Exponential RPA And AI At Nielsen
Measurement and predictive analytics have long driven Nielsen’s two companies – Nielsen Global Connect, which focuses on what consumers buy, and Nielsen Global Media, which focuses on what people see. Customers depend on Nielsen to provide accurate, actionable information and a complete picture of complex and changing marketplaces. And in order to continue to meet these expectations, Nielsen had to look inward and transform.
Nielsen began driving transformation through artificial intelligence and automation tools in 2017. Deborah Fassi, Senior Vice President of Transformation and Automation in Paris, initially focused on how automation, machine learning and other types of AI can transform Nielsen’s internal processes. She is also increasingly focusing on how these tools can help better serve Nielsen customers so they can understand their customers and markets.
After switching to the job, she realized that some of the most dramatic benefits – at least in the short term – could be achieved by automating existing tasks and processes. They often involved structured activities that could be performed by robotic process automation (RPA) and related tools, and could also benefit from the information integration capabilities of RPA. She founded an RPA competence center in 2016 and hired Oleg Royz as Vice President, Global Digital Transformation and RPA. The original goals for RPA were ambitious. Fassi and Royz hoped to scale RPA globally and provide automation services in Nielsen’s various businesses in 104 countries.
The development of RPA at Nielsen
RPA should always be part of a mix of technologies used to accelerate Nielsen’s transformation, along with machine learning, artificial intelligence, and investments in scalable cloud platforms and services. Fassi and Royz believed that RPA would be a great solution to some of Nielsen’s fragmented processes, as it could pull information from multiple systems and provide integrated data for the processes. They chose UiPath as the corporate standard for RPA.
Although the goal was to scale RPA quickly, the initial adoption was slow. Fassi and Royz initially expected large savings from RPA productivity, but had little data to support these performance cases. You have spent a great deal of time evangelizing for the technology, both in terms of its benefits and in terms of its reliability and safety. Managers across the company did not fully understand the value of the technology and were reluctant to allocate resources to implement it – and were responsible for hard savings.
Fassi and Royz eventually took a new approach, supported by Nielsen executives, to implement RPA with the help of dedicated RPA developers and focus on saving time rather than reducing costs. They promoted the technology to each manager with a team to make their business more efficient. Managers would appoint someone on their teams to work with the CoE and provide process knowledge, and the CoE would develop the RPA application for free for the manager’s budget. The RPA CoE would eliminate the financial risk and the implementation risk. The entire development would be financed centrally. Managers would have complete freedom in how they would use the hours saved and the RPA CoE would only keep track of the hours saved from manual labor. As Fassi put it:
With this new approach, acceptance and enthusiasm increased rapidly. Oleg Royz’s CoE has now worked with forty different business units across the company. The pace of RPA development increased dramatically. Many employees volunteered to take UiPath training courses and became RPA champions and ambassadors across the company.
According to Fassi, the CoE measures its success in terms of automated working hours – each full-time equivalent is 2000 hours per year. The CoE implementation team has a maximum throughput of 300,000 hours of new automation per year. So far, savings totaling 450,000 hours can be documented. Fassi says she doesn’t need to know what will happen to the hours saved. “That’s up to it,” she notes. She doesn’t think this will result in a huge job loss as the business grows. “We definitely saw wear and tear on some teams, and RPA makes it easier to deal with.” Some aspects of Nielsen’s business grew quickly and the teams that supported them were very stressed. They found the RPA offering very attractive and found it helpful in dealing with the growth.
Fassi also has AI and machine learning in their portfolio, and Nielsen is trying to connect RPA robots with AI and other tools. You’ve already combined RPA with Optical Character Recognition (OCR) to extract information from documents and Natural Language Generation (NLG) to create automated summaries of business presentations. Some of their robots are also integrated with chatbots and can make calls.
Some of their RPA applications are also enhanced with machine learning. The most common application is text classification so that customer inquiries or work instructions can be forwarded to the right people or departments.
They also use RPA to integrate with a wide variety of enterprise applications. Processes in which ServiceNow, Salesforce.com and SAP are involved can be automated with RPA robots. In the past, they have used people as “integration glue” to submit data and create metrics. Now they leave more time for creative thinking and value-adding activities.
According to Fassi, the exponential increase in RPA usage has started to drive cultural change. Executives enjoy the opportunity to improve margins through greater efficiency and to strengthen trust through higher information quality. Managers can use the tools to better understand their teams’ financial performance and test new ideas. Each employee can think about manual and repetitive tasks that they perform and whether they might create their own automation. And many of the improved internal processes ultimately benefit customers as well.
Fassi argues that automation is quickly becoming a competitive advantage for Nielsen:
RPA started slowly but has grown exponentially. Now that our operational processes are becoming increasingly automated, they are becoming much more efficient. We are now collecting a lot more data and we need to be able to analyze it better for us and our customers. Automation makes that possible.