Robotic Process Automation

 RPA sounds like science fiction.  Are robots real?  Do they have a role in the future of banking?  The answer is simple:  RPA is relevant wherever you have repetitive processes that have very clear business rules.

With RPA, businesses can automate mundane rules-based business processes, enabling business users to devote more time to serving customers or other higher-value work. Others see RPA as a station stop on the way to intelligent automation (IA) via machine learning (ML) and artificial intelligence (AI) tools, which can be trained to make judgments about future outputs.

Using RPA tools, a company can configure software, or a “robot,” to capture and interpret applications for processing a transaction, manipulating data, triggering responses and communicating with other digital systems. RPA scenarios range from something as simple as generating an automatic response to an email to deploying thousands of bots, each programmed to automate jobs in an ERP system.

RPA is an emerging form of business process automation.  It is currently being deployed in many banks around the country, but is still in its embryonic stage.  RPA is not an extensive user of Artificial Intelligence (AI) yet.  In most cases, it is technology designed to automate manual, highly repetitive, drudgery processes around the bank.

A good example is the calculation of capital gains and losses in a fiduciary tax department.  The calculation is straightforward, and machine can do it more efficiently than humans – in terms of both accuracy and speed.  Another example is a small element in consumer debt collections, where two systems do not talk to each other so you need to copy information from one system to the other.  Or, when a consumer loan is approved, a robot orders the title and appraisals.

The first prerequisite of RPA is a thorough understanding of your processes, including all exceptions and intricacies.  Then, selecting the best candidates for automation is a difficult task.  There are thousands of candidates throughout bank operations, and resources are limited.

One possible set of criteria for prioritizing process automation candidates includes:

• ROI (will the RPA pay for itself and more?  Will it free capacity or reduce headcount?)

• Quality (error reduction opportunities)

• Consistency – will RPA improve consistency of process and thereby enhance compliance?

• How frequent is the process to be automated?

• What will be the impact of automation on customers and employees?

• The process lends itself to business rules

• The process is not complex

Generally, RPA is considered a threat to people, but, in reality, in can enhance employee happiness and well-being.  Many processes, especially in bank operations, are a drag on the human spirit:  they are repetitive, tedious and unsatisfying.  Why not free people to improve their contribution to the organization, use their brains more effectively and stop spending time on unrewarding, no value-added but yet necessary processes?

RPA can be applied to whole processes, but it’s best to start with snippets or small, simple processes.  The robot can be trained to follow whatever rules you give it, and it will do exactly what you tell them, which is why it is important to fully understand and map out a process before you introduce RPA to it.

In some cases, such as fraud detection, you can use the robot to identify likely fraud suspects, then use people to work the leads.  You can then expand the robot’s role to gather data and complete other, non-judgment related tasks.

Where should RPA be housed?  The knee-jerk reaction is IT, but the deep knowledge of bank processes resides in operations.  It makes sense to build a collaborative team and empower Operations to select the best candidates for RPA, even though many IT departments want to keep control of robot access.

The finance department is another area that should not take control of this initiative, due to their focus on payback and headcount.  This can create a negative connotation to RPA which will be difficult to overcome.

There is also understandable resistance in replacing people, and RPA indeed can displace humans.  But note that the proposed use here is not to create a learning machine but to replace people with robots where the human touch does not add value to the process (think credit scoring at the best and worst credit cases).  RPA can be used to assist front line staff in decisioning fee waivers, for example, a perfect combination or person and machine.  In many cases employee engagement has improved, not diminished, after RPA was deployed, since the more hateful part of their job has been automated, freeing them to engage in other, less tedious activities.

Organizational buy-in for RPA is critical, as it is for any new process, initiative or technology.  It is often not recognized as mission-critical, which causes it to be pushed to the back burner too often.  An agile organization can accommodate RPA experimentation with very small items to test different technological solutions and RPA applications, learning from the experience and iteratively improving the RPA process selection.

A progressive CEO can make all the difference when it comes to RPA and similar automation.  With the CEO’s support you can develop a set of metrics to demonstrate the value of the process not just in terms of dollars and cents but also in terms of improved employee engagement scores, KRIs and other metrics.   

There are many pitfalls to RPA implementation.  Here are some thoughts as to how to avoid them.

1. Set and manage expectations

Quick wins are possible with RPA, but scaling RPA is difficult. Ambitious claims about RPA from vendors and implementation consultants haven't helped. Go in with a cautiously optimistic mindset and do not over-promise.

2. Consider business impact

RPA is often propped up as a mechanism to bolster return on investment, free capacity or reduce costs. Improving the customer experience can also be a major benefit. For example, airlines employ thousands of customer service agents, yet customers are still waiting in the queue to have their call fielded. A chatbot, could help alleviate some of that wait.”

3. Involve IT, lines of business and compliance early and often

Operations folks without technical expertise are using cloud software to implement RPA right in their business units. Often, the CIO tends to step in and block them due to lack of standardization, governance and collaboration.  Involve IT and other partners from the outset to ensure they get the resources they require.

4. Poor design, change management can wreak havoc

Many implementations fail because design and change are poorly managed.  In the rush to get something deployed, some companies overlook communication exchanges between the various bots, which can break a business process. Before you implement, think about the operating model design; map out how you expect the various bots and the human element to work together." Also, don’t forget to consider the changes new operations will have on an organization's business processes. Plan for this well in advance to avoid business disruption.

5. Don't fall down the data rabbit hole

A bank deploying thousands of bots to automate manual data entry or to monitor software operations generates a ton of data. This can lure executives into an unfortunate scenario where they are looking to leverage the data. Then, instead of a simple project to automate a simple process scope creep and heightened expectations doom the project to fall short.

6. Project governance is paramount

Another problem that pops up in RPA is the failure to plan for certain roadblocks.  A simple example is a change in your password policy that isn’t transferred to the bots, resulting in lost data and much frustration.  RPA support must constantly check for chokepoints where their RPA solution can bog down, or at least, install a monitoring and alert system to watch for hiccups impacting performance. 

7. Control maintains compliance

There are lot of governance challenges related to instantiating a single bot in environment, let alone thousands. Even basic questions such as, is the bot male or female, can generate much debate.

8. Appoint a person to be in charge of the RPA program

The RPA leader develops business cases, calculating potential cost optimization and ROI, and measures progress against those goals. They are often supported by a small but nimble group that adds discipline and accountability to the program.

9. Don’t forget the impact on people

Some banks get so enamored with RPA that they neglect to involve HR, which can create some nightmare scenarios for employees who find their daily processes and workflows disrupted. Ultimately, it’s still all about people..

Initial RPA deployment is often rocky, including organizational strife, employee concerns, resistance and political jostling.  However, successful deployment brings with it high demand, which often can’t be met with existing resources.  Once people understand the selection criteria for RPA they will help you assess suitability and continue to improve the overall operation of the bank through small increments.

One key to success is the organization’s change management process and agility.  A good methodology to use is ADKAR, which can be utilized across many change areas in the bank.  RPA is an extreme example of change, as it is so foreign to so many of us, and is still in the emerging phase.  Nevertheless, organizational agility is critical, and experimentation with new technologies, including RPA, should be integrated to your culture to facilitate growth, efficiency, creativity and survival.