Human Resource Management is supported by a very broad and deep range of transactional HR processes. Of these, Payroll Processing probably contains the greatest administrative activity.
Payroll is an “omnipresent” process, highly repeated and rules based. It is an activity that demands high service quality and accuracy, whilst (being viewed as a commodity) requires that the cost of delivery is low.
As a result, payroll presents the greatest opportunity for simultaneously improving the consistency and quality of service delivery while driving down the total cost of delivery.
“Just do it right” but also ‘’do it cheap”!
These seemingly inconsistent corporate demands have caused payroll to assume a degree of importance beyond the face value of a standard process that pays people accurately and on time.
Payroll is highly repeated, governed by both statutory and fiscal rules as well as defined by company guidelines and audit requirements. Gross-to-Net calculations are basically arithmetic calculations; bank transfers are the most common accounting processes undertaken by employers. Pretty much everyone in employment gets paid every week/month. So why is it that payroll doesn’t run at 100% all time; every time?
Clearly, the unplanned interruptions to standard processes by input errors; non-standard time and attendance processes; the range of different forms of input media and approaches as well as off-cycle changes dictated by business events doesn’t help. Equally, the changes dictated by local and country level fiscal and statutory changes also adds a degree of “churn” to a standard process.
However, our approach is to get the quality right at the start of the process through the adoption of self-learning artificial intelligence (AI) and automated processing of standard activities. The benefit of an “autonomic” approach to work processing is that it is self learning, designed to adapt to unpredictable changes while hiding the intrinsic complexity from operators and users.
Very rarely are companies’ payroll processes fully integrated. There are typically free-standing time and attendance systems, compensation systems that may have manual interfaces; “by the event” payment policies; shift management issues, less than fully integrated approval and exception handling requirements and even country-specific gross to net payroll engines.
When the challenge from the business is to do this right but also do it cheap, the dissonant response is to add check-points; programme runs/re-runs; off-cycle payments and all sorts of work-arounds.
The ability to deploy autonomic processing through the use of robotic process automation (RPA) and AI allows operators to eat through much of the repeatable processes. Operating at the user interface layer of applications and working between existing systems allows the autonomic platform to mimic what individual payroll administrators have been doing to keep the system running. Displacing this work and allowing the payroll experts to “get ahead of the game” has a positive effect both on total cost of operation but a remarkable improvement on service quality and “calming the noise” of payroll errors.
The result is that the process expertise is freed-up to view the whole process and ensure interventions are identified early and acted upon decisively. They become the point of expertise defining the rules that the robotic process automation enacts – right every time without the costly effort of fully integrated HR/Payroll systems.
Payroll Process Examples
As an example, the following processes have demonstrated tangible benefits from the deployment of robotic process automation
New starters / leavers
Contractual changes – position, benefits, roles, responsibility
Permanent changes – increases in salary, adding benefits, contact details, address changes
Temporary changes – bonus payments, expenses payments, overtime
- Validation and reconciliation of payroll
Deductions – pension, union, travel card, loans
… and the results. A robotised process typically takes four to eight weeks to implement and delivers an in-year pay-back with the following outcomes:
Accuracy of 100%
Auditing requirement reduced by 80%
Cycle time cut by 40% (including exception handling and auditing)
Completion rate (i.e., work flowing straight through) on Absence ~87%
Completion rate on Attendance ~ 96%
FTE savings of between 80-90%
Human error (at the data input or processing stage) is still the largest point of failure in any HR Process, payroll is no exception and with the scarcity of true payroll expertise, expecting robotics to take the “leg-work” out of paying people makes absolute sense.