Thanks to the abundance of IT service management (ITSM) tools - technology suites that enable many of the ITIL-espoused, ITSM best practice processes - corporate IT organizations have a wealth of ITSM data. It sounds great, but is it really? When I point to the differences between data, information, knowledge, and wisdom; the wealth of ITSM data isn't as useful as it could be without additional "processing."
Some might say that we are lost in a sea of ITSM data, uncertain of both where we are and how best to navigate to where we want, and need, to be. It doesn't matter which way we look at the water, it's the aquatic, and data, version of "snow blindness." There's so much data that it's hard to see things clearly, or to use that data wisely.
The Quality of ITSM Reporting and Analytics is a Known Issue
An enterprise's ability to use the data stored - although some would say "trapped" - inside ITSM tools is a continuing issue for enterprise IT teams. A great proof point is the Service Desk Institute (SDI) report: "Life on the Service Desk in 2016," which highlights reporting and analytics as both a top frustration with, and top required innovation/improvement in, ITSM tools (via two separate survey questions).
In a third question, the "inability to easily produce metrics and reports" was voted as the thing that causes services desks the most pain. With 53% of respondents pointing the finger at reporting and analytics, placing it ahead of other service desk pain points including: outdated ITSM tools, struggling with knowledge management and self-service, and budget constraints. Unfortunately for enterprise IT teams, it was also top when SDI produced its 2013 report.
Why Does Traditional ITSM Reporting and Analytics Hurt So Much?
Part of this pain relates to the core ITSM technology - that reporting and analytics is seen as an "add-on" to the ITSM-process enablement (incident management, say), and an add-on where "good enough" is probably sufficient to get through the request for proposal (RFP) and purchasing processes. The data is there but customers can't use it in all the ways they would like to.
Another part is the ongoing requirement for manual effort, with it still not uncommon for reporting-pack creation to take days. And with this manual effort comes unwanted costs (possibly including data re-entry), the probability of human error, and the inability to see meaning in what can be large data sets. The ITSM tool might have a basic facility to list the top ten problem areas using incident categories but, beyond this, customers are reliant on the limitations of human comprehension, and trend and pattern recognition.
It's consequently a big opportunity for the application of artificial intelligence or, more specifically, machine learning and predictive analytics to ITSM operations.
Benefitting from Machine Learning and Predictive Analytics
Most of us will already know the benefits of automation, and machine learning offers a similar portfolio of efficiency and effectiveness advantages:
- Increased speed of execution
- Cost reductions
- Improved customer experience
- Reduced human error
- Increased task adaptability
- Reduced need for human intervention
Plus, with the specific application of machine learning to predictive analytics, there's the ability to manage, and get insight into, data sets that are too large for human analysis (and possibly even comprehension). Where predictive analytics can be described as:
"… encompassing a variety of statistical techniques from predictive modeling, machine learning, and data mining that analyze current and historical facts to make predictions about future or otherwise unknown events" ~ Wikipedia
Applying Predictive Analytics to ITSM Operations
There's already a number of opportunities where predictive analytics can be, and already is, used to support ITSM activities. These will grow as our thinking matures, including:
- More efficient IT support - speed is an important facet of modern-day support and customer service; customers tell us so. Predictive analytics can be used to speed up support transactions - both those that are service-desk-agent-led and those where end users employ the self-service channel (including the use of machine-learning-enabled knowledge management).
- Identifying and predicting issues - predictive analytics can identify common issues (what ITIL would call problems) or the signs of adverse things approaching, including in the assessment of change risk and capacity planning calculations. It can also offer up the most-likely resolutions for identified issues, using a recommendation engine, and schedule preventative maintenance.
- Predicting future IT service trends - this could be predicting the demand for new or existing IT services, or the future levels of IT support personnel required to maintain, or to improve, service. Predictive analytics can also be used to predict future levels of customer satisfaction based on service-level and operational trends, or proposed changes the status quo.
- Improved knowledge management - predictive analytics and machine learning offer a number of capabilities that significantly improve an organization's ability to exploit its knowledge. From intelligent search and recommendation engines, to automatically identifying and filling knowledge gaps. Chatbots and intelligent autoresponders are both high-value use-case scenarios of these search and recommendation capabilities - with the technology able to provide faster solutions with a potentially better customer experience than traditional human-to-human (H2H) support methods.
Thus, ITSM is a corporate capability that's ripe for predictive analytics enablement. Both in terms of people being empowered by the technology and/or the technology taking over manual tasks previously undertaken by people. With the result that: staff are more capable; ITSM activities - including issue resolution - are more efficient; staff and management has greater insight into services and performance; there's a better customer experience; and there's potential for operational costs to be reduced.