A senior finance person once asked me to report on the “real” total cost of payroll. Like most payroll professionals, I immediately ran a query of employee earnings and employer expenses. However, this finance person was looking for inclusion of other factors such as how much it costs to produce, audit, grow, and sustain the payroll operation year over year.
“How should I even go about attempting to answer that?” I thought.
This inquiry was the start of a growing set of questions organizational stakeholders would ask. They were not just looking for data, but an analysis of data. Many of you may be asked similar types of questions such as: What departments have seen the greatest fluctuations of overtime over the past five years and why? Have employer expenses risen or fallen over the past decade? What is the median withholding bracket of our employees? If you work in an organization focused on understanding how its financial resources are expended, you can expect to get some variation of questions like these.
When I was first approached with these questions early in my payroll career, they were usually met with a blank stare and an awkward silence, later followed by a laugh and a comment like, “I don’t know” or “ask finance” or “go bother someone else with this.” Over time, I have been fortunate to learn from some fantastic finance leaders and data specialists who helped me identify the sources and value of data all around me and then how to interpret and report on this data. In this article, I’d like to share a little of what I’ve learned about three essential and powerful research tools and perspectives: artificial intelligence (AI), analytics, and big data.
Artificial Intelligence
If you search Google for the term “artificial intelligence” (AI), you are sure to find many definitions attempting to simplify a very complicated concept. My definition is that AI refers to technological software and hardware designed to replicate tasks traditionally thought to be human. For many, this invokes images of robotic arms working in factory assembly lines or supercomputers interacting with humans in conversations or chess matches. These are only some uses of AI. Anyone who plays video games can attest to a revolution in gaming due to the integration of AI making computer simulation not just more competitive but improving the overall gameplay experience. The spirit of AI is to build computing systems that respond to this ever-growing need for our computers to be smarter. That is, a system that will not just return an answer to our request, but one that analyzes all of the factors, conditions, and variations available and delivers a result tailored to our specific situation—for instance, not just suggest a movie to watch, but also when and where to watch it.
AI has made its way into payroll as well. In payroll, human resources, and finance circles, there has been a lot of conversation regarding the use of robotics, specifically robotics process automation (RPA), to streamline administrative tasks. In case you have never heard of RPA, the following is a 2014 definition provided by the Institute for Robotic Process Automation and Artificial Intelligence (IRPAAI):
"RPA is the application of technology that allows employees in a company to configure computer software or a ‘robot’ to capture and interpret existing applications for processing a transaction, manipulating data, triggering responses and communicating with other digital systems."
As an example of potential uses in payroll, think about employees who work in multiple departments doing multiple jobs at different rates of pay, or those who earn shift differential with special rules. While there are payroll and time clock systems that track where employees work and at what pay rate, many systems rely on certain employee input (such as logging in a department code in the time clock) or some manual calculation done by the payroll preparer. Having a time clock system embedded with AI could track and calculate this data for you and then deliver the data needed to produce pay to payroll in the format required for processing. RPA is an example of an AI technology that helps free up payroll resources to focus on the review and production steps. If you work for an organization with hundreds or thousands of employees or manage a payroll operation with various interdependent validation steps, or steps that take a lot of time, RPA can significantly reduce manual effort, increase accuracy, and reduce overall processing time.
Analytics
This “smarting” of our technology is due in large part to a mixed usage of the “Internet of Things (IoT)” and analytics. IoT refers to devices that use the internet to collect and exchange data that is used in its function. Think about how your smartphone does more than send or receive calls but acts as a digital assistant, helping you keep up with appointments, stay connected to family and friends, and providing access to businesses and services you use. Most of the devices we use every day are getting smarter, from TVs to cars to credit cards. There are even new lines of smart appliances, such as refrigerators that can send you a text message to say what you are out of while you’re out shopping. Think about many of the newest automobiles that can now sense traffic around them and initiate “safe” maneuvers, park themselves, and in some cases even drive themselves. IoT works hand in hand with robust analytical programs that look at how consumers interact with products.
I like to think of analytics in two parts: data collection and the method of handling the data. The collection of data refers merely to how we receive the types of information used to reach a conclusion or better understanding. For example, how would you go about understanding how your customers perceive a new product or how a policy change affects your employees? Many companies use social media as a powerful resource to collect data about their customers. Others solicit feedback on their websites or through polls. The handling of data refers to how the data is prepared, vetted, and then analyzed to reach a result. The task of considering the vast number of opinions and perspectives can be daunting. Thankfully, big data is a data science methodology used to simplify studying large quantities or varying complexities of data. Within big data are four different types of data analytics:
- Descriptive analytics—Looks at an issue from a historical perspective. For example, how many employees have we paid in the last five years? How many of those checks needed to be corrected? This approach is ideal only when you need to know what has happened or is happening. It cannot give insight into why.
- Diagnostic/discovery analytics—Attempts to understand the reasons behind a result. For example, if you discovered that 3% of your total check production was due to a correction, you could use a diagnostic analysis to understand the reasons behind these corrections. Is there an earning or deduction that calculates inaccurately? Are hours for a specific department consistently reported late? Decoding the reasons behind a result can improve process efficiency and even save the company money in many cases.
- Predictive analytics—Looks at data to predict what could happen in the future. We see an example of this kind of analysis a lot during political elections in the form of voter polls that attempt to predict which candidates might win an election or determine how voters feel about a candidate’s position on the issues. In organizations, it can give decision-makers insight into what employees are more likely to want or not want, and how those choices might affect net pay and employee satisfaction.
- Prescriptive analytics—An advanced form of predictive analytics that utilizes A/B testing or automation to optimize solutions to an existing problem. Often, you can find that a problem has many answers. The prescriptive analysis looks at which option is most suitable to the result you are ultimately seeking.
I work for a client that produces payroll for more than 80,000 employees spread over 31 countries. For my client, there are no such things as “simple” changes to payroll, as every policy change needs to be vetted to consider downstream impacts to all aspects of employee pay.
Big Data
Many major retailers have long identified the value of using big data to understand better what their customers want/need. One such example is Walmart. A 2004 New York Times article explains how Walmart used big data to determine how to stock its stores for an impending hurricane. An excerpt from the article follows:
Hurricane Frances was on its way, barreling across the Caribbean, threatening a direct hit on Florida’s Atlantic coast. Residents made for higher ground, but far away, in Bentonville, Ark., executives at Wal-Mart Stores decided that the situation offered a great opportunity for one of their newest data-driven weapons, something that the company calls predictive technology. A week ahead of the storm’s landfall, Linda M. Dillman, Wal-Mart’s Chief Information Officer, pressed her staff to come up with forecasts based on what had happened when Hurricane Charley struck several weeks earlier. Backed by the trillions of bytes’ worth of shopper history that is stored in Wal-Mart’s computer network, she felt that the company could “start predicting what’s going to happen, instead of waiting for it to happen,” as she put it. The experts mined the data and found that the stores would indeed need certain products—and not just the usual flashlights. “We didn’t know in the past that strawberry Pop-Tarts increase in sales, like seven times their normal sales rate, ahead of a hurricane,” Ms. Dillman said in a recent interview. “And the pre-hurricane top-selling item was beer.”
While this example is from more than a decade ago, there are many modern-day examples of analytics and big data working together. For instance, if you use Netflix, Amazon, or Facebook, you are likely familiar with the movie, product, and friend suggestions feature from these sites. How does Netflix know what you might want to watch? Or Amazon know what products might interest you? Or Facebook understand what people you might know? The answer is by analyzing previous selections you’ve made, evaluating them against hundreds if not thousands of transactions, and using powerful predictive or prescriptive analytics tools to offer these suggestions.
So how does big data help the payroll professional? Recently, HRO Today published an informative article to discuss some of the ways payroll data is used in strategic decision making. “Many payroll providers are mining the unique data produced by their clients' payroll runs to glean valuable insights into the effectiveness of these organizations' incentive compensation plans, regional employee productivity, nearshoring and offshoring programs, and even talent management initiatives,” the article said. “This insight fosters better strategic decisions on where to best deploy company resources—labor chief among them.”
Value of Data Analysis
Truthfully, you may read this and think these tools are way more than you would ever need. For many payroll professionals, Microsoft Excel is the most powerful tool they will ever use. Still, the real beauty of AI, analytics, and big data is that they are scalable to meet your specific organizational needs. If your company sells products, these tools can help do everything from track inventory to understand better the consumers who use these products and even help suggest other products consumers might want. If you work for an organization that provides a service, these tools can be used to measure the performance of your service, compare it against industry standards, and continually improve your service. However your organization chooses to use these tools, it’s beneficial for payroll professionals to know how to use them. It is no longer enough just to process payroll; professionals are increasingly asked to interpret what the payroll data means. These tools help make that analysis more accessible and accurate than ever.
Daniel Thompson Jr. is Senior Manager of International Payroll and Accounting for UTIO and a member of the Board of Contributing Writers for PAYTECH.
