The Impact of Automation on Metrics

The original “help desk” was usually one person whose job it was to take notes about computer issues from users […]

Category: Article

The original “help desk” was usually one person whose job it was to take notes about computer issues from users or programmers (mainframe in those days) and pass the information along to the mainframe company. The mainframe company would send an engineer to address the issue. The sole task of the help desk was to note and report the issues.As personal computers became part of working life, the help desk started to do more. Most were still “catch and dispatch” desks, again capturing information and reporting it to the people who would go to the deskside and affect a fix.

Time went by, computers became more and more important, and the desk started attempting to resolve as many issues as possible without passing information to other groups. More recently, we’ve measured support effectiveness and efficiency by first level resolution (FLR), first contact resolution (FCR), time to resolve (TTR) and as many other metrics as any of us can imagine.

About a decade ago, we started talking about the shift-left strategy, bringing more complex work to level 1, thus cutting the number of escalations, the time the customer was left waiting, and the cost of support. Part of the shift-left strategy was also to move the simpler and more repetitive issues into a user-facing knowledge base so that end users could resolve as many issues as possible by themselves. As the simpler fixes moved out into self-help, level 1 was left with more complex issues.

At least for a period of adjustment if not permanently, the support center’s metrics would be changed. Because of increased complexity, both FCR and FLR would decrease, and metrics such as talk time, handle time, and TTR would increase. Since the support center had long considered these key metrics and focused on improving them, this was all combining to make support “look bad.”

Artificial Intelligence, machine learning, bots, and other automation tools are poised to make a large impact on support. The prime targets for automation are the very things that produce high FCR and FLR, keep talk and handle times down andtraditionally speakingmake support “look good.”

Your basic support (level 1 if you are tiered) is about to get squeezed between the work shifting left from escalation groups and the increased level of difficulty in resolving incidents and requests coming in from customers because the easy ones are being handled without human intervention. With machine learning in play, the degree of difficulty will continue to rise, because automated systems will learn how to resolve more and more issues over time.

The result can be (and should be) that what makes support “look good” is truly helping the whole organization move forward, not the accumulation of sets of numbers in a report each week or month that virtually no one cares to read.

With the “easy-peasy” stuff gone from the daily workflow, new vistas can open up, including:

  • More strategic thinking and action
  • Streamlined processes to provide better end user experience
  • Improved communication on your team and with other teams
  • Collaboration to improve performance across the board
  • More planning time for things like business continuity and disaster recovery
  • More project involvement

You know; all those things you’ve been wanting to do but didn’t have time for.

Yes, the skills needed will be elevated or changed, as mentioned in this post. Yes, the work will be more complex. But that’s where your brains are needed. This is going to be an exciting time to be in the business of providing and supporting the technology that makes organizations run.

Recent Posts

Quick Links

Refund Reason