There are multiple important things to note when considering going through the RPA journey and trying to identify processes for automation – for example, if a process is rule-based and has high frequency, it might be a good candidate for automation. But some organizations stray from the path to success due to high expectations and sometimes – certain myths and misconceptions. This can be caused by lack of general understanding about what RPA is and how to use it.
You can find a short explanation on what RPA is and why it is useful for business in our earlier blog post.
That is why it is crucial to clarify some things that are often misunderstood by business users. The myths listed below and the explanations added to each of them should help everyone interested in RPA get a better idea of how to embark on a successful journey. But it is important that this information is not only read by the person responsible for RPA in an organization. The whole RPA team, the top management and all of the employees should also be introduced to it, because otherwise misplaced hopes and fears may appear that can negatively affect the reputation of RPA within the organization. Without any further ado, let us dispel the most common RPA-related myths.
1. Robots will replace humans
This myth is probably more popular among lower tier employees who perform tasks that are fit for automation. Top management normally does not think about it (unless they actually expect to reduce headcount via automation). People tend to be afraid of losing their jobs when their main tasks and responsibilities are actually routine manual work that does not require a lot of cognitive functionalities. This is also related to some other popular mini-myths that feed this major myth:
- Physical robots (humanoids) are used for process automation;
- Every process can be automated;
- RPA works around the clock with no supervision, support and maintenance.
Some of these ideas may originate from modern science fiction, but these are all myths. RPA solutions are not created to replace humans. These solutions are here to automate specific tasks in order to achieve very specific goals (e.g. process large volumes of operations, eliminate errors, improve data security, etc.). This does not only result in no job losses, but also significantly improves employee satisfaction, because once they no longer have to do these mundane tasks, their time can be used for more interesting activities that add greater value for the business. For example, finance and accounting specialists can do less data entry, which is often boring, and more data analysis or higher quality assistance to top management. Sales representatives no longer have to work on preparing contracts, update customer data in CRM systems or prepare sales reports, but instead they can focus on finding and retaining new customers, improving customer relations, etc.
RPA is not a perfect tool that can automate everything. As we have already discussed in an earlier blog post, a process can be automated, if it is rule-based. McKinsey estimates that less than 5% of all jobs can be fully automated at this point. But then a further 60% involve activities, around 30% of which can be automated. This means there are very few people that could actually be replaced by robots in their jobs. But there are quite a few that can have their jobs changed and improved by automating processes. And this does not even take into account the new job roles that are created by the use of RPA software. RPA solutions do not appear from nowhere, robots do not create other robots. And only some implementations are advanced enough to ensure that robots administer and supervise the work of other robots. Normally these tasks require humans. This obviously means the need for specialists in roles that did not exist 5 years ago. While that may seem scary to some, a job like that tends to be much more interesting and also valuable for the business (thus more satisfactory for the employee as well) than jobs that can actually be automated.
2. Process automation = Process improvement
RPA solutions are designed to automate processes. But if there is one thing that we tell each of our clients, it is that every process needs to be reviewed and improved to be efficient prior to attempting automation. Process improvement does not happen automatically. And process automation does not equal process improvement. To some extent, RPA solutions can make processes more efficient, because it increases the speed of operations and eliminates human errors. But if a process is inefficient, has faulty instructions or produce inconsistent or unclear results, the “robot” will simply do the same – but much faster. And to top that off, the biggest issue here is the fact that the robot – unlike a human – will not understand that it made a mistake, if it did everything by the book, but the book (a.k.a. the instructions) was wrong.
There is one thing to note here, however. And that is the fact that normally, when preparing for process automation, process descriptions and instructions are prepared (if they are not in place already) and reviewed thoroughly. This helps identify improvement opportunities. We thus always tell our clients – and yes, we are repeating ourselves on purpose here – that process review comes before automation. We can help find solutions that improve processes and make them as efficient as possible. And we are more than willing to do that, because once that is done, process automation is much smoother and the resulting RPA solutions are more stable.
One can thus state that process automation is one of the tools used for process improvement. RPA solutions improve processes, but the effect is completely indirect. It is there simply because RPA deployment experts and consultants who actually know their business will make sure that processes are reviewed and improved prior to automation. But it does not happen automatically.
3. RPA is only good for cost reductions
Automating processes can definitely bring significant benefits cost-wise. And it does not only come in the form of headcount reductions – costs can also be reduced due to higher product quality and reduced error rates (meaning lower re-work and warranty costs), greater agility and faster processing (meaning lower costs related to seasonality, and re-training after organizational changes), etc. However, this is by far not always the main reason to go on the RPA journey. An earlier blog post already described the benefits a business can achieve via automation. But repeating is the mother of learning, so here is what process automation means:
- Greater productivity due to the capacity to work around the clock,
- Increased efficiency, because tasks are initiated instantly and are completed much faster with no human errors,
- Improved customer experience via high quality service,
- Higher quality of products and services,
- Standard, predictable and easy-to-audit processes, as they are always performed in the same rule-based manner,
- The ability to scale easily as additional robots are simple and cheap to deploy when needed, which leads to human employees being able to focus on more value-adding activities,
- Smoother control of business cycles due to advanced robot and workload management,
- Improved employee satisfaction because of more interesting tasks and the opportunities to learn new things,
and a lot of other benefits that can sometimes bring much more value to the business than cost reductions.
4. Only large companies can afford RPA
Business process automation obviously has its price. First of all, there are RPA software licences that need to be purchased. Then there are some additional cost related to hardware and software required to create the necessary environment for the robots to run (such as dedicated computers or virtual machines, operating system licences, robot user accounts and their email, MS Office and other software licences, etc.). There are also likely to be some consultancy required in the beginning, which comes at some extra cost. And then finally, depending on the path an organization decides to take (whether its training internal employees, hiring new staff or outsourcing the services from a service provider), there are the development and maintenance costs. All of this can make RPA look like an expensive endeavor that only large organizations can afford.
The truth is that RPA implementation comes with an investment that is insignificant, when compared to the usual alternatives – business process management systems (BPMS), business process outsourcing or establishing cost centers in low-cost countries. A software robot may cost anywhere between 1/9 and 1/3 of an annual salary of an FTE, depending on the choice of software and the deployment service provider. But at that cost, a robot will work around the clock with no breaks, sick leaves, weekends, holidays, etc. It is thus calculated that a single minute of an RPA solution running should be valued as roughly 15-20 minutes of human time.
However, not all RPA initiatives are that successful. There are quite a few that fail completely. And then there are some that only manage to reach an ROI of 30% during the first year. This huge gap between the results of a successful implementation and one that is not that great is generally determined by how ready an organization is for RPA. Taking the time to plan things and do the homework before starting leads to faster deployment, greater accuracy and less errors and rework. As we have already mentioned in an earlier post, in order to truly enjoy the benefits of RPA, it is an absolute must to have a good understanding of internal business processes, set clear long-term goals and short-term priorities for the overall RPA journey. If it is planned correctly, the costs of RPA will be insignificant, when compared to the returns it brings.
5. RPA is only applicable to some industries
There is a very popular misconception that processes can only be automated in companies that operate in some very specific fields, such as IT, financial services or companies that rely heavily on multiple different software packages. But as we have already noted almost any rule-based process that runs on a computer can be automated. And pretty much any organization has processes like that, regardless of its size or industry.
This myth is likely fed by the fact that the early adopters of RPA for internal process automation were banks and other financial institutions. While RPA has definitely proven its worth in this field, more and more organizations from different industries have been reaping the fruits of process automation during the past few years. This includes companies in IT, healthcare, manufacturing, shipping and logistics, retail and wholesale etc. Some of these industries have very specific processes that can be automated (e.g. claims processing in insurance, loans processing in banks, handling patient data in healthcare, etc.). But most of them actually have processes that apply to multiple industries, such as issuing invoices, calculating salaries and wages, generating periodic reports, processing customer orders, etc.
This has also been proven by our own experience. Our clients come from very different fields and industries, including engineering and consultancy, manufacturing, retail, e-commerce, etc. And when we do work with these companies, we do not just automate processes in a single department. We generally go company-wide and automate all sorts of processes in various departments, such as finance and accounting, IT, sales, procurement, logistics, production, QA, etc. We can thus safely say that RPA can bring huge benefits and competitive advantages to players in pretty much every industry there is.
6. Processes can be automated without involving IT
Most RPA software vendors brag about how anyone can use their products. RPA software is generally created in a way to seem like it is very user-friendly and easy to use, so that people who have no programming experience could use it with no help from the IT. It is presumably enough to understand the process and anybody from within the business can automate it. This is true to some extent – it is indeed quite easy to automate a simple process. But in order to create a solution that lasts and is stable, it needs to at least follow certain rules and practices:
- Advanced exception handling must be in place to make sure it does not break down or cause errors;
- The solutions must be as universal as possible to retain stability when systems and other software changes;
- Sensitive data must be encrypted and secure;
- The most efficient means and procedures must be used for data input, processing and output, IT infrastructure, etc.;
- Separate automation flows must be created in a way that it is easy to control, manage and adjust, when there are changes in the IT infrastructure and systems;
- It must be clear and simple for the end users to use these solutions and handle situations not only when everything works fine, but also when things go wrong.
In addition to that, most business processes that are good candidates for automation, are obviously much more complex, than simply opening an MS Excel spreadsheet and typing in a few figures. Thus, experience and understanding of IT infrastructure, best practices for automation and the ability to employ existing assets and resources within the organization proves very useful. Otherwise, when “anyone” or “everyone” is developing RPA solutions, it is very likely that this beautiful initiative ends prematurely, without even reaching any actual milestones or benefits for the business. Or it might end up becoming a huge headache when external consultants have to be hired to handle the mess, instead of developing quality solutions right away.
On the other hand, it is in fact true that RPA solutions are normally developed quickly and rather easily, without any complex system integrations or changes in the existing infrastructure. It may thus create a false impression of experienced RPA developers being able to completely avoid any contact with local IT. But that is also a very faulty practice. The internal (or outsourced) IT do not have to manage the RPA implementation or own the project. Neither do the developer team must come from the IT department. But co-operation with IT is a must, because they are in charge of all necessary hardware and software, user rights, databases, etc. And so, instead of avoiding IT, much better results are achieved when IT is involved in the RPA journey, but does not own it.
“Business users” may not exactly like this, because the idea of avoiding IT for one reason or another is usually one of the selling points for RPA. On the other hand, IT people sometimes oppose RPA, because they consider RPA to be a temporary patch instead of a long-term solution. However, with the help of good internal communication and change management, both sides of this mini-conflict usually end up very happy with the results of a successful co-operation, and the solutions created are indeed long-lasting and stable. Also, the IT department usually finds some internal manual and mundane administrative processes to automate as well, once they see how RPA works.
The myths described above are the most popular in the current business environment. But there are more. There are always people who are skeptic, afraid of changes or misguided by others, that emerge in every organization. This leads to false beliefs that, when not addressed properly, may put a halt to any improvement. That is why it is our mission to introduce RPA and the great benefits that come with it to business users in a way that is a simple and clear as possible. When both the top management and the lower tier staff understand that robots will not take their jobs, and that RPA solutions are really not that expensive, but come with benefits to companies of all shapes and sizes, then the RPA journey is smooth and the end results are amazing.