Ventana Research Analyst Perspectives

When artificial intelligence emerged from the labs and vendors started offering it as a component of their software, many contact-center buyers shied away from it. From their point of view, AI and machine learning tools were new, expensive, relatively untested and had an uncertain use case. This stance was understandable, as contact center professionals are traditionally expected to be risk-averse when deploying technology into their operations. Contact centers are, by design, supposed to be hardened, mission-critical sites of high reliability. There has historically been a bias towards avoiding new technology, deploying only when it has been thoroughly vetted across the industry.

AI and machine learning have crossed that threshold. AI tools are embedded in most vendors’ portfolios and are deployed in applications that span the customer experience and lifecycle, not just in contact centers. There is still skepticism, naturally, and many potential buyers do not fully understand or appreciate the benefits to be gained by implementing AI tools in their centers. However, Ventana Research asserts that by 2024, 3 in 5 organizations will have turned to real-time analytics to provide guidance to contact center agents during customer interactions, resulting in better customer satisfaction scores. We also believe it will help reduce costs and relieve agents of much of the mundane work that leads to burnout and turnover.

The first and most widespread application for AI in customer service was in chatbots and Intelligent virtual assistants. When automated conversational bots first rolled out, vendors needed to distinguish their capabilities from that of traditional interactive voice response, which uses a tree-like menu structure to guide callers to a desired endpoint. That endpoint could be the extraction of some relevant information (like an account balance or shipping status) based on a data dip, but most likely an IVR would terminate by routing a caller to an appropriate agent based on caller input.

Chatbots had to offer more conversational characteristics to justify the expense and complexity of deployment compared to earlier self-service tools. Hence, AI and machine learning — combined with natural language processing — were called upon to make customer interactions with bots more lifelike and productive. For many contact center professionals, this was their first exposure to the capabilities of AI/ML systems.

Vendors continue to develop AI tools, often as a platform service enabling existing applications to work better (or differently). Multiple AI platforms are available from major vendors — ranging from IBM, Salesforce and Google — to more targeted contact center developers like NICE, Verint and ServiceNow, plus small firms that leverage some of the larger vendors’ platforms. AI is available everywhere across the industry, so buyers can usually find it in the software stack of multiple vendors. It has progressed far enough that it is not a “shiny object” distraction, but buyers do need to be continually educated and updated on how it can be used.

The value of AI in contact centers is that it opens the door to analyzing data sources that were not commonly explored for CX purposes. Unstructured data like voice recordings provide valuable insights that can immediately be acted upon, but only if the data is easily accessible and can be manipulated by non-technical line-of-business users. Without automated systems that incorporate AI/ML, much of the useful data is simply too complex and difficult for most operations teams to access.

It has taken only a few years for vendors to move this technology from the developmental stage to where it is deeply embedded within the tools that contact centers and other teams use every day. It is now a component of applications used in many disciplines, including customer engagement platforms, tools for agent management, call routing, forecasting and scheduling, and knowledge management. The advantage of having AI/ML embedded so deeply is that it effectively disappears into the stack — users need to know almost nothing about the mechanics of how ML works, as it has become a nearly seamless part of established reporting and analytics applications.

As a result, buyers seem interested in deploying tools that incorporate AI, and less likely to see it as technology for its own sake. AI and ML have opened up a new avenue of approach for operations teams to link up with one another across departments. The current emphasis on building automated workflows and processes can only succeed in an environment with more shared access to information and an easier way to sift through information. Automated systems that incorporate AI/ML can enable all of the contact center data to add value to customer and agent experiences.

In the second part of this report we detail the most effective AI use cases developed so far. More insights on how call centers are using advanced technology can be found in these Analyst Perspectives:

• Field Service Transformation for CX Differentiation
• Customer Service and Support: Expanded Role and Need for Software
• The Voice of the Customer Is Really a Chorus of Voices
• Intelligent Virtual Agents Are an Imperative for Digital Self-Service

Regards,

Keith Dawson