All you need to know about Process Automation
What is a business process?
It is important to understand what a business process is: it is above all a set of tasks and global steps (end-to-end) to achieve a specific objective. We can illustrate this with an example of supply chain:
A customer order on the internet will result in a succession of tasks: sending an order confirmation email, preparation of the order, stock modification action, transmission to the carrier, delivery and reception of the order by the customer for the main ones.
All of these tasks are necessary for the proper delivery and invoicing of the customer.
The tasks are often listed and time-stamped in several different systems (ERP, CRM, xMS..). This must allow an operator to know at any time where the order is and what its status is.
Process Automation and its potential
Automation is a technique or a set of techniques aimed at reducing or rendering useless the human factor in a process, with the objective of productivity gains.
The potential of automation is huge. PwC estimates that 45% of tasks performed by humans can be automated, which would allow savings estimated at 2 trillion.
It is certainly thanks to this potential that companies have embarked on a crazy race to automation without obtaining all the expected results…we will try to understand why and give some good practices in this area.
Succeed with your automation project
1. Efficient automation is only possible after having built or restored an overview of the process operation
Once all the steps of the business process have been identified and listed in your various tools, the use of process mining becomes key. It will allow you to monitor the entire process and make their functioning readable.
Thanks to the information logs, process mining reconstructs the main steps, the routes taken by each unit and gives information such as the timeframe, the value of each unit, the actors, etc.
This transparent view based on real facts allows you to make automation decisions based not on how your processes were designed but on how they are operated in the field, otherwise known as “as-is”.
In addition, process mining provides you with a vision of the history, a real time vision and also a predictive vision of your processes. You thus have a temporal and dynamic vision which will be able to guide your choices and help your decisions.
2. Determine which parts of the process need to be automated
Once this overview is reconstructed, I will have to determine which parts of my process are automatable or need to be automated first.
Here again, process mining will support your decision. Thanks to Machine Learning and AI, you will be able to easily identify and categorize process anomalies: excessive delays, non-compliant paths, rework, bottlenecks, etc. For each element you will have the root causes, which will give you valuable information on the choice between a simple optimization because easy to implement (change of provider for example in the case where a provider is less good than the others) or if an automation is necessary.
This is often the case for:
- recurring and voluminous tasks
- tasks where human intervention causes errors and therefore reworks
- simple and numerous verification tasks
3. Choose the appropriate technologies for the parts to be automated
The automation can be done internally via simple developments. It can be for example to create a link between 2 of your information systems to enrich them with data and avoid double data entry.
It can be, as we have seen, AI or ML which allow, thanks to algorithms, to carry out autonomous tasks in an automated way by recovering a context. For example, machine learning can be used to check compliance at different stages of the bank credit process.
Finally, RPA allows the use of robots to perform simple tasks such as questions and answers in chatbots.
Fields of application of Process Automation
Automate a conversation with a customer
RPA makes it possible to handle requests that are often simple or without added value. A chatbot could be an example of simplified customer relationship processing. Indeed, some necessary steps of qualification of the customer context (asking for name, first name, email address, order number, etc.) can often take time. Entrusting this to a robot saves time for the operators and allows them to be more involved in tasks with higher added value.
Avoid human error
Manual processes necessarily involve malfunctions or errors. In accounting, for example, invoices may be sent to the wrong people, errors in due dates or errors in amounts. In this case, automated processes can reduce error rates and accelerate processing times.
Control quality and compliance
Automating verification tasks in a consistent and systematic manner can strengthen your quality system and improve compliance.