As chatbots infiltrate the workplace, start with the IT service desk

In the coming year, millennials and the Gen Z generation will account for almost 60% of the global workforce. And if there’s one thing we know about this cohort, it’s that they like to communicate via technology, whether through text, collaboration portals, videoconferencing or some other digital channel. So why are we seeing so few chatbots being used on the IT service desk? While it’s true there are challenges to overcome, the ultimate benefits outweigh the effort involved. In a recent study, users reported an overall positive experience with using a chatbot, especially when it meant avoiding queuing up for human help.

Some answers emerged in our recent work with an energy and utility company that sought to improve the end-user experience and agent productivity by empowering teams with AI and automation tools. The company wanted to implement an AI chatbot solution for Level 0/Level 1 IT support to address  service requests, FAQs and status checks on trouble tickets.

A chatbot-first approach to IT service

We worked with the utility to implement a chatbot solution using our WorkNEXT platform. The chatbot serves as the first point of contact for users. Trained to understand language specific to the enterprise and armed with an existing knowledge base, the bot helps tackle frequently asked questions. Once it was integrated with the enterprise IT service management (ITSM) tool and a robotic process automation (RPA) system, the bot was able to address user queries on service requests, ticket creation, ticket update and ticket closure. When the bot can’t fully remedy the issue, it transfers the call to a human agent.

Over time, the bot has helped the business accelerate problem resolution time and meet its key performance indicators, including an average speed to answer of less than 10 seconds, a 60% increase in end-user satisfaction, a 36% deflection of emails to the chatbot in the first 60 days, and a 40% chatbot resolution rate. By making the AI chatbot the first point of contact for users, we ultimately expect it to scale up from handling 10% of the utility’s ticket volume to 70%, thereby improving the operational efficiency of the enterprise.

Overcoming challenges

To enable these benefits, the business needed to overcome some key challenges that are common to any company seeking a service desk chatbot deployment. These included:

• Resistance to change. Many office workers are accustomed to interacting with one another on a daily basis to resolve important issues. The idea of inserting a chatbot into the problem resolution process can seem unfamiliar and even unnecessary for these workers, particularly when the benefits aren’t communicated. This is why it was imperative to conduct aggressive adoption and change management measures to educate end users on the benefits of this new support channel and guide them toward it. Simple measures included mailers, posters, banners, gamification and roadshows to highlight the benefits and seek feedback. These were followed by more sophisticated measures, including the use of cognitive- and user sentiment-based responses to represent human emotions. This automatically improved adoption. Further, it’s essential to start with use cases that are simple and are sure to give an accurate response to the end user’s query, thus encouraging employees to trust the chatbot’s capabilities. Over time, additional use cases/scenarios can be added in an ascending order of complexity. It’s ideal to follow a three-stage onboarding process, where stage 1 consists of basic use cases (i.e., ticket management, FAQ, simple search responses from the knowledge base), stage 2 covers service requests, and stage 3 includes incident resolution.

• Cost perception and uncertain ROI. The cost of building and implementing an AI chatbot can seem high, particularly to an IT support organization. Licensing and ongoing support can range from $60,000 to $300,000 annually, depending on the size of the business. The more use cases, the higher the implementation cost. Businesses need to revamp how they measure the returns on that investment. Near-term benefits – such as reduced wait times and abandon rates – translate into quality of service rather than monetary value. Over time, however, labor costs and ticket volumes will decline, bringing about a reduction in operational costs for the organization.
• Complex technology to comprehend. Organizations that have not yet interacted with AI can find chatbots complex to understand, at least initially. It’s important to understand that the virtual assistant (VA) is separate from the AI chatbot platform. The VA serves as a front end with which users interact, whether asking it to send an email, set up a meeting or even ask questions via verbal commands. The VA is powered by the AI platform at the back end, which enables it to understand human speech and convert it into executable commands through natural language processing (NLP). Natural language understanding (NLU) and machine learning (ML) are the technologies that form the basis of the AI platform. Given conversational AI’s still emerging status in mainstream business, there are still many different approaches to how bots consume and interpret knowledge. Organizations’ needs will vary based on the line of business and the type and maturity of the knowledge content. Further, some organizations deploy multiple chatbot platforms to take advantage of their varying functionalities and capabilities, such as which ITSM tools they integrate with, the channels through which the bot can be reached and their NLP capability. Because no single chatbot can be implemented universally across industries, the choice of bots will depend on the specific functionalities required. However, many organizations lack sufficient knowledge or experience to make this choice on their own.

The mainstreaming of chatbots

In the consumer world, the use of chatbots for service interactions is growing every day. As these encounters continue, chatbots will also find their way into the workplace. According to Gartner, by 2022, 70% of white-collar workers will interact with a conversational platform on a daily basis.

By overlaying the IT service desk with chatbot capabilities, businesses can reap the advantages of efficiency and a modernized experience while gaining experience with this advanced technology.

Siddarth Oruganti and Vijayta Dhingra from the AI-powered Service Desk practice in Cognizant's Digital Workplace Services contributed to this blog.