Would you let your payroll system suggest which employees deserved a pay rise?
It’s not difficult to imagine a world run by robots. We’ve heard experts tell us it’s just around the corner and that it will completely revolutionise the way in which we work.
Although the word “revolution” is often overused, it’s a near certainty that Artificial Intelligence (AI) will be exactly that. Like automation and the Internet before it, AI is set to change the way every industry works – payroll included.
In fact, PwC estimates AI could contribute up to $15.7 billion to the global economy by 2030. This is leading to a lot of fear, amongst managers and employees, that it may have detrimental impacts on job security.
In fact, almost a third (30 per cent) of Australians are not welcoming of AI into the workplace, according to SEEK. But, according to this research, one of the main barriers to adoption is our lack of understanding of these new technologies.
So, do we really need to be concerned?
Going beyond automation
Payroll is already ahead of the curve when it comes to automation. Where once a payroll professional could be expected to manage 50 employees, that same employee can now provide payroll support for thousands of employees.
This is all thanks to incremental automation. From compliance to automated calculations of wages, tax deductions and overtime, there are many areas of payroll that have greatly benefited from automation. AI is the next step forward in the evolution of payroll technology and will bring greater insight, efficiencies and understanding to the way in which we do our jobs.
AI could provide employers with insight into their workforce efficacy. For example, it will notice that multiple staff are staying back later and responding to out-of-hours emails more and, through its assessment of this data, could recommend the hiring of more employees or suggest pay increases for those employees.
Further to this, AI could greatly improve compliance. For example, in the US only 33 per cent of respondents to a 2014 Deloitte survey said they track their employees’ domestic expenses. An average payroll fraud scheme, like a worker claiming time they didn’t actually work, could last for 30 months before being discovered, according to the Association of Certified Fraud Examiners.
This is where AI comes into play. It will be able to curb the losses from such wrongful behaviour by tracking and analysing the data in real-time.
AI will be payroll’s helping hand
There are many other benefits AI will bring to payroll but one starting to seep into organisations more is the use of chatbots. AI-powered chatbots are able to answer a range of basic questions that an employee might have.
The most common use is assessing an employee’s query and either answering it or directing the employee to the right person. This reduces the time payroll might need to spend on menial tasks and basic questions, helping focus their energy on becoming a strategic function of the business.
This is how AI will stand out. It will do the heavy lifting and admin roles in the payroll department, as well as better connect it to other parts of the business. For instance, it may notice an employee has begun to take regular half days and reduce their overtime allowance. It could then alert the payroll department or the employee’s manager to flag a potential issue; they might be taking time off for personal reasons. Functions such as this will be especially useful to large organisations, particularly as more staff enjoy flexible working conditions or work remotely.
However, the question remains – will AI replace a payroll manager?
Be informed, not alarmed
Although AI will introduce incredible technical efficiency, it could not possibly replace a human. After all, while AI may give you the data to justify a pay increase, would you want to negotiate salary with a robot?
In many aspects of payroll, this is critical. When employees have sensitive queries regarding their payroll, they want to talk to a real person. No matter how sophisticated a chatbot is, we will never truly relate to them.
Payroll still needs the human touch, even if we continue to go full steam ahead into AI technologies. When used well, it can help improve the efficiency of payroll teams and enable staff to focus on innovation within the payroll function.
But, as it stands today, AI is not ready to deal with our highly dynamic payroll environments. Automation, for instance, ensures workers are paid on a schedule. But, there is not likely an AI function that can accommodate for spontaneous adjustments, like a payment outside of a salary cycle…yet.
Payroll will inevitably become more automated and AI will enhance the technological capabilities of payroll software. But, it’s still a long way off from replacing humans. For payroll managers and businesses looking to take advantage of AI, it is crucial to build it up gradually. That means learning what tasks AI could help us to achieve in the future and understanding where, as a payroll professional, we can add further value to the business.
It’s a case of being informed, not alarmed.
Sandy Forrest is the general manager at Ascender Public & Enterprise Services.
AHRI’s elearning for teams and organisations offers consistency of content ease of delivery and is a cost-effective online training solution. Give your employees the flexibility to gain new skills at times that suits them and meets your organisation’s needs.
Payroll process is certainly highly automated now with potential for greater AI intervention. It is how we will handle payroll in the future with flexible and remote working; working from different countries with different currencies & pay curves.
There are for example intelligent computer screen monitoring for work time tracking.
However payment for outcome is also likely to increase from the current say room servicing in hotels & CAD to a broader range of professional roles.
Great article. The opportunities to improve compliance and employee outcomes through AI are endless, and exciting. Although it’s a 15.7 trillion impact to the global economy, not billion as stated. I thought the number seemed modest.