Have You Considered BI for IT Service Management?

A few months ago, I blogged about the fact that, while we were getting “excited” about Cloud and Social in the context of IT service management (ITSM), we were somewhat neglecting the impact of Mobile on our ability to deliver high-quality IT services (Social? Cloud? What About Mobile?). At the time, with the title of the blog tantamount to IT buzzword bingo, I chuckled to myself that all I needed was to throw in a reference to Big Data and I could have called “house.”

What do we do with all the data imprisoned within our ITSM tools?

Big Data? No, not really, more BI

While the Big Data perspective will be seen as a little too “large” from an ITSM tool data perspective (the Wikipedia definition of Big Data describes it as “data sets whose size is beyond the ability of commonly used software tools to capture, manage, and process the data within a tolerable elapsed time”), I can’t help think that these considerably smaller ITSM data sets are still ripe for the use of business intelligence (BI).

We have so much valuable data stored within our ITSM tools and, while we leverage existing reporting and analysis capabilities to identify trends and snapshots such as Top 10 problem areas, do we really mine the ITSM tool data to the best of our ability?

If we do (I can’t say I have had ITSM tool vendors making a song and dance about their capabilities), is it something that is both easy to implement and use?

Why am I bringing this up now? Are things changing?

Hopefully yes.

Over the last couple of months, ITSM tool vendor RMS has been talking to me about (and demonstrating) their new ITSM offering RMS Vision. Powered by NeutrinoBI, it is described as an Agile or Self-Service BI solution that can be applied to any ITSM tool not just RMS’.

The product blurb offers descriptive phrases such as:

  • “provides the ability to access and analyze all information immediately and intuitively.”
  • “make fast business decisions without cumbersome information collation.”

Whether it is using RMS Vision or another Self-Service BI tool, you can’t deny that there are many opportunities to make better use of all the information we store in our ITSM tools.

What can I do?

While there are obvious uses in the context of trends in volumes, response metrics, and costing information, I can see other, more interesting, uses such as:

  • Mining incident data as part of a service portfolio management or service catalog creation exercise. Using service desk data to help create a service taxonomy; we have the former but often lack the latter.
  • Taking a detailed look at the incident classification hierarchy to understand its fitness-for-purpose and opportunities for “tweaking” that make it better suited to both operational use and back-end reporting and analysis.

I could go on (with a little more time to think) but I consider YOU far better positioned to consider, and to articulate, how we could best use BI for ITSM.

So what do you currently do (by way of leveraging BI in the context of ITSM)? What would you like to do?

The ITSM Community is far smarter than the combined analyst community, so throw me some ideas :)


Please check out my latest blog ... http://blogs.forrester.com/stephen_mann


I've always used BI

I've always used BI principles and tools to mine and extract data from helpdesk and other databases. In recent years Microsoft have made it very easy for SQL Server databases with the continual improvement of Analysis Services. Even more accessible is the Excel Pivot Tables and now Power Pivot. Many people are already familiar with Excel, so it becomes a very usable BI tool at the desktop level.

I wrote a short tutorial for HelpMaster a while ago on how to connect Excel directly to its SQL Server database, but this connection/reporting method applies to any database. Any ITSM product that has a database that can be connected to via OLEDB, ODBC, or native connections makes Excel Pivot tables easy.


For small data sets as you mention, you can get away with just using direct queries that are constructed with a BI-ish schema. For larger data sets, it is best to use OLAP and cube it up.

I think most solutions should have this capability.

I remember a few years ago doing a study on contact methods between helpdesk operators and clients, based on contact method and gender. I was looking for patterns on contact with the client based on the respective genders. In many data sets that I examined, there was a higher incidence of helpdesk operator males calling females via telephone as opposed to email or other mechanisms. Go figure!

BI is good fun for the data hounds. Helpdesk and ITSM data is extremely revealing, and you're right, there is not enough of this style of deep-analysis in the market.

Thanks Rod ...

... I knew people would be doing it :) Thanks for the offered advice.

Connecting BI Semantics and Syntax

I think two separate points are being discussed here. Stephen's point asks where are the greatest gains from the application of BI to mid-sized ITSM lie, and I think it's a very valid question. However, Self Service BI opens such a wide exploration window it can be difficult to identify the prime areas of benefit as anything other than specific examples. I think it's all dependent on the individual need at the time; so it's a little like asking Ranulph Fiennes what his favourite commute is. Although the tool is very new I've already seen Operational Managers, Contract Managers, Financial Managers and Service Delivery Managers' eyes light up in realisation of what that mass of ITSM information could do for their individual roles by asking their questions from their perspective, particularly when other data sources or data islands are pulled in. It also varies day to day, just as business does.

Ron's point avocates Excel for ITSM BI. I agree that Excel is a very powerful tool and most of us familiar with Excel can readily use pivot-tables for where the schema size is tiny enough that the fields are memorable and intuitive. That's simply not the case for mid to enterprise ITSM instances with 1000s of data items and 100s of KPIs....asking a senior manager to relate performance measures to specific fields, or combinations of fields, across many tables and multiple data sources is where the power of semantic-driven self-service BI such as in RMS Vision, comes to play. It does not pre-requisite Excel data analysis skills. Meta data provides that link between semantics and sql syntax and empowers the manager to ask questions in their native language, without understanding what combination of fields represents that meaning. This is where the raw-level analysis provided by Excel is practically limiting.

RMS Vision recognises the powerful data manipulation that Excel provides, however, and that it often holds personal or silo'ed data, so has mechanisms for using it in the the analysis loop.