Unsupervised Improves Internal Helpdesk Bot Experience for Fortune 100 Company

Download Case Study

The Problem

A major professional services company with more than 100,000 employees sought to reduce volume and increase FCR for their internal help desk call center. In addition to the traditional call centers, the firm had invested in chat bots to resolve issues faster.

However, their chatbot and call routing solutions had an initial configuration, but quickly grew out of date as employee needs changed overtime.

Wanting to better utilize chatbots and call center resources, this large enterprise invested in Unsupervised to gain a better understanding of their helpdesk and improve employee experience with it.

Transform Insights to Actions

Unsupervised was able to find insights to provide a more thorough understanding of what was happening across the global contact centers and chatbots in days. With previously unknown insights surfaced from our AI, they quickly identified areas to improve their call center experience and reduce costs

While new insights emerge every week as their data refreshes, a few critical areas of improvement emerged immediately:

1) Tenured Agents Handle Priority Cases Poorly - This insight was driven by the realization through Unsupervised patterns that high priority cases had a lower FCR when handled by tenured reps. Previously, high priority cases were routed to the most experienced available agent, but because these cases are so infrequent, they actually were best served by newer agents who had most recently been trained. Changing how they routed these P1 - P3 issues saw a significant boost to FCR in coming months.

2) Login Requests Are Poorly Reported - Through patterns they realized more tickets could be classified as login issues from the case description rather than the type field in ServiceNow. Realizing this, they were able to change their call routing and chatbot intents to look for login issues within the description and route accordingly.

3) New Opportunities to Leverage International Resources - It had been assumed that their international call centers handled issues worse than those in the US. Through Unsupervised, they realized that their Tier 1 international resources actually had a higher than average FCR for all items handled via chat and email. This enabled them to design more ways to pass tickets on through these mechanisms to the international teams to reduce cost and increase FCR.

Unsupervised continues to help this large firm find unknown insights buried within their complex, unstructured data. They now are able to tune their call center experience and chatbots to reflect the needs of their employees today rather than that of months ago.