Is your most fancied FP&A RPA initiative not delivering the anticipated benefits? You are not alone.

A couple of years ago Deloitte conducted a survey of 400 international companies. According to the survey, 63% of the respondents failed to meet delivery deadlines for RPA projects. Even for the ones that saw the light of the day, delayed bot implementations increased the costs pulling down the ROI and in hindsight making it a bad investment. Another study by EY found 30% to 50% of initial RPA projects to fail.

So, what are the most common pitfalls? And more importantly, how can you sidestep them and increase your odds of a successful RPA implementation?

1. Not selecting the right process for RPA. A key thing to remember is that despite the word “process” in RPA, normally, not the entire process is automated but the ‘tasks’ within this process. It is important to ensure in advance that you select processes that have a significant component of highly structured tasks that are rules-based and have huge volumes. Otherwise, the benefits will not justify the investments. A potential RPA candidate should have all three parameters (highly structured, rule-based and sizable tasks). In my experience, while the first candidate for automation is chosen with much care, on its success, there is a rush to onboard a lot of processes and this is where it can break. The most common and quick return areas are P2P followed by regular reporting and analytical tasks.

2. Not doing an in-depth and end-to-end assessment of the target process beforehand. A common mistake many companies stumble upon is missing doing a 360-degree assessment at the process mapping stage.  For instance, when the month-end reporting process goes away to the shared services / outsourcing organization, the core finance will still like to have an inhouse “coordinator" type of a person to ensure hygiene, do the right classification / codification, interpret the business needs of the expenses, review the variance reporting comments etc. The cost of this person is missed in the business case. This hidden cost comes up later and creates disillusionment.

3. Automating a badly designed process. Speeding up a broken process does not help. Before implementing RPA, step back and use the opportunity to relook at the entire process to identify business needs for each task, eliminate unnecessary steps and duplications, standardize as far as possible and make sure the internal controls and checks are in place. Only once you are sure the newly defined process is lean and robust, move to automation.

4. Not understanding the full costs of an RPA initiative. Even if a company takes care of the second point above, many organizations don’t realize that RPA requires an ongoing investment of time and resources. It is not a one-off project. There are multiple components of the costs involved in an RPA solution. Implementers in their hurry may take cognizance of only licensing costs and initial implementation efforts. What is also important is to understand and consider costs of maintenance which would be periodically required. Chart field changes, business reorgs necessitating cost center changes, user interface changes in the applications, security patch updates, any other changes in the underlying programs can disrupt the RPA and would require constant monitoring and updates. Do not undertake an RPA initiative just under peer pressure. You will see a lot of finance guys chest-thumping on social media about 90% automation of their processes etc. It is important to do your own assessment before taking the leap.

5. Choosing customized in-house solutions. In my experience, the best payback is in using commonly available tools in the marketplace. Building an own in-house IT team to design specialized local solutions bottoms-up for finance processes can be very expensive and can also face issues in scalability and flexibility in the future.

6. Not having the right structure for RPA ownership and governance. Ideally, an organization-wide RPA CoE structure across all functions is best led by a small team of actual business users and finance. This ensures a high probability of sustained success. IT should be only the enabler and should not be expected to be taking a leading role in designing and maintaining the programs.

7. Talent Management issues. A lot of projects fail or do not generate adequate payback because the people involved are not sharing their inputs because they fear that an RPA implementation may result in a job loss for them. A future career roadmap should be discussed with those who are affected by the successful implementation. This will help organizations to deal with employees’ fear and avoid possible sabotage.

Bottom line: in order to implement a successful RPA initiative, you need to select carefully, take a long-term view of both costs and benefits and get support from your team to ensure that your FP&A RPA implementation is your Repeatedly Proven Advantage.