Editor's Note: This piece originally appeared in our sister publication, DOCUMENT Strategy Media. We thought our Mailing Systems Technology readers would find it useful, as well.

What is the value of a piece of mail? A large insurance company or healthcare provider that depends on mail campaigns for customer transactions and marketing initiatives might estimate the cost of a simple direct mail piece at less than 50 cents. Even counting the cost of paper, printing, postage, machine maintenance, and related facility expenses, it’s not a bank-breaker. However, when an enterprise sends out a large mailing — whether from marketing, billing, or claims — and it receives a portion of it back as returned mail, the cost of non-delivery can be substantial.

Not only does each piece of returned mail represent a loss of the original 50 cents, but it also immediately requires an additional $1.25 to $1.50 to scan the address, fix the associated databases, and remail it. For a large enterprise that receives an average of 1,000 pieces of returned mail a day, this expense can easily add up to more than $250,000 a year.

If this isn’t enough, a piece of returned mail also represents a missed opportunity to attract a potential new customer or communicate important information to an existing one. Returned transactional mail —including contracts, bills, or payment reminders—can result in a loss of revenue, increased administrative costs, and a poor customer experience.

For many large enterprises, the job of processing returned mail and correcting addresses in source databases is not insignificant. In fact, business operations leaders have long sought the benefits of outsourcing and shared services centers to eliminate the duplication of effort, reduce costs, and enhance speed, accuracy, and auditability of these important back-office processes.

Looking at Robotic Process Automation

Today, some of these decision makers are adding a new arrow to their quiver: robotic process automation (RPA). Because RPA software can automate repetitive, algorithmic processes that involve structured digital data and straightforward business rules, it seems, at first glance, the perfect solution for taking on business processes like those needed to handle returned mail.

In the shared services center of a large insurance company, for example, RPA is being used to filter returned mail and designate it as either a duplicate, addressed to a deceased person, or in need of an address change. It then automatically updates the information in the dozen or so related administrative systems and processes a piece of mail to the correct address. Filtering these items allows employees to focus on the work items that truly require address verification, and cleansing the data means that the company can get mail to its customers in two days instead of 30. Like so many improvements to back-office processes, streamlining those around returned mail has a multiplier effect: The more quickly the company can update correct information across its related systems, the less risk it faces for additional rounds of returned mail.

While RPA bots can run these processes 24 hours a day and do the job of full-time human equivalents at a reduced cost (for an average of $5,000 per robot), business operations leaders are finding other upsides equally—or even more—compelling. When employees, who understand the intricacies of back-office processes, work alongside RPA bots, they are freed from the repetitive and mundane work of sorting address changes and cutting and pasting data from one system to another, turning instead to the strategic work that contributes to continuous improvement and operational gains down the road.

4 Tips to Leverage RPA

Despite these compelling benefits for document management, RPA is not a plug-and-play deployment. To realize the full potential of this new technology, business operations leaders should keep in mind these important points:

1. Engage early with information technology (IT).
Educate yourself about what a robot can and can’t do. Conduct the appropriate due diligence to align permissions for the bots by involving the risk management and information security teams as well as IT. If you are hoping RPA will scale safely across data systems, taking time upfront to sort out the security requirements will get you there faster.

2. Design an RPA operating and technology model that fits the company.
Take into account the relevant people, process, and tools across the organization. This will then form the foundation to build an RPA center of excellence using the model as a foundation. Put into place best practices so the company can optimize the deployment and the organizational impact it can have.

3. Focus on core end-to-end business processes.
Break down processes into their component parts and map the steps involved. Identify where one application must talk to another. Because the back office has traditionally run on the nuanced activities and decision points of human individuals, RPA bots must be carefully configured to mimic those activities. Let decision makers involved in the process know that, during configuration, an RPA bot is set to work based on a specific set of business rules; it executes on a script with no chance of wayward interpretation and poses no risk of altering the process or system.

4. Identify and evaluate technical integration challenges before deployment.
RPA merely automates the keystrokes and decision points of a human worker for an assigned process based on defined business rules. Bots will not reduce keystrokes or exceptions, as they are not designed to create a future-state process. This means that if, during a certain business process, an employee must check another screen with undefined rules to capture related information 20% of the time, the RPA robot must be configured to identify such exception scenarios, and not all processes can be 100% automated with a bot. Take a careful look at the applications, data sources, unique identifiers, data fields, and the variety of commands needed for a smooth implementation.

Of course, Rome was not built in a day, and automating returned mail and other business processes in the shared services center of a large enterprise will not be either. By carefully designing an RPA operating and technology model before deployment, business operations leaders will avoid the pitfalls common with the adoption of emerging technologies and get the most out of their investment.

Kevin Tesch is a Director of Robotic Process Automation (RPA) at ISG. He has more than 23 years of experience in project, process, and change management in both business- and technology-focused projects. Follow ISG @ISG_News.

Dave Zamorski is the Senior Director of Enterprise Print Services at ISG and has extensive background in print management, production, workflow automation, and communications. Dave specializes in optimizing document-intensive environments and working with clients to develop a clearly defined print and digitization strategy that best meets their long-term objectives.