4 RPA Mistakes And How To Avoid Them

Automating its accounts payable processes enabled a Switzerland-based medical and pharmaceutical giant to achieve 90% improved processing time. It reduced manual effort by 10% and saw a return on investment within two months. So the results can be significant. 

In order to unlock the full potential of RPA, you must be aware of these pitfalls from the outset. Here’s how to avoid accidentally damaging your own efforts. 

1. Picking ‘low hanging fruit’ 

Our first common RPA mistake is for firms begin their journey with ‘low-hanging fruit processes’. Those that present a combination of high potential and low complexity. In finance, this may include automating existing Excel sheets used for tracking of accounts payable exceptions. 

While this may be the surest way to get stakeholders and executives on board, over relying on this process-by-process tactic can lead to problems down the line. Process-by-process automation often focuses on tasks that meet a certain threshold of time saved. However, these may not actually be the tasks taking up the most time in an employee’s day.

For increased overall impact, you should consider the full automation of the finance department as a whole. This allows firms to focus on exactly which time-consuming and data intensive processes occupy their employee’s time. 

For example, NHS Shared Business Services (SBS), chose full production automations within its finance and accounting service from the outset. By stepping back and creating a clear plan, NHS SBS could identify the most challenging processes, gaining employee buy-in.

In just four years, 250 of the most time-consuming tasks had been automated. Crucially, these automations interact and work together to create an entire eco-system of robots. While this is a longer-term approach, firms will experience greater bottom-line improvements in the end.   

2. Underestimating the importance of re-engineering processes

A question that clients often ask is: “should we re-engineer our processes before we implement RPA or automate the ‘as-is’ processes if they have enough potential?” My advice: do it simultaneously.  

Automating an inefficient process is not a way to fix it. Trying to apply RPA to a flawed process will never work. However, to re-engineer a process without considering RPA first could lead to missed opportunities. Neither method will leave you with the most efficient outcome. 

To automate a process, you must have intimate knowledge of how the process works at a keystroke level. Reconsidering process efficiencies while on your RPA journey will save time on having to revisit automation scripts down the road. 

3. Not involving expert business users in process automation

While IT buy-in is crucial for the success of any automation deployment, it would be a mistake to neglect those employees who want to share their own knowledge and insight during implementation. 

RPA allows business functions to take ownership of the technology without much IT knowledge. Finance professionals do not need to know how to code to create their own robots. 

Task automation and simple departmental processes therefore, should be driven bottom-up. Having an attended robot perform something at the request of an employee presents no more security risk than the employee doing it themselves. This creates enthusiastic ambassadors for automation, rather than a resistance to change that a top-down only initiative can produce. 

In reality, every organisation has automation ‘creators’ and ‘consumers’ and everything in between. That is why businesses should support these employees with specific knowledge of a process to take on RPA training and thus become citizen developers. 

Heritage Bank, Australia’s largest mutual bank, started implementing automation in 2017. In the time since, the bank supported those employees who wanted to become ‘champions’ that drive automation within the business. This approach, the bank estimates, may pave the way for citizen developers to become the main source of automation solutions within the business by 2022.

Empowering employees within financial departments to become citizen developers is crucial to driving RPA adoption, increasing internal capacity and even lending a hand to the IT department. However, to maintain control and security over such an initiative, there is a need for some basic governance principles and a supporting technology.

This includes allowing people to create automations for themselves unsupervised, but not sharing these with other employees without curation and certification from a central Centre of Excellence (COE). 

4. Not taking employees along for the ride

While certain members of the finance team may be more suitable to share their insight into RPA than others, automation roll out is most effective when employees feel like they are taken along for the ride. 

This could take the form of companywide education programmes or even offering upskilling and reskilling options to give others the opportunity to join the RPA journey. For organisations eager to automate while still remote, a number of certification programmes also exist online.

RPA is a powerful tool for finance professionals, and if implemented effectively can move finance roles away from data gathering, data entry and bookkeeping and towards advisory responsibilities where they can best use their strategic skills. 

While there is no one fool proof method for implementing RPA – every finance department faces different barriers to efficiency – certain common pitfalls should be understood and considered from the very outset of an automation journey in order to set you on a path to RPA success. 

Gavin Mee is Area Vice President Sales for the UK and Ireland at UiPath