Robotic Process Automation: Do Robots Dream of a Personalized World?
September 12, 2016
To cope with the rising customer expectations of the current ‘click and buy’ generation, organizations need to offer their customers a highly personalized, context-driven experience.
<<< Start >>>
<<< End >>>
The required processing speed and complexity - combining data from various sources and devices - cannot be delivered by a human workforce alone anymore. Could Robotic Process Automation (RPA) bring organizations the level of contextual intelligence that is required to make the first step towards intelligent automation and fully customer-centered products and services? Let’s explore.
One of my first jobs as a student was in the administration department of a services organization. The organization provided repair services for household equipment like washing machines, dishwashers and vacuum cleaners. It was my job to register the daily pile of order forms – with handwritten time and material registration – into the order system. When that was done, I had to create an export of those orders, print them out and type in the data into the billing system. At the end of the day, my colleague and I would print reports from both systems and verify our actions. Then we would provide the prints to our supervisor, who would go through our numbers again. In case of corrections, we needed to repeat the whole cycle, until everything was correct. Or seemed to be correct. Back in the 90’s, this process ran 8 hours a day, with two people typing in the data and transferring data from one system to the other and one person doing the final checks.
As you can imagine this whole process was not only very repetitive and dull, it was also error-prone. Every order took approximately 5 – 10 minutes to be fully processed. I remember at some point I made the joke that I felt like a robot, doing the same thing over and over again. After just two weeks I quit this job and became a part-time postman, a role which I considered more challenging and less dull.
So here we are, 25 years later, in the era of robotics. It is fascinating to think how much has changed around us. Still, a lot of processes remain unchanged in essence. Let’s take a new look at our case of the invoicing process at the repair services organization. The process consists of the following activities:
Today, service engineers register time and material directly on their smartphone. The data is instantly sent to the back office and stored in the order system. The electronic registration of time and material already takes out the manual copying of data. Sending the data from the order system to the invoicing system could be a simple export run, but in many cases, there are still some calculations required to bridge the gap between time and material and the actual bill the customer receives. And that is where the robots come in.
By using Robotic Process Automation (RPA), we are able to create a digital workforce that can handle complete process cycles, from processing an order to producing the final bill and registering the incoming payment. This includes merging of data from different systems, simple calculations, verification of data and making corrections; 24 hours a day, seven days a week.
<<< Start >>>
"By using Robotic Process Automation (RPA), we are able to create a digital workforce"
<<< End >>>
There is no doubt that robotics lead to huge cost savings, by taking out manual repetitive work. The cost aspects alone offers enough motivation for most companies to consider investing in RPA. But what about the customer? How does the customer benefit from this all?
By introducing a robotic workforce, the error percentage in processes will dramatically drop. Robots don’t make errors; they simply follow the exact instructions that were provided to them by their human operators. These instructions are either modeled in a visual workflow or taught to the RPA tool by recording and analyzing the process as being performed by a human operator. Often a combination of these two approaches is used, to have optimal control over the processes. By reviewing every possible scenario during the test phase, the chances of errors in a process are minimal.
Naturally, customers will benefit from the lower error rate. Plus, customer requests will take less time to be processed. But I doubt if RPA alone is enough to cope with the rising customer expectations and the required level of personalization. One of the drawbacks of these kinds of systems is that they have difficulties in handling exceptions, meaning that for every unique case a separate process needs to be defined and tested. So much for personalization!
On the other hand, by taking out 90% of the repetitive regular work, RPA opens up new possibilities to handle these special cases. Experts that first had to spent most of their time on performing repetitive tasks, can now focus on unique and complex cases and on the interaction with the customer. No doubt, this is an important first step towards fully customer-centered products and services.
So what are the reasons for organizations to implement RPA? The beauty of RPA is that it optimizes the use of existing systems, rather than replacing them. This results in quick implementations, with a minimal impact on the existing infrastructure and systems. The impact on the organization that implements robotics is of course much higher. Entire workforces can be reorganized, often resulting in people losing their jobs or being moved to other departments. And while some organizations choose to use RPA to improve their relationship with their customers by providing better services, others will simply choose it because they need to reduce their costs and become competitive again.
Organizations are implementing RPA for obvious reasons. But is it enough to be able to offer the highly personalized experience the customers of today are expecting? Robots work continuously and don’t make mistakes, but are they capable of making complex decisions?
Robotic Process Automation provides us with a system that acts in straight forward, repetitive processes. Like a human process handler on steroids, so to speak. And although this is already a huge step forward for most organizations in terms of efficiency and accuracy, it doesn’t provide us with the level of contextual intelligence that is needed for organizations to offer their customers the highly personalized, context-driven experience they are expecting. It is a first and necessary step towards the introduction of an intelligent, self-learning ecosystem.
In order to be able to create systems that know and act based on contextual data, we need to abandon traditional process-driven approaches and move towards the goal-based approach of an Intelligent Automation Platform. Robotic Process Automation will still play a major role in such a platform, but the orchestration of activities will be handled by a totally different technology. Let’s explore this topic in my next article!
What developments can you identify in a world in which technology is becoming more and more personalized? Is your organization responsive enough to cope with rising customer expectations? Get involved and leave a comment. If you would like to know more about contextual intelligence and how to shape it, please get in touch. Interested in the field and would you like to discover your career opportunities? Contact our recruiter.
Enjoyed this article? Check out the other publications by Kees van Mansom: