Building the Business Case for Robotic Process Automation
Robotic Process Automation (RPA) is gaining traction across industries such as banking, insurance, healthcare, and telecom and is being used to automate labor-intensive processes. It is attractive for business operation executives, as RPA-based process transformation initiatives do not require a lot of dependency on the IT team; projects can go live in a matter of weeks and do not require significant investment. In spite of significant media coverage and reports of successful adoption, there are multiple barriers to the adoption of RPA in enterprises. A major reason for this is the challenge in projecting how it would benefit the business directly and in showcasing to senior management how RPA would provide return on investment.
The first question to address is whether you need RPA at all. The best candidates for RPA are operational processes that are low in complexity with fewer exceptions, have high volume and thus high manual effort, and deal with sources of inputs that are structured in nature. Examples of structured data sources are web-based fields or spreadsheets where the input data resides in a fixed field and which a trained robot can consistently locate. Processes that are more complex in nature can also be considered and will require the workflow to be predictable and capable of being modelled and a tool capable of organizing the input source, if unstructured. Processes which have too low a headcount to justify an investment in RPA, are already candidates for an IT automation initiative, have high volatility (i.e., frequently undergo workflow changes), or are exceedingly complex are bad candidates for RPA.
Consider the following costs when building a business case.
Initial robot software license cost: The primary cost is the license for the RPA tool. Two common pricing models are a fixed-license cost for each instance of a robot that is deployed and a cost per process.
How do you determine the number of robots required? In the first model, you need to estimate the number of robots required. Ideally there should be a dedicated robot for each process, but the total number of robots deployed can be optimized by exploring whether one robot can handle multiple processes. Each process would need its own time slot on the same robot, as a single robot cannot execute two or more tasks simultaneously.
After this, the next step is to choose the right RPA tool. The tools’ capabilities and costs, especially with respect to those of a full-time employee (FTE), are the two primary factors to be considered here. The location of the FTEs for the target process or function is important when evaluating tool cost. A robot can cost as little as one-third of the price of an offshore FTE and as little as one-fifth of the price of an onsite FTE.(Source: Institute for Robotic Process Automation)
Robot training: Robotics tools provide an interface to configure an automated process and train a robot. Corresponding software development lifecycle costs are involved in this implementation, and they rise with the complexity of the automated task.
Process evaluation: Consultants and robotics SMEs can be engaged to evaluate a business process for automation suitability, automation effort estimates, and tool recommendations. The cost of such exercises tends to go down over time as a company’s internal competency to perform such evaluations improves with increasing deployment of automated processes.
IT infrastructure: Robotics tools usually use a client server setup, and while the client setup requirements match the standard configuration of most desktop-based software, the server will host a control tower to manage multiple robots and the database for automation scripts, audit files, and logs. Some robotics tools do not come with a built-in database component and will require a separate database instance to be linked and the required database licenses to be available. There are tools with a SaaS approach, which reduces the costs for onsite infrastructure maintenance, but this configuration may not be ideal for industries like banking, where customer privacy and security regulations apply.
Annual software license: Most popular robotics tools have an annual license fee, which comes with technical support and upgrades for a fixed period of time.
Ongoing robot maintenance and support: Automated processes require support for exceptions that may not have been covered when modeling the business process for robot training and changes to the process that can happen post-implementation. The second case is an ongoing expense and includes the cost for retraining the robot and for a process expert.
Changes in operating model: Changing a business operations model to a robotics-enabled one will require factoring in the cost of training the workforce to work alongside and with a robotics tool and the cost of redeploying operators to newer areas and training them in the skills required in the new areas.
Robotics CoE: Companies that have embarked on the robotics journey will need to maintain a Robotic Center of Excellence consisting of representation from operations, business, and IT teams who will identify use cases, engage with robotics vendors and consultants, maintain robotics architecture and infrastructure, lead automation program planning and governance, establish best practices, take charge of the socialization of robotics across the organization, and manage soft aspects like post-automation workforce training and redeployment.
Benefits of robotic automation
An assessment of the benefits must start with defining why robotic automation is necessary and the pain points that are being targeted. They are usually in the following three areas.
Reduction in manual effort: The primary hard quantitative benefit is the reduction in manual effort. Once the suitable set of processes has been identified, the total cost of manual labor in these processes and the percentage of manual effort reduction that can be expected after robotic automation must be identified to quantify this benefit. A simple way to model the expected manual effort reduction is to baseline your model using results that have been seen in your industry.
A data-driven methodology for projecting the savings from RPA would be a proof of concept–driven approach. A set of processes in a business function, each one ideally being of a different complexity level, can be automated, and the results, in terms of efficiency improvement and manual effort reduction, can be extrapolated to all the suitable processes in the function to arrive at an estimate of the overall benefit that can be expected.
Efficiency and process throughput improvement: The impact of efficiency improvement for a particular process and its impact on the bottom line depend on the nature of the process. Customer-facing processes like bank account opening, when automated, improve the customer experience and lead to better customer acquisition and retention rates. Automating finance functions like the accounts payable process lead to on-time payments, lesser penalties, and better cash flow.
Error reduction: The business impact of errors in operations depends on the function and can be significant in customer-facing operations or regulatory- or compliance-related functions.
Other benefits: Another area where benefits are realized, which corresponds to reduction in the manual labor for activity, is savings in facilities, systems, and similar resources.
“Soft” benefits include better employee productivity and utilization of manual resources in more high-value tasks. Also, a robotics tool can be configured to store activity logs and can give senior management better visibility and control of operational metrics. Audit trails can also facilitate better compliance.
Enterprises are adopting RPA in a big way, and making the business case to senior management for adopting this technology in your organization involves careful research and analysis to quantify the benefits of change.