We are living through interesting times. The march of the bots continues, and businesses are reaping benefits through lower-cost customer interactions. This is having a fundamental impact on the role of the contact center and how the humans within are used. Business models will need to change, and IT platforms once considered too expensive will be deployed, allowing further improvements in customer satisfaction and service.
As the simple processes are effectively automated-away through intelligent bots, the processes that remain can be characterized into one of three types:
Ungoverned Processes
For those who have embraced chatbots and robotic process automation, it is well understood that those processes that are highly repeatable and relatively static represent the low-hanging fruit from which cost savings can be easily derived. Unfortunately, things are not always what they seem. I met the chief information officer of a large insurer recently. Its back-office was, so I thought, a perfect example of highly repeatable processes that could be modelled programmatically. Indeed, the steps required to validate a request for insurance, confirm the terms, check the history, and update broker systems were reasonably simple and sequential. The problem was that so many of the IT systems were ungoverned and unstructured. Requests in free-flowing, poorly formatted word documents arrived by email and then required highly skilled parsing by humans to identify the relevant pieces of information. As these items were gleaned and fixed, the appropriate set of destination systems were populated. Unfortunately, these were provided by third parties who seemed to take a cruel delight in making changes to their systems on a regular basis without telling anyone! Thankfully, the CIO had the foresight to understand that not all repeatable processes are also programmatic.
Complex Processes
Consider, the process of a customer moving from the perspective of a telco operator. This is the classic use-case in which any attempt to automate anything beyond the simple case will end up in a world of pain. I remember dealing with a large quad-play operator in which the Home Movers IT program was seen as a poison chalice only suitable for those with sufficient experience to be able to scope the work. It soon became apparent that a significant proportion of use-cases do not follow the happy path and that the company needed to deal with multiple date changes, non-responsive suppliers, broken equipment, wayleaves, incompatible tariffs, geography-specific products, provisioning failures, product incompatibilities, etc. In theory, every one of these scenarios could be modelled, but this would require so much effort and ongoing change that it would be too expensive to build and fall short of what a suitably trained contact center advisor could achieve using her natural intelligence and common sense.
Critical Processes
Our largest customer, JP Morgan Asset Management, uses highly qualified financial advisors to deal with ultra-high net-worth individuals and institutions. It offers advice and market insight that is gleaned from more than 20 systems and considerable personal experience. It will also need to explain, human to human, the operation of complex financial instruments and make sure that clients understand the risks and benefits of proceeding. While certain intermediary steps can be automated, the nature of the process is one in which the advisor's route through systems cannot be known ahead of time and that new applications or analysis might be required depending on prevailing market conditions.
These three types of human-centric processes are commonly referred to as exploratory processes and are the focus of what the financial services industry has been calling interop for more than a decade. The very latest platforms are designed to weave desktop applications, including their user interfaces, into configurable workspaces that give humans the freedom to explore the data without constantly searching, copying/pasting, and updating systems. Perhaps more importantly, they are also designed to deal with hand-offs from chatbots in which the conversational context that has been gleaned becomes instantly available and is used to launch the correct applications with the correct data context.
The idea, is therefore, not to integrate too far, but to allow humans to quickly uncover data and make insights without becoming the means of application integration. This is the natural evolution of the role of the contact center and one that enables humans to focus on tasks that machines are not yet capable of performing.
James Wooster is chief operations officer at Glue42. He has an extensive background in software infrastructure solutions and has spent his career in enterprise sales and field operations. Prior to joining Glue42, he was vice president of U.K. and Ireland infrastructure portfolio sales at SAP, vice president of of Software AG's Australian business, and UK/Ireland managing director of TIBCO.