LuLoo wrote:I asked a similar question some time ago....modules.php?name=Forums&file=viewtopic&t=5135.

The answer to most of these questions often begins with "it depends". If your emergency changes are preventing or reacting to Major incidents then you need to look at the cause of the incidents and get problem management to work on the causes.
If you analyse the failures you can recommend improvements to stop repetition or similar failures.
Sometimes, as long as you're capturing cause, trending and reporting there isn't much more change management can do unless your change management function sits in a very influential part of the IT department.

On the face of it your figures don't look too bad but 15 emergency changes a week sounds a lot.
Thanks for the other thread to review. Despite my efforts, I didn't find that one so it was good to read up on it.

To clarify a bit more, we have 5 regions that all filter through the process so while there may be 12-15 emergency changes per week it really breaks down to a few per region per week (on average). We have probably a dozen internally managed data centers along with around 6 that are outsourced and centrally managed. Between the application and infrastructure side, we probably have at least 40,000-50,000 CI's although actual count is unknown due to a poor CMDB (current project running to resolve that).

I will say we are very immature in all other ITIL streams so our problem management is almost nil and while the right answer is to go down that path, it's simply not available to us at this junction. We also do try and analyze the data but with around 1000 changes per month to touch, we simply don't get around to reviewing everything due to a small team.
User avatar
Diarmid
ITIL Expert
ITIL Expert
Posts: 1894
Joined: Mon Mar 03, 2008 7:00 pm
Location: Helensburgh

Wed May 16, 2012 5:52 am

An acceptable number of failed changes is zero.

An acceptable number of emergency changes is zero.

Of the two it is the emergency changes that can most plausibly be accepted above the acceptable level, but not much above. This is because an emergency change can be required for circumstances reasonably outwith both your control and your powers of divination.

thus far, I'm in agreement with your sentiment.

However, I do not believe that Gartner or anyone else outside your organization has anything useful to offer in the way of figures.

There are two main platforms from which to manage improvement in this area.

The first is to measure what you have and aim for better by analysing the background and making improvements and the second (which should be first) is to respond to your customers (and to your service costs) pain areas and fix those.

Much more important than the percentage failed and emergency changes is their nature and impact and that is why outside figures are useless to you - they cannot be mapped to your situation.

As you say, the granularity with which you can manage this is constrained by your resources. Therefore it is so much more important to focus on the pain areas than the numbers.
"Method goes far to prevent trouble in business: for it makes the task easy, hinders confusion, saves abundance of time, and instructs those that have business depending, both what to do and what to hope."
William Penn 1644-1718
Post Reply