As a society, we’ve been fascinated with robots since the mid-1950s. My favorite was Rosie from The Jetsons—I’d love a robot maid! While we are still some ways off from having animatronic co-workers and household help, we do have software robots that—through robotic process automation (RPA)—can help improve the performance of today’s contact center agent.
RPA comes in a couple of formats:
Traditional RPA (Unattended Automation)—RPA can automate repetitive, rules-based tasks or processes, such as cutting and pasting from one system to another, or navigating between screens in a process flow. It can handle these tasks independently and automatically, freeing contact center agents for higher-value activities, such as engaging customers. RPA can also automate these tasks for other channels, such as website inquiries and online chat.
Attended Automation—This form of RPA sits on the employee desktop and offers guidance on next best actions based on real-time, contextual application triggers. Employees can select a “show me” mode for the attended robot to walk them through the process, or a “do it for me” mode in which the robot executes the next step or series of steps.
Hybrid Automation—By combining unattended and attended RPA, you create a synergy between the software robots and agents, enabling them to pass work between each other, reducing overall handle time. Agents can work on the next steps in parallel, and then pick up the task/process once the robot has completed its steps.
By combining unattended and attended RPA, you create a synergy between the software robots and agents.
Let’s look at how this works.
Scenario 1: Before RPA—The agent greets the customer and verifies her identity, issue or request type. He either puts the customer on hold while he captures the relevant data from multiple systems, or explains what he’s doing as he navigates systems. Once the data is populated on his screen, he reconnects with the customer if on hold, answers her question or confirms her order request, and wraps the call. He then performs the after-call work to update the systems and send the order for processing. A typical call could last between 4 to 9 minutes.
Scenario 2: After RPA—Now, let’s apply RPA. The agent still greets the customer, verifies her identity and identifies the issue. Simply initiating the process launches RPA, which prepopulates applications based on the call type. RPA pops up a script with suggested answers to the customer’s questions or a next best action based on the information gathered by the robot. Perhaps the customer is inquiring about a new credit card. The script gives the agent the info he needs to respond. If the customer confirms her desire to open the card, the agent can either have the attended robot show him the steps or complete the process automatically. The agent is now free to focus on the customer and determine if there are any other services that might benefit her. He can wrap the call, and RPA can either fully (unattended) or partially (hybrid) execute the follow-up work, such as initiating the process for card creation.
The call might take 3-1/2 to 5 minutes, and the agent would be able to immediately jump on the next call vs. spending time in applications completing the transaction. This can save an additional 2 minutes, for an overall savings of up to 4 minutes per call. Multiply this by the number of agents and calls per week, and you can quickly see the opportunity for potential savings and faster, more accurate service.
Examples: RPA and Seasonal Help
One of the largest telecom providers in the APAC region was launching a new service and anticipated hiring 15 new FTE to meet expected demand. Instead, the telecom deployed RPA unattended to automate order processing tasks for the simple service requests, such as quoting, billing and reporting, freeing up the seasonal employees to handle the inquiry calls for its new offering. The telecom was able to reduce turnaround time on these requests by 70%, from 12-15 minutes to 4-6 minutes.
A healthcare insurer needed to hire and train additional workers to meet peak demand during the upcoming open enrollment period. Given the complexity of plans, data and processes involved, training was lengthy. While some tasks could be automated with RPA unattended, most of its work is complex and requires creative thinking. The insurer leveraged RPA attended automation to shorten the time required for agents to train and become proficient. RPA attended automation provides just-in-time, on-the-job guidance with desktop prompts that:
- Remind employees of the next step or questions to ask in a process.
- Flag fields that are overlooked or improperly completed.
- Prevent employees from proceeding with actions that are not compliant.
- Perform the work for employees in some cases, such as copying and pasting data from one system to another.
The insurer was able to reduce classroom training by two days and time to proficiency by one week improving accuracy and turnaround times. The new-hires who leveraged attended automation had a noticeable improvement in first-call resolution, along with CSAT scores about 5% higher than previous classes.
So while the seat next to you might not be filled with a robot anytime soon, RPA can be that helping hand in the contact center to eliminate tedious, time-consuming tasks, reduce hold time and after-call work, and help agents and seasonal workers onboard and get up to speed quickly. I’m still holding out for my own personal robot maid, but for now I’ll be happy with RPA as my co-worker.