Kathy Gibson reports from SMEXA 2015 – Using big data can help IT service desks save time and money while improving the service they offer their users.
According to Kevin Leslie, SPM director at HP Software EMEA, the concept of bimodal IT has fundamentally changed the nature of services.
“In the past, it was all about ITIL, ad we thought that companies could compete better if they implemented ITIL.
“Now, it’s possible that ITIL has had its day. In the past, there were users and there was a service desk. Nowadays the service desk might not even be involved in many service queries. A lot of assumptions are breaking down.
“ITIL is a framework, but it’s not necessarily the issue when you are implementing an IT service desk.”
Using big data in a service desk is a new concept that can offer a good return on investment that can save hours for people on the service desk while improving the mean time to repair (MTTR). “Users want their issues fixed quickly, so if you can reduce it that’s good.”
Customer satisfaction can also be improved, Leslie adds.
The traditional IT service desk has structured data in the middle, but it’s missing machine data and human information. Big data, on the other hand, uses the free text from human input as well.
“Your systems can read what you’re typing, interest it, and guide your use of the systems based on that.”
Machine data, on the other hand, could be used to interpret images or error messages.
Charl Joubert, change and configuration manager at the University of Pretoria, points out that all organisations sit on a gold mine of data in their service desks. “We needed to understand what our users want,” he says.
UP has about 62 000 full time students, 25 000 distance students, 33 000 certificate course students, 11 campus companies with 600 staff members and about 7 500 UP staff. This is all supported by an IT staff of 284 people, nine of who are IT service desk agents.
“We couldn’t do the traditional telephone or email channel for service desk,” he says. “Self-service was really the only way to go.”
UP is on its fifth version of the IT service catalogue now, and Joubert says it’s improving all the time.
“But you get all this information from users and we don’t do much with it to improve the service we provide.”
UP uses the HP Service Management 9.40 with smart analytics, HP uCMDB 10.20 critical services an HP Adoption Readiness Too integrated to Blackboard LMS.
HP Service Manager Smart Ticket helps to capture information from the user. “We stopped the email channel to the service desk. It was counter-productive for staff and IT staff because the information is not great. Its open to interpretation and contact starts becoming such as hassle that you start irritating the end user.”
Now, it’s easy for self-service users to submit requests, and the backend predicts the categorisation – to an accuracy of 76%. This saves the agent 60% of the time to categorise interactions, taking away human interpretation and errors.
For users, there is a service portal and also a mobility portal; and for IT, about 1 250 man hours have been saved year to date.
The mobility portal will be introduced to students in 2016, and Joubert expects uptake to improve dramatically.
Hot Topic Analytics indexes incident data, helping UP to overcome what used to be an immature problem management process. “Out of the box it supports Afrikaans, and about 40% of our interactions are in still in Afrikaans,” Joubert points out.
“We saw Hot Topic Analytics as a problem management game changer, and could immediately see where our big problem areas are, allowing us to visualise them to the management team.”
Diverse teams can also spot patterns and thus start to solve them.
UP handles about 3 500 incidents each month, taking an average time per incident of 4,39 hours. It has seen a 6% reduction in incidents, 210 fewer incidents per month and 921 man hours.
“That is impressive,” Joubert says. “And it gives us the freedom to become more proactive and create better services for our end users.”