We often say that it is possible to automate any rule-based process that runs on a computer. If a process does not require creativity or the human factor, Robotic Process Automation (RPA) technologies enable automating it. Every business can easily find such process internally. It may thus seem that the possibilities of RPA are limitless. However, RPA is not always the best option.
Our experience has helped us to create a simple framework for identifying processes that are indeed worth automating. Every process can be easily evaluated, and a decision can be made by following a few simple rules.
How to identify RPA opportunities
What is the easiest way to identify processes that are worth automating? The most important point here is a good understanding of the process. This is obvious and you might think there is no need to point that out. However, organizations often try to jump into the RPA train blindly without ever understanding their internal processes. We constantly encounter situations, where our clients have no internal resources responsible for the standardization, improvement and control of their processes. Processes tend to evolve into something completely different in such environments. We sometimes joke about how 5 people who do the same task will eventually create 10 different ways of doing it. Even though this is usually filled with irony, it is also not far from the truth.
In addition to that, the decision to automate processes is usually made by top management. But they rarely know more than the theory behind how their processes “should” work, instead of how they are performed in reality. That is why we always encourage our clients to standardize and optimize their processes first. We offer our help, if they need it. This ends up being a very positive side-effect of the RPA initiative. Business processes are not only automated, but also improved to an extent where they are as effective as possible. That is also the only way its worth to do automation.
Understand, understand and understand
But, like we said, improvement and standardization are not possible without a general understanding of the process. The ability to see the big picture and understand what internal processes contribute to it is essential. If it is not there, the client usually encounters one of the following two problems:
- wrong processes are being automated, or
- too much time is wasted trying to reproduce the correct flow of the process.
For example, one of the classic use cases for RPA is processing customer orders (a.k.a. Order To Cash or OTC). Automating this process can generate amazing returns because it:
- Has high volumes;
- Has high frequencies;
- A lot of manual work is needed;
- The work is rule-based and standard in most cases;
- Customers are more satisfied, when their orders are processed quickly.
However, in a B2B environment customer orders are often very different and each customer follows their own way of ordering goods or services – whether it is a plain text email, a call on the phone, a PDF document or something else. From the point of view of a software robot, each different type of order is a separate process that needs to be automated. And so, while it is obvious that automating the whole OTC process is beneficial to the business, having different templates for each customer increases the development cost of such a solution. It thus may be wise to put a process like that to the side for a while and focus on something else.
Is it worth automating?
Once again – any rule-based process can be automated. It the question is “should it be automated?”. Our know-how and the good practices in the RPA market allowed us to create a simple framework that can be used to evaluate each process. It not only helps to identify the processes that are worth automating, but also helps set priorities and decide what should be automated first.
The criteria are as follows (written as questions for simplicity):
- Is there a high volume of operations to process?
- Is there a high frequency of tasks?
- Is the possibility of human error high?
- Is the process predictable?
- Can exceptions and their handling be described in rules?
- Is the task manual and monotonous?
- Does the task have to be performed quickly or during non-working hours?
- Does the task involve sensitive data?
- Does the task use structured data?
- Is the process mature and stable?
- Can savings be measured?
The more positive answers to these questions, the higher the benefit of automating such a process. That is not only because automating it may reduce cost, but also because of improved service or product quality and processing speed, lower rework-related costs, higher data safety, better customer and employee satisfaction, etc. If there are multiple candidate processes, these criteria can also be used to form a heat-map for priorities and also a road-map for the whole RPA initiative. However, before attempting to do that, an organization should have clarified their goals for the RPA initiative.
Correct goal setting is a topic for a whole separate article, which we will definitely post. Nevertheless, it is necessary to note that it is almost impossible to set priorities correctly without a clear long-term goal. Of course, if an organization has some urgent issues or some special processes that absolutely need improvement and automation, it makes sense to start there. But what if your goal is not just cost savings? What if you need to make sure that:
- tasks are performed and are not forgotten if “the responsible person is gone” (e.g. backups of files and documents or preparing an important periodic report);
- tasks are performed outside of your normal working hours (e.g. preparing an inventory and production report and sending it to procurement before they come to work);
- tasks are performed right away without any delay (e.g. reacting to customer inquiries and processing them quickly, instead of “when the account manager has time”).
These and similar processes are related to a secondary and a lot less marketed business value of RPA solutions. It is harder to measure, because these benefits do not result in headcount reductions. But if that is more important than, for example, handling a high volume of transactions, the priorities in the heatmap should be set accordingly.
The criteria above can definitely be used as a basis for identifying RPA opportunities. But you should never believe that the same shoe will fit every foot. Every business is unique and special in their own way with different internal processes, long-term goals and short-term priorities. Every RPA journey is thus a new adventure with unexpected results – both nice and not so much.
We are here to help you make the decisions. If you would like to know more, contact us and we will take you on an amazing ride through the RPA journey.