Insights from The Lab

Robotics in Banking with 4 RPA Use Case Examples

Article by : Chris Wilds

Robotics in banking

Robotic process automation (RPA) is about to change the way banks conduct business faster than any other new piece of technology available. You can think of banking robotics as the “consumerization” of banking automation – meaning that front-line employees will eventually be able to automate their own work if trained properly. For the banking industry, robotics are a new, and completely underutilized way to increase productivity and keep repetitive, manual-labor-intensive processes at a minimum.

What is robotics process automation in banking?

Robotics in banking is defined as the use of robotic process automation software like UiPath, or Blue Prism, to install desktop and end user device level software robots, or an artificial intelligence workforce, or assistants, to help process banking work that is repetitive in nature. Once set up and implemented, banking robots take control of mouse and keyboard actions such as opening applications, clicking, copying and pasting information from one banking system to another, sending emails and other labor-intensive “low-value add” tasks. These robots work at the individual data field level and act similar to an Excel macro across banking software systems.

Established banking institutions have historically relied on multiple legacy core systems in which operational benefits had been oversold during the sales process, and more often than not, the implementation under-delivered and fell short of automation promises. The result is an industry in which large amounts of manual work has to be done outside of and in-between multiple core banking systems in order to reconcile and transcribe data to process transactions. In an industry prone to large-scale white-collar work processing, this snowballs as banks and their IT departments struggle to merge different legacy systems into a coherent workflow.

Robotics enables the banking industry to integrate “the last mile” across business units like it has never been able to do before. RPA use cases can be applied to a wide range of processes, including; retail branches processes, commercial lending, consumer lending, loan processing, underwriting, and anti-money laundering, just to name a few.

Robotic process automation in banking is minimizing the need for extra repetitive manual work tasks, data reconciliation, and transcription—sometimes by up to 70%. While the idea of implementing robotics into banking operations may seem daunting, it’s easier than most companies selling it want you to think. But buyer beware, if you have the same big technology consulting company that sold you legacy systems with “straight through processing” coming in to try to sell RPA, you might want to consider other options for robotics implementation. Below, we’ll discuss some RPA banking use cases that show how surprisingly easy it is to implement robotics, provided you do the required front-end process analysis and standardize work first.

Top 7 Benefits of Robotics in Banking

Retail and commercials banks are facing increased pressure from management, shareholders, and external competition (like fin-tech companies) to reduce costs, increase quality of products, and speed up processing of back-office work. Robotics, when paired with the right type of process analysis, can help banking operations management tackle most large scale and routine data movement tasks with speed of implementation like never before. We are talking about automation in weeks – not months or years.

Benefits of robotics in banking can be financial while operationally improving back office processes and the customer experience. Banks can save money on labor while doing more with less with RPA.

The following list compares benefits of robotics in banking to traditional automation:

  • Banking RPA does not require new core IT infrastructure change or upgrades—it’s a low-cost layer that sits on top and across all currently installed banking applications.
  • Robotics in banking does not require coding experience.
  • RPA for the banking industry is nimble and robots can be tested in short cycle iterations.
  • A banking robot can be installed or updated in less than a week when banking processes change.
  • Front line employees can be trained to maintain and “manage” their own banking robots.
  • Banking robotics can increase morale of human workers by reducing boring data entry work.
  • Banking robots can work 24/7—365 days per year. Banks don’t have to pay a robot overtime, don’t have to pay the health insurance, or worry about them quitting.

What is a RPA Use Case in Banking?

A banking robotic process automation use case is defined as the process of documenting a list of banking operations actions or steps that take place on front-line employee level computers, or other electronic devices, that is used to automate information movement across banking applications. Banking RPA use cases are used as process “blue-prints” by IT consultants, or your own staff, to implement automated scripts that run across multiple IT systems at one time that process data.

How do you develop RPA banking use cases? The 3 step process is detailed in the info-graphic below.

Step 1: You must identify sub-processes on process maps where banking robots can be implemented.

Step 2: You must prioritize and evaluate all of the banking sub-processes and use cases to see which yield the most benefits.

Step 3: Finally, develop the use-case requirements, rules and key strokes that the banking robot must take over.

3 Examples of Robotics Use Cases in Banking:

RPA in banking yields the most benefits when led with deep front-line process analysis and a desk-level work standardization plan across the organization. The only way to realize the maximum benefits of your robotics is to standardize all manual work, then implement the RPA use cases afterward.

Three high level examples of RPA in banking are below:

  • Reduction of consumer loan processing time from 30 minutes to 10 minutes by eliminating copying and pasting of customer information from one banking system to the next
  • Increased accuracy of new bank account opening requests – reduced down-stream errors, and improved system data quality by eliminating data transcription errors from inbound new bank account opening request emails into the core banking system
  • Increased auto loan processing speed of customer verification by automatically validating customer data on government websites, such as the DMV, tax payment, or property appraisal sites

A Detailed Use Case of Robotics in Banking Operations—RPA in Consumer Loan Processing

In this robotics in banking use case example, we’ll follow Cathy, a consumer loan processor who’s getting ready to process a loan of a prospective borrower. Normally, Cathy would need to process each loan request 80% manually at the bank, which would take her at least 20 minutes per customer. She had to cut and paste information between email, multiple loan processing systems, credit bureau, and several government websites.

The pre-RPA banking use case process is as follows:

  • When a customer requests a new consumer loan or line of credit, a call center representative, branch employee, or website captures the data into loan processing system #1.
  • Once the information is received by Cathy, she runs a manual credit check by transcribing data from loan processing system into an external website to pull the credit report.
  • She then saves the credit report as a PDF and attaches it to loan processing system #1.
  • Cathy then copies and pastes the credit score into a field in loan processing system #1.
  • Once the credit check is complete, data is transcribed into 2 other different core banking systems from loan processing system #1.
  • The final step is for Cathy to log into a government website to the validate customer address and appraisal from supplied documents by copying all of the information from loan processing system #1 and pasting it into the website to validate the address of the customer requesting the credit.
  • The confirmed information is then printed as a PDF and attached to loan processing system #1 by Cathy.

This was all done by hand, 20 times per day. 80% of Cathy’s day is spent copying and pasting.

Now, with the help of banking robotics, Cathy’s loan processing job is going to be very different. The banking RPA use case will now process the work as follows:

  • Cathy receives the loan package in the system as she did before.
  • Cathy opens her UiPath robot, it then logs into loan processing system #1 and pulls all the information needed to process the credit check.
  • UiPath opens the credit reporting website and runs the credit check by pulling the information out of loan processing system #1.
  • UiPath creates the PDF copy of the credit report, attaches it to loan processing system #1, and copies the credit score into the credit score field in the system.
  • The above steps only took one click of a button compared to the 80 clicks it took before.
  • Then, UiPath pulls the loan data received in loan processing system #1 and transcribes the data into 2 other core banking systems. Again – with one mouse click.
  • The Banking robot logs into the government website, enters the necessary data to run the address check, and validates the property appraisal and customer address automatically.
  • Finally, the RPA bot saves the address check and appraisal PDF to the to loan processing system #1.

What used to take Cathy 20 minutes to do now takes her 5 minutes, and she is able to focus more on delivering an exceptional banking customer experience instead of just moving data around. This sounds like something that you might want to try out, right? Let us show you how below or even do it for you.

Want to see a banking RPA use case for the commercial lending process? Follow this link: Commercial Lending – Banking RPA Use Case

Additional RPA use case examples in financial services can found below:

Finance and Accounting – Accounts Payable RPA Use Case

Financial Services RPA Use Case

Health Insurance – Claims RPA Use Case

Property and Casualty Insurance – Claims RPA Use Case

Implementing RPA Use Cases in the Banking Industry – Work Standardization Must Come Before Robot Installation

The above example is just one of the many ways banking RPA is making the lending process faster and more cost-effective; saving both the banks and consumers time and money. So how does one go about implementing RPA use cases into a bank? We’ve highlighted the key points that must be completed to undertake when you try to do this on your own, or with the help of a professional.

How to scope the banking RPA use case project

The first step to an effective RPA use case implementation in the banking industry is determining a manageable scope. Think of a small test drive, then broaden your scope to include more once you have learned your first RPA lessons. It might seem tempting to come out swinging with an enterprise-encompassing scope, but don’t fall into the trap. Despite what some consultant firms might tell you, or try to sell you, starting a robotic process automation banking project on a large scale is a great way to fail fast.

Processes where RPA can be implemented in banking include:

  • Account Maintenance
  • Trial Balancing
  • Accounting Onboarding
  • Account Closing
  • Loan Origination
  • Account Reconciliation
  • Account Processing
  • Commercial Banking Operations
  • Card Services
  • Exception Processing
  • Fraud and Risk Review
  • Item Processing
  • Lockbox Operations
  • Print and Statements
  • Wire Administration
  • Direct Loan Underwriting
  • Loan Servicing
  • Collateral Management and Imaging
  • Mortgage Lending
  • Escrow

Financial institutions have multiple divisions in the back office that process and transcribe data all day long, including; credit operations, loan servicing, fraud prevention, collections, and loan operations. Of the multiple functions that banking RPA helps with, one we’d recommend starting with is loan or deposit operations.

Loan operations employees deal with large amounts of data spread out over different systems, fields, and forms. Loan officers, processors, and credit analysts are loaded with manual work, customer information, and have to deal with redundant data processing from multiple core systems.

Start your RPA use case project with a group of 5–50 employees who you know are open to change.

We’ve learned through experience that it’s best, to begin with, a pilot scope to reduce project failure risk and reduce project learning curve costs. This means that identifying use cases for RPA in the banking sector for just a few tasks to start, rather than for each and every process that a bank executes all at once. Org change is less painful if dished out in small doses. And trust me when I say this, once your employees see a banking robot in action they will be excited.

Determine baseline operational cost to calculate the total benefits realized from implementing RPA use cases in your retail bank

To realize the full scope of financial benefits that could be realized from banking RPA use case implementation, it’s essential to have a baseline for operational costs before attempting to add robots to your bank. Make sure to budget for this task to usually take a week or two to do. You will have to coordinate with the HR department at your bank to get the cost for each employee in scope.

It’s also a good idea to measure the costs after the first initial attempt to implement RPA, as well as subsequent attempts over time. This will allow you to show the costs and savings achieved over time with the use of RPA. The best way to make a case for robotic process automation in banking is showing a lot of green to people who are on the hook for benefits, i.e., senior management. So, make sure you have the proof of resulting cost savings before you run around RPAing everything!

Analyze current state processes to determine your robotics in banking use cases

Roll up your sleeves, because it’s time to get dirty. Now’s the time to do deep dive process mapping in your bank. So many people always want to skip this step – and it’s the most important one. We’re not just talking a little here and there, either. You have to map all of the systems and people you’re considering adding banking RPA use cases to. A thorough process analysis of all front and back office tasks needed to carry out daily operations are essential for determining how RPA can help your bank. You need to get down to the “grain of sand at the beach” level of detail to properly document the use cases at the mouse click level.

This could be done conducting what we do here at The Lab called “day in the life of” observations via screen sharing software, or by doing face to face observations in person. We literally watch workers perform tasks for a whole week, all day long, to document where banking RPA can be used. We like to use GoToMeeting, Webex or Team Viewer for the off-site observations. Off-site analysis reduces the cost of having an on-site consultant to analyze the current daily banking operations processes.

Once you’ve completed your shadowing, process mapping software program, such as Microsoft Visio, can help you to visually represent your steps to build your banking RPA use cases. Every detail that goes into the workflow (down to mouse clicks and keystrokes) must be represented. A good way to document this for the IT department or RPA vendor hand off is to use BPMN 2.0 stencil, as this is the “coding level standard” way to represent process flows. Then, you get to repeat it all over again with the next employee and so on and so on, until you have a sample size large enough that it validates your findings. It might not seem like a breeze, but remember, no pain no gain.

Now on to the really difficult part – you must standardize banking processes before robotic process automation implementation — but not after. Here’s the deal: you shouldn’t be thinking of pressing that banking sector RPA button if you’re not up to analyzing and standardizing your current manual processes first.

Banking processes might be different across financial institutions, but they also vary across employees internally. Successfully rolling banking RPA out to scale means everything must be standardized and planned out before implementation so that it is executed properly and efficiently.

Find project managers with lean project experience focused on transactional and white-collar process work. Their previous project management experience of standardizing processes at scale will serve your RPA banking use cases well.

Hire a banking focused RPA vendor to analyze processes and develop use cases for you

Once you have your workflows standardized and mapped out, it’s time to implement your new banking RPA processes. We recommend using UiPath over Blue Prism or Automation Anywhere – it is the easiest to use. And, there’s a free Community Version that we think every business should try. You may also find it beneficial to research use cases to reduce any time lost on the learning curve that comes with using a new technology. Another option is to become RPA certified so you can do it yourself; saving money and adding a skill to your CV.

Of course, the easiest option is to hire a banking RPA consultant to both consult and handle the mapping and process standardization. They can then hand off the installation of the actual RPA technology once everything is sent to the implementation team. Prices vary among consulting firms, but we’re a firm believer in less cost and risk.


Risks of Robotics and RPA in Banking

With any new process change, or technology update – comes risk. However, when compared to long term core technology implementations, the operational risk of RPA is far lower. This is because a robot can be turned off instantly and won’t shut down your core banking processes, if you choose to do so. Also, robots don’t change entire processes and require organization wide change management, they only change individual user desktop settings. Has an Excel macro every brought down your IT systems? Most likely not. And neither will a robot if you keep them confined to data transcribing and scraping tasks.

Risk categories of RPA in banking include:

1) Operational Risk

2) Compliance Risk

3) Data Quality Risk

4) Ethical Risk

Operational Risk of Banking RPA

At the beginning of any RPA project will come risk of push back from internal staff. Staff will be intimidated by the thought of a “robot coming for their job”, when in reality the robot is simply a tool to help them reduce amount of work required to move data between systems and applications. In addition to the risk of employee dissonance, should the goals of the project not be clearly communicated, is the risk of robot down time and disruption in operations. The robot could stop working due to operating system updates – so the need for operational readiness to update and repair robots is needed. Good news is that RPA robots can be updated within a few hours should they have to be reconfigured.

Compliance Risk of Banking Robotics

When it comes to compliance checkpoints and confirmations of required process steps with banking RPA – the way robots are being implemented and the tracking of robot work tasks, just like human work tasks, is what matters most. The spreading of unchecked banking bots across the organization, especially around processes that require regulatory checkpoints, can become problematic should process managers not have an inventory of installed banking robots and what process it is they are doing. Should RPA bots that are not built to adhere to strict compliance processes become a bank’s entire process workforce, they can lead to major compliance issues. This is why it is important to have banking bots only act extensions of humans, and not replace decision making ability unless it has been fully thought out. Robotic processes require fool proof audit trails just like human workers do. Banking RPA is end user-friendly technology but the management and inventorying of them requires discipline.

Data Quality Risks of Robotics in Banking

With every incremental terabyte of data being loaded into a banking system, the reality of the proliferation of poor quality data is realized. Even if a robot can reduce the data transcribing errors of a back-office employee from spreadsheet into a system, what if the data received from the front office is already in bad order and that data gets loaded into the system? Now imagine the back-office employee being able to load 100x more data into the system – that’s a lot of bad quality data going in that will have to be cleaned up downstream. “Big data” standardization processes are an element often overlooked during RPA implementation projects that banks must consider.

Ethical Risk of Banking RPA

Ethics and robots sounds like a scene out of the Terminator movie series. But, all enterprises must constantly balance whether to invest in their people or technology. Merely investing in technology to replace staff or using outsourcing to eliminate internal head count can reduce moral – quickly. Thankfully, with RPA, it doesn’t have to be one or the other. Banking robots allow for a combination of humans and technology. Additional ethical considerations of robotics in banking, outside of the organization include those to society. What would the fall-out from the 2009 financial crisis had been if bad actor banks were able to process 100x more mortgages and sub-prime loans per day?

Benefits of choosing The Lab for analyzing, developing use cases, and supporting the implementation of banking RPA

One great option for implementing RPA into your financial services operations is to use the “a la carte” RPA mapping offered by our team at The Lab. Our model is designed around starting small, earning our client’s trust, proving our competence, then increasing efforts over time. We offer a variety of template IP options to speed up banking RPA analysis and implementation. We provide RPA solutions for financial institutions who want to step up their game, or simply keep up with the rapidly changing industry. We would love to chat with you further about our banking RPA solutions, methodology, and mapping if you’d like to learn more.