Artificial intelligence (...
When to Apply RPA in Finance?
According to research conducted by the Hackett Group, the market for robotic process automation (RPA) is real and growing. It has the potential to change the business process outsourcing (BPO) landscape, global business services (GBS) organizations, and broader business-specific processes.
While we can’t predict the future, the market is expected to grow (50%+ through 2020) transforming our idea of automation and service delivery. RPA can be a game changer because it allows companies to achieve remarkable cost savings without having to change their existing technology infrastructure and process design. Anyone (everyone) dealing with legacy systems knows just how valuable that can be.
RPA is about software that sits between other applications or websites and does complex, rule-based work that would typically be done by people. RPA can mimic human interfaces and operate computer systems just like a person would, for example, it can log in and out of an application, copy and paste data or choose from a drop-down menu. It brings the biggest benefit to companies that have a fragmented technology landscape with multiple applications or instances of an ERP.
Two big factors are driving RPA adoption in finance and elsewhere:
- Speed of Impact: RPA has a short development cycle, and yields quick benefits like reduction in FTEs, based on low initial investment with a high ROI in weeks and months, not years.
- Enhanced Capabilities: RPA-enabled processes are highly auditable and consistent and have extremely low error rates; they are also very scalable (no need to add resources as volume rises) and provide the data required to run sophisticated analytics.
Ask Yourself These 8 Questions
RPA has many applications within the finance organization because many finance processes fall within the realm of activities that fit the RPA application model, from tactical bank reconciliation to end-to-end processes such as account-to-report (A2R). There are eight questions finance executives should ask when they consider whether a specific process is ripe for robotic automation:
- Is the process executed frequently and in large volume? It doesn’t make sense to use a robot to perform a process that only happens once in while in small batches. That’s much more appropriate for a person.
- Does the process require access to multiple systems? A robot can much more quickly and efficiently log in/out of applications and cut and paste data from one application to another.
- Can the process can be broken down into unambiguous rules? Unlike people, software needs clear rules. If there are judgment calls, it cannot perform its task.
- Is the process structured and digital data available? The robots need to follow an established and well understood set of process steps and have clear access to all necessary data.
- Is the process prone to human error? Robots don’t make mistakes when switching application and re-inputting data.
- Does the process require limited exception handling? Too many exceptions and the robots stop making sense. RPA’s ROI is built upon doing large volume of repeatable acts. If many of the actions are “kicked out” for human intervention it defeats the entire purpose.
- Does the process, once started, require limited human intervention? If the answer is that the process needs a lot of human intervention than it doesn’t make sense to get robots involved.
- How complex is the stakeholder/process ecosystem? Processes that serve multiple stakeholders and exist in very complex ecosystems may not lend themselves well to fully automated solutions because they require different outcomes.
Case in Point: Cash Application
Some finance processes are particularly suitable for RPA. Take the order-to-cash process. One of its sub-element that has seen a lot of RPA activity is customer billing. It meets the criteria listed above.
- High frequency and high volume
- Requires access to multiple systems
- Structured with available data
- Prone to human error
- There are limited exceptions and human intervention required once started
- Primary stakeholders are the customers.
In customer billing, the robot can process the general invoice. It can then consolidate the billing information into the right format, transmit the invoice and post the AR transaction.
RPA is in early adoption phase, but it’s showing great promise in the finance space. It’s a quick and high ROI approach to resolving burning issues (bloated process cost, high error rate), as well as a longer-term automation solution to specific process improvement that does not require a complete overhaul of existing process or system architecture.
RPA adoption is forecast to grow by a factor of 10 over the next two-to-three years, according to the Hackett Group 2017 key issues survey. Many finance organizations are already taking advantage of robotics and will lead the next wave of capability improvements and cost savings. Keep these three things in mind when looking at RPA solutions:
- To be effective RPA must be considered within a broader automation roadmap and service delivery model for finance, including longer-term outsourcing and GBS strategies.
- Look at different tools for different situations. There are multiple vendors offering solutions specific to particular needs.
- Best success comes from optimizing processes and combining them with RPA.