Should I do an RPA Proof of Concept?
We often hear the question, “Why do an RPA proof of concept (POC)? Doesn’t RPA work right out of the box?”
Improving your chances for success
Companies that do an RPA POC have a much better chance of achieving success in their RPA initiatives. The POC process lays the groundwork for a larger automation program with ongoing success. A POC also allows the team to prove process automation with existing systems, applications, data. and processes within their current environment. A POC can result in improved profitability, increased cost savings, fewer errors, improved quality and scalability, and more rapid deployment.
Selecting the right process
Choosing the right process for a POC increases the chance of adoption. A recent EY study showed that as many as 30 to 50% of initial RPA projects fail. Most RPA failures stem from picking the wrong process to automate – one that is too complex, has unconsidered variations, or doesn’t provide adequate business value. Processes should be well understood before you automate them.
The ideal candidate process for an RPA POC:
- Reasonably sized, preferably as small as possible while also providing benefits.
- Relatively simple with few “Ys” in the road, where the primary path covers 80 percent of the process and requires only one person to perform the task.
- Non-mission critical tasks with a high degree of user frustration which will increase user approval and result in a higher likelihood of follow-on RPA initiatives.
- Less complex to automate, for instance, avoid virtual environments. For example, Citrix as it requires surface automation techniques. This is achievable but may add unneeded complexity early in the adoption cycle.
- Highly repetitive, mature, and stable processes that are preferably documented with no significant changes expected soon.
- Well defined (even strict) business rules with no human judgment/decision making. Decisions are rule-based and logical and can easily be depicted on a decision tree i.e. input – process – expected output.
- Large transaction volumes help ensure a solid ROI with more robust testing.
- Digital and structured data inputs such as stable databases or templates: JSON, XML, XLS, and CSV; avoid data from printed, scanned, PDF, hand-written, or paper documents.
- Available systems for testing which can provide for many test cases.
- Quickly generate a high return on investment which will be based on your driving factors for automating such as reduced operational risk.
- Requires the user to access multiple systems to complete the task.
Once an appropriate business process has been identified and the scope has been established a POC can be conducted. Once the POC is completed, it provides a convenient tool to demonstrate to stakeholders how true automation can work in your environment. In addition, it establishes confidence in building automation across the enterprise.
Steps for establishing and executing a successful POC
- Planning: Establish the business case which should include outlining the goals, the criteria for success (KPIs), the project scope, the high-level ROI expectations, and budget.
- Assessment: Determine the appetite of management, stakeholders, and employees for attacking this specific process.
- Process Preparation: Perform a detailed walk-through of the business process selected for automation. Remove any steps that are outdated, extraneous or redundant. Process mining software can be especially helpful in clarifying how processes currently work versus how they should work and in pinpointing bottlenecks.
- Technical Preparation: Establish a technical environment that can support the requirements of your proposed POC. This might include a laptop or desktop that has the RPA products necessary to complete the POC installed. This platform will also include security access to the systems upon which the processes are currently performed.
- Development: Develop the RPA code necessary to affect the POC. Recording the process upfront can be extremely useful as it allows the developer to review the process repeatedly during the development process. Task mining software is a great option for recording tasks.
- Demonstration of results: Schedule and perform a demonstration of the automation for the key stakeholders. Validate the success criteria to ensure goals and objectives have been met.
- Selling the value: The next steps should include “sales activities” to generate excitement for the personnel that will be using the automation. The introduction of any new technology into someone’s job should be handled carefully. Change can be difficult for some individuals and their acceptance will be critical to the success of your automation efforts. Help them to understand how automation will assist them in doing their job.
If you are just starting your RPA journey, a POC is your best first step.