Triple-A Trifecta pioneer Corning takes aim at data-driven RPA

In 2018, the global technology services and business operations industry continues its obsession with all things automation. HfS’ research agenda in this area is informed by what we call the HfS Triple-A Trifecta (see Exhibit 1), where value can be found at the intersection of the formidable change agents of smart analytics, robotic process automation (RPA), and artificial intelligence.

Exhibit 1: The HfS Triple-A Trifecta finds value at the intersection of data and automation

As organizations experiment, pilot, and operationalize various automation initiatives, we advise that data curation should be a key consideration in these discussions—what meaningful data is generated by automation that can further be used to improve the overall processes? Instead of considering machine learning and deep learning as linear steps in progression after RPA, enterprises can ground their investments in sound data curation that can inform all automation initiatives. However, this is an opportunity many organizations still miss when embarking on their intelligent automation journeys.

Unlocking value in the Triple-A Trifecta at Corning

Our research allowed us the opportunity to learn about the RPA initiatives coming together at an organization where the automation leaders are doing exactly this—discovering and planning at the intersection of data, analytics, automation, and AI.

Chad Keenan, the Director of shared services at the high-tech materials science company Corning, was at a decision-point in 2015. His organization was looking at traditional outsourcing of business process services to augment in-house capabilities. They quickly realized that some of the service providers they were evaluating didn’t have a strong vision for process automation and analytics, where the company believed their future lay. So, Corning Shared Services decided to embark on its own automation journey with support from Corning IT and external partners like Deloitte and Ataway. Keenan recounts: “I remember sitting in a control room, looking at a client dashboard and realizing that the service providers are only allowing clients access to as little data as contractually required, and efficiency gains would only be measured around those data points. We pivoted from traditional outsourcing to focusing on how RPA could accelerate value and change our operating model, which was competitive in the BPO approach.”

Keenan has been working in business transformation for over 20 years, and so he approached this more thoughtfully and looked at what his organization wanted out of the program holistically. He wanted a closed-loop delivery model with data as a core component of it. As a part of standing up their RPA Center of Excellence (CoE), they picked processes including order-to-cash and record-to-report. Those are pretty standard candidates, but they also stood up a metrics database to capture the data created by the automation and plugged in a Business Intelligence (BI) tool to show overall performance metrics to process owners. It became quite evident that there would be value from this real-time process automation data. Their primary use case was to have mobile dashboards relative to transactions flowing through automations. They had wanted to conclude their business case around efficiency gains, but then they started thinking about how data sets can benefit the overall end-to-end process.

A shared services organization usually does not have insight or control across broader process chains. Departments such as finance and accounting have their own processes and methods, and while some of their work transitions to the shared service center, the center is usually not in a position to dictate process changes to internal client departments. With the new insights, Keenan’s team started to pick apart processes and metadata from suppliers and customers; then they were able to focus on uncovering data for components of the process where they didn’t have visibility.

Next phase—using curated data to feed machine learning algorithms

Keenan comments on the next steps: “The insights from fully mechanized dashboards and process operation analytics on those operations have become the next wave of data. We are now working with our data scientists and automation advisors to see how machine learning can be used, particularly on our business and process insights.” Corning’s operations teams knew they would get good metrics. However, they didn’t anticipate how much rich data insight they would gain into supplier practices, supplier engagement, and customer practices in areas such as dispute management, credit, and collections.

It’s not all clear yet, but Keenan believes analytics in 2018 will give his organization much better knowledge of its supply- and customer-base, relative to how the overall organization interacts with them. This is therefore a great example of how the Triple-A Trifecta of RPA, analytics, and AI is working together to deliver actionable insights back to Corning, from what previously were black-box business processes.

Keenan has spent so long in business transformation that we had to ask him if all this was feeling very familiar. “What we’re doing now feels different because of the pace of change with which we are advancing,” he says. “But, what is still the same is the need for change management.” The larger business process change agenda is absolutely critical to getting these investments operationalized correctly. The impact on people and the “future-state” business processes needs to be considered front-and-center during these bold programs, to ensure that the business can fully absorb the benefits realized. It’s a factor we noted as key to overall RPA CX in our report The HfS Robotic Process Automation Customer Experience Big Picture View.

The bottom line:

Automation practitioners must realize and exploit the linkages between data, analytics, and automation technologies to get past basic cost benefits and create new value for their organizations.

In formulating business cases, think of AI, analytics, and automation in the context of your process—not as it is, but how it can be reimagined using data, considering upstream and downstream activities that impact it. Taking a data-driven approach to RPA is one way to think about this, thereby elevating the discussion beyond pure cost-reduction to value-added actionable intelligence.

Corning is already on this journey—a Trifecta pioneer re-contextualizing its automation discussions around data-driven RPA.