The Definition of Insanity Revisited

A recent Pulse Report revealed that 62% of IT service management (ITSM) leaders are considering or actively pursuing AI projects to help reduce help desk call volume so they can divert CIO budget to strategic initiatives. That sounds great, but will they succeed? Not if they choose technology that is repackaging the same solution employees have already voted against.

ITSM tools have traditionally relied on portals and knowledge base articles to provide employees with answers to questions in the hope of deflecting calls from the service desk. However, portals deliver a fragmented service experience that is difficult to use, resulting in average employee adoption of 10 to 15% with less than 1% ticket deflection. In spite of that, some vendors have recently introduced chatbots that offer Google-like search engines to deliver fewer and better search results of knowledge articles via portals. Does that make sense to you?Einstein is widely quoted for saying, “the definition of insanity is doing the same thing over and over again–but expecting different results.” While improved search may sound like a great idea, most organizations do not have a great knowledge foundation, which means these chatbots are trying to do the same thing (i.e., searching on knowledge articles) and expecting different results (i.e., thinking that fewer articles will deliver better answers).

Knowledge articles have not worked for a number of important reasons:

• Most organizations do not have a great knowledge foundation, which means chatbots will only have responses to a small number of issues
• The help desk is too busy closing tickets, so knowledge articles will never be a priority, which means many will be out of date
• Knowledge articles are written by help desk agents in their technical vocabulary, so when employees ask questions in their normal language, answers won’t be found
• When an answer is not known, employees need to go somewhere else to interact with the help desk, teaching them to go straight to them via email or phone next time
• Teaching the chatbot new answers requires creating more knowledge articles, which the team has no time to do

History has proven that employees are not willing to sift through knowledge articles and submit help desk tickets through portals when they have Alexa-like ease of use and immediacy for answers at home.  If you are one of the 62% of IT leaders actively pursuing AI in the form of a chatbot to reduce help desk call volume, make sure the chatbot you choose is not trying to improve on technology that has already proven ineffective, or you might be flirting with insanity. Look for an innovative approach.