Last year we talked about some common pitfalls of robotic process automation, and made some suggestions for avoiding them. For instance, not setting clear enough objectives to guide your automation strategy, disproportionate reliance on the IT department to the expense of unsatisfactory involvement of the business units, or insufficient testing of software robots prior to launching — are just some of the things that might go wrong, and prevent you from leveraging RPA at its full potential.
Get ready for robots. Why planning makes the difference between success and disappointment is a recent EY report which raises attention to the fact that 30% to 50% of businesses fail to achieve the promised reduction of human error, increased productivity, improved regulatory compliance, customer satisfaction, or decreased costs. UiPath emphasises that technical challenges, as well as defective strategic thought, e.g., lacking a “bird’s eye view” over the implementation process, account for a good part of the failed automation ventures.
How advanced is RPA in Australia?
Companies are becoming more and more aware of the benefits robotic process automation can bring about, e.g., improved productivity, accuracy, operational efficiency, or customer service.
“54% of Australian organisations have reported cost savings of 15% or more from implementing robotic process automation during the past two years, according to Accenture.”
So Australia nicely illustrates the global automation-oriented trend. More concretely, according to a Telsyte survey, by 2020, the robotic process automation market is expected to reach $870 million in the area, which represents a 45% growth compared with 2016.
Have you started implementation yet? If yes, the next step would be scaling up to an enterprise wide automation formula. As we’ve said before, scaling is indeed a ‘must’ in order to make the most out of automation. Just don’t forget to keep an eye open for what might go wrong, so that you can make the most of the benefits of robotic process automation.
Top robotic process automation (RPA) mistakes to avoid in 2019
1. Narrow focus on cost reductions, without a holistic, longer term perspective
RPA is an end to end change program. Software robots work best when part of a ‘mixed strategy’, combining the right RPA tools, process engineering and human expertise. The larger picture should also include other technological developments, like analytics, natural language processing and AI, in a complex journey towards digital transformation. A means — ends hierarchy, increasingly granular top down, is the first necessary ingredient.
The recommended way to circumvent this costly mistake is rather simple: make sure you start with a broadly focused, long term business case, specific yet flexible, allowing the inclusion of innovations that form a coherent pattern.
2. Expecting enterprise wide scaling to come naturally after the proof-of-concept
There’s no denial that the pilot phase is necessary when implementing RPA. It is the perfect occasion to learn by doing in a sandbox environment, which allows the possibility to optimise the software robots without risks. However, it is certainly not also sufficient for expanding automation across the whole company. Wishful thinking that once the concept has been proven for one process, it can simply be generalized for all processes, is among the notorious RPA mistakes.
The good news is that it can be avoided with a certain degree of realistic forethought, and with a coherent, centralised vision of the whole automation journey. A Centre of Excellence might be what is needed to ensure a holistic perspective.
Waste makes haste, so your patience should handle a journey duration of nine to sixteen months, during which bots are implemented to automate various well-chosen business processes. This is the long and winding road towards sustainable RPA.
3. Not leveraging assistance from the IT department
As our partners from UiPath put it, the best case scenario for RPA implementation is at the junction point of business (a generic name for operations functions) and IT. Both are vital players in robotic process automation viewed as a team sport. It’s best if you ‘render unto IT the things that are intrinsically IT-related’: automation codes and their impact on issues like scalability, security, or fault tolerance.
In so doing, you allow the people from operations to zero in on managing productivity outcomes, by making use of their in-depth process understanding.
4. Leaving everything on the shoulders of the IT team
A coherent, long term plan requires integration of the individual software robots by means of strong administrative control. Moreover, as the automation journey progresses, it is likely to bring about considerable changes in the process structure of the company.
Bots need retraining, and employees who delegate part of their work to bots need re-skilling in order to take over higher-value jobs. People in charge with the operations functions can handle these aspects much more efficiently. In conclusion, division of labour should be taken seriously and applied in practice in order to avoid this robotic process automation mistake.