7 insights I wish I had when building RPA solutions for audit

Gartner reported that the revenue generated by worldwide Robotic Process Automation (RPA) software will reach 2,9 billion dollars in 2022. This raises the question of what the potential for RPA in audit could be. Considering the high number of repetitive tasks, shortage of financial auditors, and the need for innovation, there are many opportunities for RPA in the audit industry.

Regardless of the enormous potential for RPA in audit, many companies struggle with how and where to get started. In this article, 7 insights will be shared on setting up an RPA solution to automate audit procedures.

1: Think the business case through

Before developing the concept of the RPA solution, make sure there is a strong business case. Even though this sounds obvious, it’s one of the most common pitfalls when setting up an RPA solution. To set up an effective RPA solution, there is a minimum of data elements required for the process to be more efficient than a manual process.

An example of an inefficient RPA solution would be a robot to fill in a client-specific questionnaire that is based on professional judgment and client knowledge. To fill in the questionnaire, input data from the auditor is required, so the robot sends out a survey that needs to be filled in by an auditor. As you can imagine, this leads to frustration for the auditor and the process is not more efficient than when an auditor would fill in the questionnaire directly.

An example of an efficient RPA solution would be to let the robot fill in the client-specific questionnaire based on the information that was already included in the prior year's audit file or publicly available information. In this way, this will save the audit team time. Also, no interaction with the audit team is needed, which brings us to the next lesson learned.

2: Minimize interaction with the audit team

When setting up the RPA solution, think about the concept of the RPA solution. Where can the RPA solution retrieve the input that is needed for the process? Try to minimize the interaction with the audit team in this step. Too much interaction between the RPA solution and the audit team could result in:

  • A higher turnaround time – Considering one of the advantages of RPA is that the solution runs 24/7, the most important argument to limit the interaction with the audit team is that this will lead to a higher turnaround time. Due to working hours, leave or just a busy calendar, the reply to the RPA solution could be later than appreciated.
  • Dissatisfaction of audit team – Extensive interaction with the audit team can lead to dissatisfaction of the audit team. Receiving emails from a robot is easier perceived as a ‘spammy’ message. Because there is no human behind the emails, this might also result in the message being perceived as irrelevant or low priority.

Only when interaction with the audit team is required for a certain decision to be made, the interaction can take place. Preferably limit the number of interactions to one certain moment, for example at the start of the process. Include all relevant decisions in this first and only moment of contact. Also, predefine what should happen if the audit team does not reply promptly.

Another option would be to let the entire process be executed by the audit team. For example by setting up a template to extract data from forms automatically. The template is set up centrally, but the automated solution will be run by the audit team themselves, therefore directly covering the needs of the auditor.

3: Involve stakeholders early

When developing the RPA concept, identify and include the relevant stakeholders. Involve them as early as possible to set up the business case, structure, and concept. Stakeholders that are required to be included from an early stage are audit team members from different levels, a methodology partner, an IT specialist, an implementation manager, a project manager, and a learning manager.

One of the key stakeholders is the auditor. Including audit knowledge in the process of setting up the concept of the RPA solution is key for a successful product. Including an auditor in the process at an early stage will lead to a better fit of the solution to the needs of the audit team. This will ultimately lead to higher usage and adoption.

Including the industry knowledge is not only preferred for a successful setup but it is also needed from a regulatory perspective. For RPA solutions to be used during an audit, the solution needs to meet the auditing standards and other regulatory requirements. Include the relevant stakeholder(s) that can determine whether the solution is robust enough to use in an audit, preferably from the very start.

4: Responsibilities must be clear

Even though involving all stakeholders will lead to a better RPA solution, the risk arises that no one will have a complete overview of the product that is being developed. A recommended approach would be to assign a product owner that is responsible for the final product. The product owner will work closely together with the project manager that is overseeing the progress of the product.

One of the risks of setting up an RPA solution is the supposition that 'the robot will fix it’. However, auditors should always apply their professional judgment and check whether the work has been performed correctly. When brainstorming the concept of the RPA solution, do not forget the review process. Make sure the reviewers are well accommodated in performing their review, this too will lead to a higher appreciation for the RPA solution and ultimately higher usage and adoption.

5: Map out all exceptions

Once the basic constructs of the RPA solution are set up, map out the exceptions that might occur in the process. Every RPA solution has a ‘happy flow’ (the perfect flow that you wish every audit file or audit procedure would follow), however, the reality is normally more rigid. Involve auditors from multiple levels, disciplines, and industries to find out how files are built up or how certain procedures are being performed. Let’s take the documentation of auditing the account operating expenses as an example. There are at least 50 ways how you can document that procedure, and all 50 are correct. Brainstorm with the group on the most common exceptions and how the RPA solution should address those. Remember that great minds do not necessarily think alike. By involving more opinions, the RPA solution will be better deployable and therefore more useful.

6: Test, test, test

Even though the concept of the RPA solution is perfect, it’s important to test the RPA solution extensively. Performing tests will make the RPA solution more robust. It’s important to test out how the RPA solution will respond in different scenarios. When testing, you will come across multiple occasions where the RPA solution is not working as you were expecting. Better address this at the testing stage, update the design, and test again. No pain, no gain.

7: Celebrate and communicate successes

The first time the robot runs successfully? Communicate. The first time the robot handles a specific exception? Communicate. Milestone reached? Communicate. A lack of communication is one of the main reasons auditors do not embrace change. Especially with tools like an RPA solution, which can sometimes feel like a ‘black box’, transparency is key. This also means being transparent about situations in which the robot did not work perfectly. Being honest about what an RPA solution can do in an audit, but also about what it cannot do, will add trust to the RPA solution. Also, it will make the role division between the robot and the auditor clearer: always review the output of the work performed by your digital colleagues.

Altogether, there is huge potential for RPA in audit as long as it’s set up and implemented in the right way. I see the implementation of RPA in audit as a necessary step to transform the audit and make the profession more sustainable for the future.

Back to Blog

Related Articles

The Voice of the Auditor

Auditing is subject to constant change, regulation, and technological advancements. The sweeping...

Everything you need to know about DataSnipper v4.1

More automation, friendly interface, and overall efficiency. DataSnipper version 4.1 offers a...

Financial Statement Suite Add-on v4.1

In version 4.1, the Financial Statement Suite add-on delivers more automation with a fresh...