IT Service Management Blog
Follow information regarding IT Service Management via this blog.

How DOES one measure the success of a CMS?

The first question is, are the citizens happy?  Are your consumers satisfied?

 

Are all your use cases being fulfilled?  Can you quantify the trustworthiness of the data in your

CMDB, its actionability?  Do all your CIs have a consumer?  Do you know the approximate time and cost of onboarding a new consumer?  Do you have a process to control the quality of the data coming into your CMS?  Right, I didn't think so.  Besides, except for the first two questions, these are all Configuration Management questions, not business questions.  Even so, these may seem like lofty goals when maybe your definition of success is "stop so many bad changes from getting through."   But we have to peel it back to the underlying problems to really get anywhere.

 

Now I can't talk about measuring the success of a CMS, without talking about measuring the success of that CMS' consumers, or about how to establish and control the elements of that success.  So "How to Measure" can only follow close on the heels of "How to Control and Improve".

 

First, Stop Making it Worse
For starters, don't let just any ol' provider put stuff in your CMDB.  Make a process to rationalize the authoritativeness and accuracy of all providers, including discovery.  If you have a noisy, untrustworthy CMDB and you don't have a process to manage the quality of onboarded data, that's why.  Maybe go read the CMS Best Practices Library CMs Strategy Guide.

 

Get Help
Now there are a lot of very un-CMS-like things you must do up front to end up with a CMS.  It's like drawing a picture of a chair:  you have to draw a lot of very un-chair-like lines before things start looking like a chair.  So you have to do a lot of Security work, Network work, Application work, etc. before you have a shot at building a really good CMS, and this generally translates to getting some good help up front.  So please, get some help, if you are in a position to.  It might make the difference between a high-actionable CMS and the next failed system in the Vicious Cycle of Failed Data Systems.  We help our customers WAY up front when we can.  Planning is key to a successful CMS.  I often hear how someone wants to implement a CMDB but they can't describe any use cases!  How can you succeed like that?  With great cost and difficulty, or not at all.

 

Use an Approach at Least Has a Chance of Success
It's a natural question to ask, "where's the ROI"?  After all, CMDBs are expensive.  It's not that you can't measure CMS ROI, but it's much easiER to measure the ROI improvement of the consumers enabled by the CMS.  The real value bubbles up from the CMS through the layers of the other processes that provide and consume via the CMS.  So, as you improve the KPIs that ARE visible such as MTBF and MTTR, you can show the value of the CMS by the part it played in enabling the improvement.  A basic approach is to correlate the timelines between the CMS deployment and the KPI

improvements.  If you can build efficient and repeatable configuration management processes, that benefit the data-exchanging processes and improve their visible KPIs, then you can demonstrate CMDB ROI. 

 

After all, all a CMS really does is make it easier to make the other processes talk to each other (of course under the right conditions and in the right way).  And I'm talking deployment here, not just operationally.  You will more or less constantly be onboarding new use cases and new consumers and providers.  This means your onboarding processes must be efficient and repeatable.

 

Additional Value Never Hurts
You're likely to find some unexpected value along the way, for example, finding single points of failure, unrestricted local admin accounts, personal hardware, etc.  I call these "Aha!" moments. 

 

Set the Bar High and Keep it High
Aha moments are great, but don't let them introduce scope creep or bypass your onboarding processes.  Once you set a bar, strictly enforce it, and it will become easier to hold a high bar once people "get it".  No Bob, we aren't going to onboard your Nifty Spreadsheet o'Servers because you calculated the chassis color from the latent webcam cache memory.  Don't discover that kind of stuff for a single consumer.  Your best friend is "We already have a provider for that attribute."  End of Bad Data.

 

So if you can put only good data in your CMDB, and if you can deliver that data to the right consumers in the right way, and by doing so you decrease some entropy somewhere up in your ITSM, would you call that success?  If you can do this, you can improve your ITSM KPIs, while lowering IT TCO, and increasing IT's agility - and to the degree IT is aligned with it's business - the agility of the business. 

 

Summary

Focus on 1) the processes 2) data quality (quality = timeliness, authoritativeness, and accuracy) , 3) documentation, especially on your integrations,  4) carefully manage your integrations, it's easy to break them, let them get out of control, etc.

 

The takeaway?  CMS ROI may be indirect, but there's a lot of it once you get it going.  

 

If you are in need of some CMDB ROI, give us a call.   Thanks!

 

Comments
Rajesh Somani | ‎11-11-2012 06:49 AM

Really useful article . We also store a range of HP Repair Manuals and guides 

 

Admin

www.itoperationguide.com

Leave a Comment

We encourage you to share your comments on this post. Comments are moderated and will be reviewed
and posted as promptly as possible during regular business hours

To ensure your comment is published, be sure to follow the Community Guidelines.

Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Search
Showing results for 
Search instead for 
Do you mean 
About the Author
Jody Roberts is a researcher, author, and customer advocate in the Product Foundation Services (PFS) group in HP Software. Jody has worked ...
Featured


Follow Us
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation.