The digital landscape is rather complicated these days. Integrating systems of different ages, technologies and suppliers sometimes seems like an impossible task where each solution only raises new problems. To take the next step, you need a simple and effective approach. That approach is RPA.
But what exactly is RPA? And how do you use it? In this guide we answer the most important questions about RPA.
First, let's talk about that term. Robotic Process Automation does not involve a robot. At least, not the way you see it in films or in images of car factories. Our robots are software robots. And yet they are 'real' robots, because they take over work previously done by humans. RPA robots operate the same interfaces as human workers. They fill in forms, read e-mails and click on 'save'. They just do it all much faster and more accurately than a human could ever do it.
Software robots are much simpler and more flexible than their hardware counterparts and can therefore be trained and deployed quickly, easily and at low cost. And because they're made to imitate people's work, they work with your existing systems. You read that well: to automate processes with RPA, you don't have to change anything at all in your current IT.
Setting up and implementing RPA can be done very quickly compared to adapting the systems themselves. But it is important to take the right steps, in the right order.
Before you can automate a process, you need to know what that process looks like. Even if you already have process descriptions, it is important to validate them again. Processes develop quickly and the 'paper reality' sometimes wants to lag behind the real world. In this analysis phase you can also see where you can make the most profit with RPA. In this phase it is a good idea to have an external consultant take a look at your processes. The perspective of an outsider often provides new insights.
The software robots run on their own system. Technical integration with your back office is not necessary, because RPA communicates with your systems via the same interfaces as an employee. This ensures that the technical set-up of RPA is usually not a major challenge. The most important part of the work is setting up and testing the interactions with your systems. Depending on the software you have chosen, you do this with code, in a low code or no-code environment, with or without help from machine learning.
There are two sides to the rollout of RPA. Of course, especially in the beginning, you want to keep your finger on the pulse when it comes to the processes themselves and the performance of your robots. But it is also very important to help your employees redesign their work. Their working day is drastically changed by RPA and leadership is required to ensure that these changes are properly incorporated into your organization. Only then will you get the most out of RPA.
In this last step, RPA's main pitfall is also to be found. It is easy to concentrate on the technique and lose sight of the human side. But a lot changes. Not just for employees. Also for managers. They suddenly have to look at their processes with a data view, where they first and foremost directed the execution.
Another pitfall is that companies want to do too much at once. A 'waterfall project' is then set up, that wants to realize a lot of things at the same time, with the risk that interpretations and implementations of processes turn out to be wrong in the end. It is therefore best to implement RPA in small steps. Automate something, see if it works and then take the next step. Fail fast, fail often. You learn from each step, and a quick success can help you get the rest of the organization on board. In this way you combine progressive insight with growing enthusiasm.
In companies and departments with long, complex administrative processes, the big advantage of RPA is immediately apparent: all manual actions related to entering, copying and checking data and a lot of actions related to assessing and interpreting documents and forms simply disappear from everyday work.
This is a huge step forward for administrative departments, accountants and HR departments. But there are also great opportunities for RPA in logistics, a sector that still uses paper forms and manual input a lot.
The advantage is clear: less work, fewer people, less costs. But that's a limited view. Automating repetitive work can also produce other, nicer results.
Less repetitive administrative work means that your people have more time for personal contact with customers. This allows you to gather valuable knowledge about the wishes, questions and complaints of customers.
Cleaning up and improving outdated and incorrect information is a huge job, for which there is always too little time. Now that you have largely robotized your primary process, you finally get the chance to work on your data quality.
The more you know, the better you can control your processes and the faster you can innovate. So use the time you gain to supplement your data with data that will help your process further. The processes you set up for this purpose can, of course, eventually be (partially) automated.
Man is not made to retype forms, but to be creative, to work together and to help other people. The working day of your employees becomes much more interesting when the repetitive actions disappear. And satisfied employees ensure satisfied customers.
Files with something special are often 'parked' until someone has time to look at them. In the meantime, a customer or colleague has to wait. Thanks to RPA, your people are free to pick up the exceptions without delay.
Whether a software robot does an operation for the first time or for the millionth time, it is done with 100% precision. Also on Monday mornings and Friday afternoons. And on weekends. And at night.
Big hustle and bustle? Then you need to find temporary staff. And on the job. And equipped with workstations, computers, accounts and access cards. Software robots can be copied at low cost and without the intervention of an employment agency.
In customer contact, it is nice to be able to say with confidence that, for example, something has been arranged 'within one working day'. With RPA you are not dependent on the availability of employees to fulfil such a promise, but you know exactly what your turnaround times are.
RPA software produces data about everything it does. This gives you deep insight into your processes and gives you the opportunity to make further optimizations. In addition, the collected data can serve as input for your machine learning models. In a sense, RPA is only the beginning of your process automation. A concrete example of this is how an e-health provider lets a machine learning model read the log files and, based on this, tells customers via a chatbot how long it will take to process their files.
A lot of data, such as addresses and phone numbers, exist in multiple systems. Often manual work is needed to keep all that consistent. In practice, your employees often do not get around to it. Automating this via the APIs of different products is often expensive and time consuming. You are also dependent on software suppliers and which APIs they want to offer. By automating this with RPA you improve your data consistency, without extra work and without long development processes.
In our daily work we come across two important misconceptions about RPA. We'd like to put them right here.
The first misconception is that RPA is some sort of interim solution, that you use until the 'real' integrations are ready. The thought is understandable, with a solution that uses existing software, through existing interfaces, rather than intervening in the infrastructure. But RPA dramatically improves the way you work with fairly little effort. RPA is also extremely scalable, so it grows with your organization . Because RPA generates a lot of knowledge about your way of working, you get more and more grip on your processes. This also helps you in the development of new systems. You can skip the long and complicated integration projects.
"Robots take over the work and man becomes superfluous. How many times have we heard that. And in theory, it's possible to introduce RPA and then move on with far fewer people. In the short term this will give you a cost advantage, but in the long term it is a losing strategy because you stop innovating. A much smarter approach is to reduce costs or keep them the same while using your people more effectively, for data quality or more intensive customer contact or to further improve processes.
Many software providers are active in the RPA market. Every year, suppliers are added and every year a few disappear. However, the top 10 is fairly stable. This is him according to Datamation:
Gartner research divides these providers into four quadrants: leaders, visionaries, challengers and niche players and identifies UiPath, Blue Prism and Automation Anywhere as the leaders of the field, with UiPath leading the field in 2019 as the most advanced, but also the most operationally deployable. Other providers focus on a specific niche, are even less far in their development or do not yet sufficiently combine vision with practical applicability.
At Node1 we work with UiPath, because it is a very flexible, hyper-low code platform that can be applied to all types of processes. With UiPath, we can always choose the approach that works best for a specific client at a specific time. The self-learning intelligence of UiPath can also help with process discovery, discovery and optimization of processes that have not yet been mapped out.
Processes become fully measurable with RPA. Agents on the computers where the work is done look at it, even with non-automated parts of the process. In this way, the software can advise itself on further automation.
So RPA is only the first step. You automate the processes as they are at that moment, possibly use the APIs that are already there, and in this way make a considerable profit compared to the old situation.
The data you collect gives you a grip on your processes. So much grip, that you can think about new ways to unlock those processes. If, for example, applying for insurance in your back office is fast and completely automatic, you can offer such a service via a chatbot, a voice interface or a mobile app.
The next step is what we call 'full integration': all processes run optimally, all data is consistent across all systems and your employees are only concerned with customer contact, monitoring and dealing with exceptions. Because all systems are integrated, you see more and more opportunities to optimize or launch new processes. Your organization changes with it. Self-managing teams have their own mandate, budget and KPI's and work independently on process improvements. There is no tightly defined roadmap , but you work through a continuous cycle of 'plan, do, check, act' on the improvements that bring you the most profit.
Every step you take provides new knowledge and insights and brings you closer to hyperautomation: the total automation of all your business processes. Is this the future? Is it an unattainable dream? Let's fall somewhere in between and keep it at 'ambition'. In any case, the whole idea of hyperautomation stands or falls with data. The data you collect about your processes tells you what the next step should be. An intelligent RPA solution provides you with that data and helps you discover, analyze and optimize processes. That makes RPA the first step towards a more efficient, more interesting and more customer-friendly future.
I understand.
