Broadening Service Desk Automation with AI

By Pat Calhoun, Chief Executive Officer
March 8, 2022

In 2016 I decided to leave my role as SVP of product at ServiceNow because my vision of the future was very different from where ServiceNow was headed. I started my career on the service desk answering phone calls from displeased users daily. Working on the service desk is hard, and that’s one of the reasons why the attrition rate of service desk agents is 41%. Few organizations can point to a great employee experience when getting help from the service desk.

I joined ServiceNow thinking I would have the opportunity to transform how employees experienced getting help. However, I came to realize that the focus of the company was primarily on digitizing customers’ back-office processes. Sure, we had built a self-service portal where employees could read knowledge articles at their leisure to hopefully find answers to their issues. Unfortunately, the poor user experience of portals has led to less than 12% adoption across all customers.

It is interesting to consider that here we are in 2022, and from an employee’s perspective, the service desk has not fundamentally changed since the 1990s. We just sponsored research that revealed that 80% of remote employees are still calling and 73% are still emailing the service desk for help. All the frustrations that existed in the 1990s, exist today, compounded by the fact that today’s hybrid workforce can no longer shoulder tap their neighbors with questions – leading to an overall increase of 35% calls into the service desk.

Moving From Deflection to Automation

When we first founded Espressive, our original idea was to simplify how employees get answers to their questions or issues. We embarked on a massive undertaking to build a platform that could understand the human language, and then tie that understanding to an immediate, personalized response. It was very rewarding to see organizations like Dexcom achieve over 60% deflection rate with Barista, our virtual agent, enabling them to shift their service agents from dealing with repetitive questions to becoming automation experts.

What constitutes automation seems to be up for debate, with some vendors claiming that providing a knowledge article to an employee is automation. (🎶 Insert Jeopardy music here. 🎶) In fact, some vendors go so far as to claim that proactively reaching out to an employee to ask about their sentiment is automation– so for them any interaction is counted as an automation. At Espressive, we call that employee engagement.

But we do not believe it is up for debate – the definition of automation is clear to us, and we soon saw the opportunity to expand the experience to shift away from simply answering questions via knowledge (i.e., deflection) to delivering end-to-end automated resolution. For example, if an employee wants a password reset, having them read an article and complete the steps is much less powerful than simply completing the steps for them. Yes, if the employee were to consider the instructions clear, then a ticket would be avoided. But we consider this to be deflection, not automation.

This led the team to expand on the Barista integration framework to enable our customer success team, our partners, and our customers to build their own integrations. Our integration framework has led us today to supporting thousands of integrations.

Some of these integrations gave us opportunities to automate tasks that have historically been handled by Tier 2 teams, such as identity management and desktop support. Today, Espressive customers enjoy a fully automated identity and account management process, eliminating manual tasks that would have historically leveraged more expensive resources.

Going Beyond Tier 1 to Automating Desk Side Support

As I look at the future, the next big opportunity for Espressive is focusing on automating tasks that have historically required desk side support. When employees experience challenges such as laptops that are running slowly, this has required an expert from the desktop support team to access the device physically or remotely to identify and resolve the issue. I believe this is where the future of service desk automation is headed – providing a way to both troubleshoot and resolve these issues without human intervention.

I also believe the future is these issues should be identified before employees even report them. Imagine a world where a virtual agent notifies an employee that their device appears to be operating slowly and asks for confirmation so that it can troubleshoot and resolve the issue. We call this “proactive support,” and we believe this is where platforms such as Espressive are headed.

And it will be awesome.

Share this post: