Nice to know, before start with RPA…
It’s always nice to start by setting expectations, so we don’t waste anyone’s time and we guarantee that if you’re reading this article, is because you’re part of the right public:
- This is not a “All you need to know” article: This is a simple and light piece of text with some things that are “Nice To Know” to start with RPA.
- This is not a “subject matter expert article”: Actually we are a full inexpert team reading, learning, trying, failing and trying again. From that we have collected a lot of experiences about RPA, so this is more like the result of a learning lessons.
- This is not a “Technical Article” to know how to build or develop an RPA: This is more like a “for dummies” article that can provide from a functional perspective, some key knowledge before starting to work with an RPA initiative.
You still here? Great. We are happy to write for you! Let’s continue, with a common understanding about what is RPA:
RPA is a powerful technology that simulates almost any kind of human action over a software user interface, but almost deleting two of the most common human weaknesses: the reduced operative capacity and the high error risk; and is specially useful when there’s not an API resource available (commonly in legacy software). Born from the well known BPM (Business Process Management) and the need to manage all the business processes. Once the companies achieve the big milestone of managing them, the next step, is to optimize. Well, here is where the RPA story starts…
The sweet & the bitter
At this moment, RPA technology is trending. But, please: don’t get carried away by the full RPA lovers enthusiasm! (By the way, yes, we are RPA enthusiasts). We have to admit that every technology has its pros and cons and RPA isn’t the exception and, we don’t want to start with the idea that everything is perfect , don’t we? Let’s see:
If you think that with RPA you can solve all your problems, let me tell you it’s important to reconsider. RPA is not for all processes. For implementing an RPA project you have to do an initial assessment of your processes and also bear in mind the following two main conditions:
- The process must be repetitive for supporting manual tasks and must follow a standard pattern. It includes, the data input, all the steps of the process and the output.
- The process should not include decision making. This would be an IA. The candidate process to automate have to follow a clear set of criteria in order to determine how to behave in an specific scenario.
To implement RPA over a process you always have to guarantee that is a stable process over time. If there’s an input data manually built, the involved users have to follow strict rules in order to keep all the variables as same as are expected by the RPA. In summary, this means that before automate any process you must have a defined and highly trust business process.
Do you want to continue learning about RPA after this explanation? If your answer is yes, let me tell you more about RPA: The main advantage of RPA is that you as person can focus on more valuable tasks instead of blowing up your mind by making boring and repetitive tasks. Furthermore, the company can increase its productivity and thus, its profits.
Even when RPA was known as an alternative to automate clicks, not anymore! Now you have a lot of plugins and integrations to avoid perform the process as same as the human user does. So, besides you can automate a human process, you can also perform this in a more optimal way not only in time terms but also in resources terms.
The Do’s & Not To Do’s:
For some people maybe this could be kind of obvious, but here, we’re going to start using my favorite golden rule: Nothing is obvious. Because just as someone read this after working with RPA for 5 years, someone else just read this and it turns into a headlight to start in the RPA world. So, here we share with you some simple advices about what is a YES and what is a definitely NEVER when we’re thinking to start with a RPA:
What to do?
- Evaluate a significant number of processes that may be candidates for automation. It would be better if the processes are of different natures and even from different departments.
- Start small, with the lowest fruits so that they are representative in quick wins. Always keep in mind what is the return on investment expected (in terms of time, revenue, error reduction).
- Have as candidates processes that have a high degree of maturity and that you know they will not suffer important changes in the short term.
- The best candidates must be processes that have a positive impact on a large number of human users.
- Consider processes that due to their volume, require investing many hours of work per day. Example: Imagine a process that is completed in 2 minutes, but due to the volume it must be done at least 50 times a day and is in charge of a group of 50 people.
- Consider all actors involved in the project. It means technology/platforms business area, and work as a unit. Evaluate the degree of stability of all the actors. Maybe the business area is stable, but the platforms involved don’t have the same property.
What NOT to do?
- Try to automate processes that have not been previously improved or analyzed for automation. Example: A process that requires an approval that must be submitted in printed paper and signed by the CEO of the company.
A process of this nature must first be restructured so that it can become a good candidate for automation. - Making employees understand process automation as an opportunity for the company to reduce payroll. This will cause employees not to adopt automation solutions, causing you to invest money that in the short term will not have a clear return on investment.
- Don’t automate errors. Don’t automate bad business flows. When the analysis of a process is done with an expert and the automation solution is based on its description, the created robot may continue to make the same mistakes that had been previously identified.
- CAPTCHA. Could seems obvious, but is not. The captcha is like the bot’s kryptonite. Avoid automate process that include read captchas or detect images (or text in images).
Whenever an automation solution is being designed, it is important to analyze it end-to-end and ensure that its design seeks to minimize possible errors within the process.
Avoid implement an RPA project with a personal computer if the RPA have business purpose. Try to invest in a good infrastructure with the capability to escalate and improve the robots.
Where is RPA going?
As well as we learned something about from where RPA comes and where is now, always is Nice To Know, toward where is going:
RPA is nowadays part of a technology trend that is intended to be used broadly in years to come., and the companies worldwide are embracing the digital transformation in order to gain efficiency in the manual activities that are prone to human error. RPA is oriented to:
Intelligent automation:
RPA is one of the first steps that companies are using to be immersed in the AI process. For the companies that are already using these technologies, including Machine Learning, the combination of bots and algorithms are reducing the repetitive tasks and helping out to build a sophisticated system where the bots can feed the IT systems that are evolving and learning from the provided data.
Customer experience:
RPA is being called to reduce bottlenecks in the interaction of the customers with the service companies. By using this technology, companies around the world can interact easily and effectively with their customers, providing to them a seamless user experience in the attention of issues or creation of new requests. By implementing RPA, the customer journey definition can help to bridge gaps that are commonly corner cases that are attended by giving support from service desk, the automation here provides a minimum SLAs time response and a comprehensive customer experience.
Cross-industry support:
Due to the implementation of RPA on a cloud service, different industries can acquire and take part of this technology for their improvements and live a digital transformation. Healthcare, Banking and Insurance business lines for instance, are part of the industries which are using RPA in their processes to handle massive amounts of data and help to store it or prune it if that is the case.
So, are you ready for the next step?…