In the last couple of years, we could hear the term RPA, or its expanded version, Robotic Process Automation, all around. It sounds very promising, especially in the era of robots and automation. It brings to mind automated facilities, smart houses, self-driving cars, voice assistants and rockets returning back to a spot on the Earth. Such innovations excite us. But what about the word process? It doesn’t sound exciting at all.
What really is RPA?
RPA is a kind of automation, concerning only the digital domain and the inside of computers. Furthermore, it is applicable only to very specific parts of software which have a user interface for your employees. Such automation only supports processes which take place in such an environment. It is much simpler than we can potentially imagine.
Let’s think of some tasks your office workers do on their computers every day: manual, repetitive and mundane. No matter the sector, role or level of the employee, at the end of the day, they always come to the point where they have to do this kind of tasks. We all do, but it definitely doesn’t make us happy. After a while, we can find that about 20-30% of our time at work includes the above scenario. Soon we discover that some roles in our company are mostly repetitive and manual.
If we can automate a vacuum cleaner and a lawn mower, why not do it with this kind of tasks?
Where can you implement process automation?
You can get all excited about automating processes and try to automate everything. However, you should be careful here, since not every process can be performed by a robot. Tasks that can be addressed with RPA need to meet certain requirements, such as:
- being rule-based – a robot is not smart at all and you need to tell it what the rules to follow are
- having a very low number of exceptions to rules – the environment must be stable, because the robot will not be prepared to handle exceptions to rules
- using structured data – a robot can’t understand open text, so you need to be specific and provide it with a standardised data structure
- being repetitive – the frequency of performed processes must be high to ensure return on investment
If your process meets these requirements, robots can help with various tasks, such as:
- processing and moving data between systems
- aggregating data from multiple systems and generating reports
- detecting specific values across systems and sending notifications
- issuing invoices and emailing them to customers + other financial operations
- processing bill scans with OCR and creating a summary of expenses
- scraping information from the internet and using it to build databases
- basic communication with employees or customers
- some phases of the recruitment process
This list doesn’t exhaust the possibilities, but let’s stop here and consider it just as an inspiration. You know your company best and can surely identify similar tasks.
The return on investment will be achievable only if the process is well analyzed, measured and implemented (more about ROI will be presented in a future post). The investment in RPA implementation is low compared to software development services.
The first benefit we can think of is cost reduction, as we do not overload employees with such demotivating tasks. For some employees it will be a threat, since their roles will be no longer needed. For others it is an opportunity, as they will finally have the time to focus on creative, more valuable and irreplaceable work.
There are several more non-financial benefits, and some of them can reduce the costs in the aftermath. The robot can work 24/7 without breaks, holidays and sick leave. It works even 50 times faster. It is a highly scalable solution – if you want to double the speed, just make a copy. It is not susceptible to human errors (the error rate can be lowered to 0 in many cases), which helps to improve the quality of your processes and their results. In the end, it encourages standardization of processes, data and workflows.