Process Improvement, AI, Robotics and the Human Touch
At the heart of every Shared Service Centre (SSC) or BPO (Business Process Outsourcing) operation is continuous improvement. How can we deliver this process at a higher quality with a lower cost? There are dozens of techniques from 6 Sigma, Lean, and Agile to name a few that the process improvement engineers, which exist in each centre, use to achieve the yearly cost reduction targets.
AI and robotics increasingly are being added as another tool in the kitbag for the process engineers to use to hit their objectives. Will all these different types of automation change the focus of process improvement and the role of the process engineers?
I would argue yes.
Historically, a process improvement engineer would concentrate on removing root causes to quality issues or look at the detailed steps an individual takes to complete a task and see if they could re-engineer the steps to be more efficient. There are different processes and techniques but they pretty much all start from the point of view, “Let’s review how the work gets done and determine what steps, techniques, QA activities or technologies can we add or changed to make the process more efficient”. If the agreed solution was to use robotics or AI, then I would argue that the first step the process engineers would focus on is, "Where are the best applications of AI and robotics?" – and, therefore, "What changes need to be made to the process to allow AI and robotics to increase quality and reduce cost?"
In other words: Identifying the application of the technology first, and then making changes to the process to allow the technology to work versus looking at the process and identifying ways to change the process to improve.
Let’s use HR as an example. Many organisations have an employee service centre. A place where employees can call, mail or chat to an HR service representative to get answers to their HR queries. Questions such as "How do I do …?" or HR issues ("I did not get paid correctly") or seeking advice about a policy or procedure ("If I am going on maternity leave what is the process I need to follow?"). As a simple example of the two approaches for how to improve the HR process of answering employee queries here are two options:
Chatbots are proven technology. So, how can we use chatbots to solve x% of the queries coming in via mail or chat? The process team would start with what questions can be answered by chat bots. The repetitive question that HR service centres get would be easy to do (queries such as, “When do I get my P11D’s?” or “When does benefit open enrollment start?”) These are highly repetitive questions during specific times of year that cause contact centre volume to spike. Chatbots could be set up to address these seasonal, repetitive queries.
In approaching the design of the work activities within the employee service centre, with AI or robotics as the first solution, how would the process engineer determine what are repetitive questions? What are the triggers for these issues? How many levels of questioning would these queries usually go (no good putting in a chatbot to read out a data point, if the person has a second follow-up question, like, “What happens if I am on holiday?” – and then you are back to a person answering the query). The process team would identify the triggers and the responses, so they could be automated using chatbots or other technologies.
But: What would this look like if you took a more traditional process improvement approach?
Keeping the same service example in mind, the approach of the process improvement team would be more focused on eliminating the calls into the centre. If they could not remove the contacts, how could they ensure that the service centre agent had quick access to a knowledge database to answer the query or, in the case of repetitive queries, the content library to “cut and paste” the response back to the employee? The starting point would be on how to eliminate the work; if the work cannot be removed, then how to make the steps to complete the task as few and as efficient as possible.
The above is a simple example, but hopefully proves the point that if you start with the solution being AI or robotics activities to drive improvements, then the approach will be different. AI and robotics are not necessarily focused on eliminating work – they take the work and concentrate on eliminating the need for humans.
The next time you look at process improvements within your SSC, or you talk to your outsourcing provider about what they are doing to enhance the operations, take the time to reflect on the starting point. Should you consider robotics, or more traditional methods for process improvements? Or is the human touch still the best answer?