Payroll Will Always Need People

I have worked in the payroll industry for over 14 years. In recruitment, I get to work with payroll people who have a range of skills, dealing with ever more complex processes, getting to grips with ever-changing legislation and government interventions.


During this time the systems I have seen used have changed considerably. Who, reading this still remembers the days of Kalamazoo clock card software or Unipay payroll software developed by Peterborough Software?

Fast forward over a decade, and we are now living in a world of ERP, SAAS and Cloud based offerings. Dedicated payroll systems have transformed into multi-faceted packages that can support entire shared services functions.

Let’s face it – machines are on the march!

In my view, any occupation that involves logical thinking and meticulous organisation is now at risk of being automated by machine learning algorithms and sophisticated artificial intelligence (AI).

It is true that much of the payroll process can be automated, but with all the required checks, balances and legislation involved in ensuring its compliance and accuracy, I still believe that businesses need a human payroll team at its core.

How much more of the industry can still be given over to our future robot overlords? The main reason that people come to work is to get their salary at the end of the month – can we risk the “computer saying no” when we live in a world of debt and credit, and we have mortgages to pay?

Payroll is a delicate, niche and highly skilled industry where compassion meets computers. Yes, on the face of it, there is nothing overtly complicated about calculating the required figures if you know what you are doing, but for every twenty standard cases, there is one who is non-standard. Every single payroll department is different and comes with its unique complexities. Can systems truly grasp the individual nature of each payroll operation? The reasons might vary, but often it is only a human discussion and an agreed compromise that can resolve the issues.

Last time I looked, computers weren’t so good at compromising.

Also, I’m not sure how happy I would be if I knew that I had to take two hours off work one morning to find out the “automated payroll system” has docked me a morning’s pay? I might have agreed with my manager that I would make up the time at home, and may have even copied in the payroll team. With a human payroll professional at the end of the phone, I can rest assured that my pay won’t be affected. Relying on an automated system would make this almost impossible. How expensive is it then for businesses to correct automated systematic errors?

If artificial intelligence is allowed to take over, tracking the different income streams of an increasingly flexible and remote workforce would require a significant amount of human effort direct from the direct managers in question. Having a payroll team takes the hassle away from them, and managers know that all issues and enquiries can be fired in their general direction.
For me, this is the key point – remunerating someone for a job well done still involves a considerable amount of judgement. Has the work been done, how much has been done and should it be rewarded as agreed? No computer can make these calls. For a payroll team, that is their professional “bread and butter”.

I am all for AI allowing people the time to add the “human” value in processes, but you can never remove the human element entirely. Everyone talks about automated trucks, but I would imagine that there will be experienced truckers behind a video screen somewhere, watching out for hidden dangers. Maybe payroll activity will change more to assessing productivity and crunching the data of the organisation, but the human element will always be there in the background.

I’m not sure that everyone will agree with me. Some see such discussions as humanity hanging onto the coat-tails of inevitable technological progress, but for me, when it comes to rewarding people for their ever more sophisticated efforts, I’m not sure that we can put the machines in total control (especially considering how easy it might be to hack them).

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *