White Paper. Big Data Turns Labor Information into Gold

Size: px
Start display at page:

Download "White Paper. Big Data Turns Labor Information into Gold"


1 White Paper Big Data Turns Labor Information into Gold

2 Executive Summary Today, manufacturers are surrounded by the buzz about big data as well as by the flood of information it represents. Why should they care? The answer is that a big data strategy creates opportunities to transform existing information about their workforce into business value. With the right data collection, analysis, and action, manufacturers can turn everyday data straw into information gold. The Big Data Buzz As with many buzzwords, big data seemed to appear out of nowhere. As fast as you can say exabyte, it s become part of our business vocabulary. In April 2010, after some initial resistance, big data became an accepted term on Wikipedia. big data now shows up everywhere from the tech blogosphere to the plans of CEOs to the pages of The New York Times. Over the last couple of years, the exponential growth of information (and the tools to handle it) has made big data part of everyday discourse. Buzzword status aside, what big data describes is real and growing at a furious rate. Each day, according to an oft-quoted IBM statistic, we create more than 2 exabytes of data that s 2-followed-by-18-zeros bytes. As a result, 90 percent of the data in the world has been created in the last two years alone. 1 No wonder the worldwide market for big data technology and services is expected to grow from $3.2 billion in 2010 to $16.9 billion in 2015, according to International Data Corp. 2 And a 2012 survey of 300 of the world s largest enterprises by Computerworld found that 13 percent had big data initiatives in place with another 38 percent planning or likely to implement such projects in Why Care About Big Data? Big data means different things to different people, but well-known data mining consultant Ken Krugler may have described it best: more than you can handle with the computer you ve got and scaling up isn t an option. 4 What this means is that the old-school model of leveraging the efficiency of technology isn t going to solve your big data challenges. It s going to take a different approach. 1 Bringing Big Data to the Enterprise, IBM Corp., accessed December 30, 2012, 2 IDC Releases First Worldwide Big Data Technology and Services Market Forecast, Shows Big Data as the Next Essential Capability and a Foundation for the Intelligent Economy, International Data Corp., March 7, 2012, 3 John Bantleman, Big Data: Business or Technology Challenge? Wired, January 3, 2013, 4 A Very Short History of Big Data, Ken Krugler, accessed December 30, 2012,

3 Then why bother? One reason, as the Computerworld study suggests, is that your competitors are. More important, the ability to manage big data creates strategic, sometimes unforeseen, opportunities. Entire companies, such as Google, have been built around the availability of massive data. Other companies have plumbed the depths of big data to cull out subtle but powerful insights about their markets. Retailers such as Target use point-of-sale big data to predict what customers will buy next, and then create highly customized promotions to win that purchase. 5 But the value of big data isn t limited to breakthroughs like these. Big data strategies also allow manufacturers to transform vast amounts of their own existing labor information into important business value. Like improvements in any aspect of manufacturing, this transformation begins with understanding the process, which can be represented by an equation: Data + Analysis + Action = Gold. Big Value Begins with the Right Big Data Data the raw material of decision making has a large impact on the gold you produce. And as with any raw material, the challenge is to reduce its cost while ensuring the quality of the final product. The right data collection techniques and instrumentation can help you meet that challenge. From manual timecards to scanned bar codes and proximity cards, there is a wide variety of methods for collecting labor data. Collection costs continue to drop as new technology becomes more widely available. Today the mobile device a FedEx driver uses to scan a package at your door is the same technology workers carry in manufacturing plants. Supported by standardized applications, such devices are allowing employers to collect data that is more precise, more useful, and more reliable, but without greater costs. Big data precision Determining the right level of precision the granularity of the data you collect and how frequently you collect it depends on the nature of the data and its ultimate use. If required for compliance, information about every instance of an event might have to be captured. If your data is being used for statistical measurement, you may need less granularity; instead, common sampling strategies can be used to minimize the cost of data collection. 5 Kashmir Hill, How Target Figured Out a Teen Girl Was Pregnant Before Her Father Did, Forbes, February 16, 2012,

4 For example, a client of Kronos Incorporated was trying to improve operations at a large distribution center. After realizing that its supervisors didn t check the status of operations more than a few times an hour, the client decided to review workflow at 15-minute intervals and then statistically average the demand and performance levels. This solution provided the right balance between useful information and the cost of processing it, while not overloading supervisors with granularity that would slow decision making. Based on the information required, labor data is typically collected in four ways, each of which provides a different level of precision and comes with different costs. Inferred data collection is a crude comparison between the amounts of labor associated with a product and how much of the product is made. It s the easiest and lowest-cost form of data collection but ultimately can be the most expensive in terms of hidden costs, labor compliance risks, and lost opportunities to improve productivity. Averaged or estimated data collection provides a closer look at labor costs. A supervisor tallies the total hours worked by a team and then allocates that time across the output of a production run. The result is an improvement in precision and accuracy, but also higher costs in collection. Exception-based time collection is used where work is assumed to be done at a constant rate, except for occasional changes. For example, an employee may be paid according to a work schedule with adjustments made for sick days or vacation. This approach, too, is an improvement over inferred data collection but still only provides approximations. Real-time tracking measures actual values, such as the in/out times of individual employees, their break and lunch times, or the hours they spend working at specific production tasks. This approach to data collection provides very precise, very accurate information but it s the most expensive in terms of collection costs. Big data integration One of the most significant values of big data is the innovation that occurs when disparate sources of data are combined. But that can be difficult because highly granular data that is optimized for a particular use is often incompatible with other uses. The challenges range from date formatting to complex issues such units of measure (e.g., cases or pounds for food processors, biweekly or semimonthly pay periods for HR professionals). This can lead to making decisions based on information that is easy to capture instead of information about what is driving the business. Manufacturers should look for big data solutions that integrate rather than isolate.

5 Big data integrity Data that is incomplete or of poor quality can have a surprisingly large ripple effect throughout a manufacturing environment. For example, if hours or components used are calculated by inference when entering production quantities into an ERP system, a single mistaken entry can create material and labor variances that ultimately lead to higher costs or late shipments. As with many quality improvements, the best way to improve the quality of labor data is at the front end. Bar-code scanners or devices that set limits on possible values can reduce the amount of poor data going in and so increase the integrity of business intelligence coming out. Spinning Gold Takes Analysis and Action Big data changes the way information is analyzed and used to take action. Traditionally, information was reported to solve specific objectives. But with terabytes of data now available to them, manufacturers can confront overarching challenges that are less-defined. As companies use big data to meet those challenges, they need people with a new profile of skills. That profile includes: Strong analytical skills. In the world of big data, correlation is commonplace. But it is a Siren s song. Masters of big data understand that causal relationships are harder to find than ever, but also more important. Big-picture perspective. With so many potentially irrelevant conclusions that can be drawn and paths that can be followed, big data requires people who can look beyond their department or project to understand the impact of their findings on the company and its market. The ability to tell a story. Because big data touches so many functional areas and stakeholders, a big data cruncher must be able to tell a story with the numbers and help others visualize the numbers value. This person must be able to put their conclusions in context. But having an a-ha! moment isn t enough. The most valuable idea is worth nothing if it can t be acted on. Is the impact measurable? Can it be used to prove a theory or dispel a myth that affects how the business is run? For example, a small change in materials may detract from product quality, leading to increased overtime and changes in how union agreements affect labor costs in future years. big data can help manufacturers connect the dots so that insights are backed with fact, made relevant to the bottom line, and create the opportunities and the motivation for action.

6 The Many Types of Big Data Gold The end result of data, analysis, and action is the monetization of an idea. But despite the headline-making breakthroughs of big data users such as Target, companies can find data gold in everyday ways that don t require large investment. For example: Single-variable change. One of the simplest uses of big data is to measure a single, previously unexamined variable, such as absenteeism or early/ late time-clock entries. Capturing this information provides visibility to managers and opportunities for more effective solutions. Finding the needle in the haystack. A more challenging problem, but one for which big data is well-suited, is to identify a relationship between two or more variables. This is often a problem that common sense and traditional solutions fail to solve. For example, the use of overtime is a frequent problem for manufacturers, and because it is a response to so many scenarios (increased demand, underperforming equipment, etc.), the cause of overtime can be difficult to determine. But at a recent meeting of the American Payroll Association, one payroll manager explained how she used data to discover a root connection between overtime and turnover increases. Management had been unable to make the association because the relevant information was handled in different departments. Once the connection was found, overtime policy was adjusted. Avoiding death by a thousand cuts. Sooner or later, every manufacturer hits the continuous improvement wall. Improvement methodologies have wrung diminishing gains out of operations, but performance is still not what it should be. As managers look more closely at production, they discover that many small, under-the-radar delays are adding up to much larger ones. These insidious problems have a cumulative effect and can seem impossible to solve until they re modeled statistically with big data on labor and production. What seem like random, uncontrollable events can be traced back to common underlying causes that can then be better managed. Predictive modeling. Manufacturing results are well-documented; manufacturers know what has worked in the past. They also can use that data to create business intelligence and support decisions about the future. For example, manufacturers that use contract employees often learn from past performance that these new workers should be teamed with more experienced ones. This creates a predictable situation. So rather than recognizing the problem at the beginning of a shift, managers can use big data to devise schedules ahead of time that take into account planned absences, seasonal demand changes, and forecast absenteeism.

7 Information aggregation. Data is often collected by individual departments to solve individual problems but then sits unused, like the straw on Rumpelstiltskin s floor. At the same time, manufacturers are looking for insight that crosses department boundaries, such as the total cost to serve a customer. This information is necessary to understand the profitability of a market or a product line, which then drives price negotiations, investment decisions, and corporate strategy. By leveraging big data, companies can aggregate information from multiple dimensions and sources and use it for multiple purposes. Manufacturers know how to turn low-value materials into high-value products. Turning data straw into information gold follows a similar process. By starting with vast amounts of clean data they already possess, applying big data analysis, and finally acting on the newfound knowledge, organizations can increase their competitive advantage from within. How Kronos Can Help Kronos helps organizations of all sizes unlock opportunities hidden within their workforce processes to control labor costs, minimize compliance risk, and improve workforce productivity. Our easy-to-own time and attendance, scheduling, absence management, HR and payroll, hiring, and labor analytics solutions and strategic services provide complete automation and high-quality information, and deliver the experience our customers expect. To learn how Kronos can help you leverage workforce management tools to enhance the strategic contribution of big data, call us today at or visit kronos.com/manufacturing. Kronos Incorporated 297 Billerica Road Chelmsford, MA , Kronos Incorporated. Kronos and the Kronos logo are registered trademarks of Kronos Incorporated or a related company. All other product and company names are used for identification purposes only and may be the trademarks of their respective owners. All specifications are subject to change. All rights reserved. MF0148-USv1