The learn as you go service desk - 2030 Week Day 5 - ITSM
16 Dec 2016
It's the final day of 2030 Week on AXELOS Community and I'd like to offer my sincere gratitute for taking part, but it's not over yet! You can still walk away with a £25 Amazon voucher, just read the scenario below and give us your thoughts! Here's today's scenario:
"Remember that scene in the Matrix where they upload Kung Fu skills into Neo’s brain. Well, we aren’t going to suggest going that far, but we’re going to get as close as we can.
Imagine a service desk where the analysts are serving hundreds of businesses, using just as many different pieces of software and infrastructure. It would be impossible to each analyst to have strong enough understanding of each business, their processes, software or service history. So instead they use real-time voice and text recognition to analyse each request as it reaches the service desk, which then searches for fixes, bugs and walkthroughs within a knowledge base and then provides each analyst with a script and/or check list which they can follow to resolve the support call.
This means that when a company in San Francisco calls about a problem they are having with some finance software, which the analyst has never even heard of before. They are presented with all the UI walkthroughs, FAQs and known errors they need to sound like they have used the software and fixed that problem 100 times before.
It really is a service desk that learns new skills on demand at the point of need, which both optimises training requirements and can provide thousands of end customers with the support they need."
What do you think, is this idea full of holes or could this really be the big managed service provider of the future? Let us know below.
I see many holes in this scenario. For one, though the real time voice recognition would translate the conversation into the languate of the analyst, dilects and slang may very well result in either the wrong information provided or the call being totally misunderstood. If the voice recognition moved passed the language barrier, the number of possible solutions could very well increase resolution time and possibily irrate many customers. There is always a possibility of something new the system has never heard of or of which there is no research available. Would the analyst have the ability to fix on their own, or will they be mind numbed which would result in an increase in resolution time and irrated customers? Too many holes in this solution for adoption as this point.
Why wait till 2030? This can be done today, All it needs is a good knowledge base and good tools to enable matching of incidents to that knowledge. Oh, and a methodology to guide the establishment of processes to support the effective use of the tools. Oh, and a culture where information sharing is encouraged, rewarded and ubiquitous. And a recognition that in our rapidly changing world the value of knowledge is greatest when it is fresh, and rapidly declines over time. And that any knowledge is better that no knowledge. If you wait until it is perfect it will probably not be available until it is no longer useful.
Actually there is a methodology that promotes these goals. It's called KCS, published by the Consortium for Service Innovation. It is complementary to ITIL but has some different views, e.g. they believe that Knowledge Articles (what ITIL calls KEs) should not wait until Problem Management has found the root cause and a workaround. If no match can be found to an existing article a draft article should be published as soon as possible, i.e. by Incident Management. Then as articles are reused they should be flagged for review, fixed, and approved based on the experience of all users.
Look them up at http://www.thekcsacademy.net/