This article is a summary of the HR Podcast episode #17: “Is HR being replaced by automation?” where our HR colleague Lene Gjævert interviewed Ken Graham, who is a part of the Intelligent Process Automation team. For those who understand the Nordic language, the podcast is highly recommended!
First, what do we mean when we talk about automation?
Automation is often referred to as Robotic Process Automation (RPA), which is the automation of a process. However, we are not talking about physical robots that take up space in the office landscape. It is not Terminator.
No, we talk about virtual software agents that can operate in an IT system like an employee: clicking the same buttons, working with the same information, downloading the same files, and so on.
Digitalisation and robotics are popular words with many different definitions. They have one thing in common: it is all about working more efficiently. When you have a digitalised process, you can add robots to speed up or increase the quality in different parts of that process.
Robots can perform repetitive and often time-consuming tasks—just like a human being. It works best when it comes to repetitive tasks with structured data, which means data that are always in the same format. A table in a spreadsheet is structured, a receipt, invoice or mail is unstructured.
How can robotics be used to lift HR processes?
HR is all about working with the people in the company; from new positions and hires to exit and pension. Today, many of the processes within an HR department are administrative, repetitive work. The goal is often the opposite: HR should engage, attract and develop the employees.
Let’s see how automation can help the HR department, in four typical areas:
Selection is one of the most time-consuming tasks in recruitment. With a structured job analysis, one might think it would be possible to screen or tag different candidates if they are “approved” or “rejected” in a process. There are a few obvious challenges with doing this: robotics need structured data in the same format.
How many different types of CVs have you seen? As an example, you can require specific education, but there are so many ways to add information about education in a CV. Still, in case you have a structured approach and also a system where each applicant can mark to which extent they meet those requirements, a selection process could be quite automated.
Amazon tried to have an all-automated process, based on historical data. It was scrapped due to one problem: their new recruiting engine was discriminating against women. In the historical data, there were more men than women that were hired in the company.
If basing the robotics solely on historical data, without having a critical look at what’s behind the data, you end up with a bias, a wrong starting point integrated into the technology that must be carefully investigated. If humans make errors, robots will fall into doing the same errors.
A part of the onboarding is to get the signed employment contract from the recently hired person. Still, many companies sign the contract with pen and paper, a process that easily can be digitised. Digital signing already exists, ready to be used.
As soon as you receive the signed contract, you need to create the new person in multiple systems: the HR system, the CRM system, the payroll system, and so on. You probably have some of the data you insert manually: name of the employee, birth date, and address.
This is information found on the contract. To create the user in multiple systems is often a purely manual process. Manual punching of data could result in faulty data, which in turn can lead to time-consuming cleanup activities.
A possible solution for improving the onboarding process is to use a tool designed especially to model a given flow of steps. As an example, the contract can be read, the user created in the access management system and HRM system. Also, an order to local IT is shipped so that the employee has a fresh laptop on the first day.
Several companies have started to look into churn prediction to find out if an employee will leave the company. If you have enough data, it is possible to find patterns in data that can reveal if an employee is in the danger zone of quitting. It is all about giving the right equipment to HR or leaders to “ring the alarm” to watch out for some pitfalls.
Other simple examples can be sending out anniversary letters or automatic ordering of flowers and cake. Instead of having to keep track of who has been in the company for 10 or 20 years this year, the robot can do it for you (and we have the tool in the HR tech team!)
In this part, you have processes like creation of the work certificates, left-over vacation days, and adding the last day of working in many systems. It contains many practical and manual tasks.
For example, sending a message to the telephone provider to let them know the employee has quit. This is a repetitive and time-consuming task. Fortunately, it all comes back to onboarding: when employees are onboarded to a system, they have to be offboarded from that same system later on. This makes it an easy job for robots.
Final advice for HR leaders considering automation
- Know your processes! If several employees are working in the same process, do they solve the tasks in the same way?
- Ask some questions! Do I need to do this task in this process? What happens if I stop doing this task?
- Have full control of your own robots! If you are making your own robots or using a system, make sure that the solution is operated and monitored when it is put into production.
- Talk about it! If you get an IT system, you develop a robot, or you automate a process: talk with the employees who are affected so that they understand its purpose, and how they should relate to it in their everyday work.
Interested in learning more about how we work with automation and robotics in Visma, and how you can get started?