3 Ways to Manage a Robotic (RPA) and Human Workforce

Recently Craig Seebach, VP of Workforce Engagement Strategy at Verint, presented at RPA Live 2020 on “Measure and Manage Your RPA Workforce Seamlessly with Your Employees.” His three top takeaways for managing a combined robotic and human workforce were:

Retain tribal knowledge of how tasks were performed prior to the automation
Set appropriate performance goals for your RPA robotic workforce and employees
Focus on your automation’s end-to end impact on your customer.

Retain Tribal Knowledge

When implementing Robotic Process Automation (RPA), the goal is to improve efficiencies, speed and accuracy. Once programmed or scripted, unattended RPA automations can perform rules-based tasks, consistently, accurately and faster than humans—until they don’t.

RPA, like any software, can break. If there’s an automated software update (e.g., Microsoft), a UI change in an application, or a change in a database schema; the robot will stop working. Or there’s a spike in a particular task and your robots are maxed out. What do you do now?

Organizations need to keep human employees performing a certain percentage of an automated task to ensure they retain tribal knowledge of how it’s done. Never hand off 100% of a transaction to RPA. Keep employees doing a percentage of the work to ensure you can triage when there’s a problem or handle work overflow to ensure work is completed—and your customer service goals are met.

Set Appropriate Expectations for Robotic and Human Workers

We’ve all heard the adage, “you can’t manage what you don’t measure.” Most RPA tools have some form of operational reporting or health checks for their automations. However, to fully capture the benefits of RPA, you need to:

Orchestrate the flow of work between RPA robots and your employees; and
Measure the impact of your automations on speed, capacity and customer experience.

Orchestrate the Flow of Work between RPA Robots and Employees

In most organizations, simple process flows have been automated with straight-through processing in industry or functional-specific business process management (BPM) systems. What typically remains? Complex, multi-touch, multi-step processes that have exceptions or potential variations in how the work gets automated. For these processes, some subtasks can be automated—however, there are points where human intervention or decision-making needs to take place.

So how do you efficiently move work between the RPA robots and human employees? How do you make sure you are not holding up the human or robot from performing the next step in the process? How do you factor in your robotic workforce when creating your capacity plans?

We have a solution called Verint Work Manager that was designed specifically for back-office operations. The solution captures work from multiple systems, prioritizes the work, and provides real-time and historical business intelligence so organizations can understand how well work is being executed—and if efficiency and service goals are being met. Your robotic workforce simply becomes another team or employee category in the solution. It allows you to manage, in a single interface, the workloads of both your RPA robots and employees—as well as the movement of work between the groups.

Work Manager dashboard showing combined RPA robot and human team.

Measure and Manage the Combined Workforce

Just like you measure the performance of your employees to see if they are reaching their goals, so too should you measure the performance of your RPA robots. And while they may have some similar performance metrics, e.g., # of items processed and % of time in production, the expectations or goals will be different.

Let’s take number of items processed. It may only take a robot 20 seconds to perform a task and the employee 60 seconds, so their goal for handle time or number of items processed will be different. The ability to set fair and appropriate goals for humans and robots will also highlight and document where you are capturing the efficiency and speed benefits of RPA.

Let’s take percent of time in production. Employees typically perform a myriad of tasks, and they take breaks, have one-to-one meetings with their bosses, participate in meetings and training, etc. They are not expected to spend 100% of their time in production.

Likewise, a robot could potentially spend 100% of its time in production if there is enough volume of the work tasks they are programmed to do. If they are waiting on work to come from, say, another department, then the robot will have downtime too. The goals for that downtime or shrinkage would be different for human vs. robotic workers.

Other metrics that might be different for your robotic workforce would be metrics that focus on the health/operability of the robot, such as # of errors and capacity utilized. Below is an example of a scorecard for a team of 5 robots. You’ll see robot #5 has dropped from 80% to 90% productive to 60%.

Upon closer examination, this robot is incurring a high rate of errors and incomplete transactions. You can dig further to understand which transaction types are failing and if there was a change in an application—or if you need to rewrite the script for that task type. It gives you insight with greater detail than you might otherwise have.

Focus on Automation’s End-to End Impact on Your Customer

You hear a lot in the RPA marketplace today about the lack of scalability. The piece that is missing is the ability to identify potential automation tasks that span across functional groups within an organization. Most automation ideas are generated by an individual or team doing a particular piece of a transaction. The recommendation is made in isolation, without understanding downstream or upstream impact of that automation on the customer or the organization’s broader operational efficiency goals.

Or organizations may rely on their Six Sigma and Business Process Improvement experts to ferret out the automation opportunities in their organization. This is a very expensive and time-consuming endeavor at the beginning of your RPA/automation journey that can postpone capturing the ROI of RPA for months.

So how can we automate this pre-work?

Verint Robotic Process Automation Discovery is the latest addition to our RPA suite of solutions. Automation Discovery “bots” are placed on employee desktops and watch what everyone on a team or department is doing. It captures which applications they are going into, how they navigate them, where they click. Using sophisticated algorithms and AI, the solution identifies commonality between employees and creates proposed process maps for the related tasks.

The maps can be validated and refined by the actual processors or SMEs. Perhaps the starting point is two steps earlier than what Automation Discovery suggests. You can quickly modify and then Automation Discovery will adjust the process maps and related recommendations. Notably, the solution provides a wealth of data to help make the decisions on what to automate easier, such as how many:

Process variations there were
Times the process was executed
People performed the process
Steps and handoffs there were.

The solution automatically presents a prioritized list of automation opportunities for the processes observed. In addition, these process maps can be ported directly into the Verint RPA solution, automating the creation of the scripts associated with the automation opportunities.

Automation Discovery is a powerful tool and a more scalable way to identify automation opportunities across departments and the entire customer journey.

Want to learn more about Verint Automation Discovery and how to measure and manage your robotic workforce? Contact us at info@verint.com.

Want to learn more about RPA and automation in back-office operations? Join the webinar on July 15 at 10 a.m. ET, where Kryon System’s Daniel Peled will interview Craig Seebach on “How Work-from-home Demands Are Accelerating Automation in the Back Office.”

Blog: How to Scale RPA Beyond a Pilot

Blog: Robotic Process Automation: Increase Capacity and Employee Satisfaction