Payroll professionals know the pain. They know that payroll operations can often be excruciatingly manual and labor intensive. Beyond the collection and preparation of data for payroll processing, checking outputs from the payroll provider or local payroll engines can be excessively time consuming.
A payroll department of a German car manufacturer revealed that 60% of their time is spent double checking pay elements of each employee’s final paycheck prior to releasing it. That means members of the payroll team are spending many days every month manually ensuring that payroll outputs processed by the local payroll partners include all of their payroll changes. Accuracy is non-negotiable. Employees expect their pay to be 100% correct every time.
Variance analysis can be used to reduce the amount of time spent checking payroll by comparing current payroll with prior periods. While, of course, the calculated values can and will move around from one pay period to the next (e.g., due to different hours worked, sick leave, overtime, bonuses, commissions, changes in tax rates, etc.), a systematic comparison of the values calculated for an employee across different pay periods does allow a quick way to flag potential errors and outliers.
So, in many ways, this is a great example of robotic process automation (RPI)/artificial intelligence (AI) assuming time-consuming tasks formerly in the domain of payroll employees. Where a person takes hours and days to compare data points and might still overlook certain deviations or patterns, software can process huge volumes of data and run intelligent comparisons within seconds.
This is especially the case if the system is smart enough to factor in things like end-of-year bonuses, typically awarded at the same time of the year. For example, if the base salary or health care deductions suddenly jump from one pay period to the next, someone should probably take a closer look. And while a variance analysis may create some false positives (i.e., flag a variance that is legitimate because the circumstances of the employee have changed), it does focus the payroll team’s review of the data on the actual variances and quickly eliminates the review of data points that are in the expected “normal” range. This eliminates the need to sift through heaps of data within the normal range. For example, this kind of variance analysis will highlight all leavers and joiners in a period—they will show up with their values increasing or decreasing by 100%, respectively.
The next level in an automated data validation methodology is to factor in the data inputs themselves to assess whether the outputs appear to be accurate. This essentially allows you to check on the changes that you have communicated to payroll. While the variance analysis above gives you a method to detect anomalies, this level of analysis gives you further certainty.
For example, the system would recognize that the number of standard shift hours submitted in the system increased from one month to the next by 10% and would check that the standard shift payout showed a similar increase. The system can also check that the implied tax and social security rates fall within the expected range.
As the machines “listen” and learn from your payroll data, they can go one step further and identify the tougher payroll data issue: data that should have been sent but was not.
Let’s go back to the shift pay example. Say one of your colleagues worked a shift pattern that included a weekend shift premium once every six weeks. The AI can detect anomalies of omission that data variance analysis or advanced data validation may not. The software should highlight this as a potential issue at the point of submission of your changes to your in-house teams or external providers.
While this kind of smart technology is becoming increasingly common, many smaller local payroll vendors do not have the means to implement it. This kind of automation is prohibitively expensive for customers to develop on their own. At the same time, it is precisely the data from remote countries that is particularly difficult for global payroll professionals to check. When you’re looking at data in a foreign language and a foreign currency, it can be incredibly difficult to even know what you are approving. So, a tool like a variance analyzer allows you to quickly perform some variance checks and shows you where the data is consistent with previous periods. By identifying where there are significant outliers, you can increase confidence levels and reduce frustrating and time-consuming manual spot checks.
Multiple Benefits of RPA
The great thing about this kind of process automation is that it benefits both the customer and the local payroll providers. It helps to reduce the manual effort on both sides and increases the efficiency for both the local payroll processor and the customer. And maybe more importantly, it reduces the number of errors that slip through the cracks that impact the employee’s payroll, dramatically reducing employee queries and associated efforts and cost to investigate and resolve them. All off this helps meet the expectation of getting things right 100% of the time.
We have looked at the benefits of RPA)/AI on the process once you have the data ready to go to your provider, but we all know we need to look upstream to the sources of payroll change data in your organization. Manually hand cranking data to get it into a format that the local payroll engine needs is another source of significant data accuracy and information security risk. What if there was a way to automate data collection processes by integrating your data sources with your payroll approval processes and then transforming that data into a format that your payroll provider or in-house teams could use without the need to intervene, manually chopping and changing data to suit local payroll needs, transferring data from one file format to another?
No more double entries and transposing of data. The new employee and demographic data that you entered into your core HR system will show up automatically in the local payroll system. We mentioned earlier the example of 60% of a payroll team’s time being spent wrestling with data and data validation. By connecting your sources of data, you will further drive accuracy and efficiency. Most importantly, you will let your payroll teams focus their time looking into more strategic issues such as payroll governance, mobility, benefits optimization, and generating real business insights from some of the most accurate data available.
So, how great would it be if you could have a consistent way to compare and validate all your global payroll data in this way—for all employees no matter where they sit, in large or small countries? What if you could plug this kind of process automation into your existing local payrolls? Well, the good news is that with the next generation global payroll solutions, now you can. So, what are you waiting for? Start to bring process automation to your global payroll operations and improve your process efficiency, your data quality, and your employees’ satisfaction.
Do you like our content? Join the GPMI community to get free education and articles straight to your inbox!
Marc-Oliver Fiedler is the co-founder and CEO of Payzaar (www.payzaar.com), an innovative global payroll platform. Fiedler has been a pioneer in building global cloud solutions at companies like ADP, Oracle, and Hewlett Packard. He started his career as a management consultant at The Boston Consulting Group and holds an MBA from Stanford University. Fiedler is a leading voice on the emerging paradigm of open-source global payroll solutions and the merits of a consolidated global payroll reporting framework.