Accelerating the
Industrial Internet of Things

Will IoT technology bring us the quantified employee?

Published on 05/24/2016 | Technology

250 1

Josh Bersin

Accelerating the Adoption of Industrial Internet of Things.



Wearables and other IoT-based devices are helping millions of us better ourselves. Can employers convince everyone to use the same technology to make work more efficient, productive, and pleasant? All that data offers great potential for workplace gains—as long as everyone’s on board with the program.


World-class athletes track how far they run, how much weight they lift, how many calories they ingest, how much sleep they get, and even how much oxygen they consume. By measuring and monitoring these variables, athletes can optimize their performance, shave seconds off their time, and gain a competitive advantage on the field. And technology that collects information and offers quick feedback helps these athletes conserve energy, keep track of daily workouts, and determine their peak performance time.

While few of us face such pressure on a court or track or field, Deloitte research suggests that we are all operating as “corporate athletes”—dealing with too many decisions, too many emails, and too many meetings in not enough hours (two-thirds of all businesses characterize their employees as “overwhelmed”). We, as business people, need tools to track our own productivity just as badly as athletes.

The desire to quantify, measure, and monitor ourselves has spawned an entire industry, with companies developing wearable computing devices, fitness trackers, and mobile communication tools at a fevered pace. Consumers bought more than 45 million wearable devices and fitness trackers in 2015, and analysts expect demand to grow by more than 45 percent annually through 2019, becoming one of the fastest-growing technology markets.

What are these wearable devices doing for us? They are giving us information on our exercise, sleep, movements, diet, and pulse, creating the quantified self, powered by an architecture of technology referred to as the Internet of Things (IoT).

But when the quantified self arrives at the office, does he or she become the quantified employee? Many employers would hope so: With oceans of data from workers’ wearables, HR departments could aim to create more pleasant and efficient work environments by looking at productivity, patterns of communication, travel and location trends, and how teams work together. But there are real obstacles to enlisting a workforce into this effort, beginning with the fact that employees aren’t necessarily comfortable giving their bosses unrestricted visibility into their movements and more.

As we as consumers spend more and more money tracking ourselves, it’s a safe assumption that most of us wouldn’t mind using fitness trackers and smartwatches to give our 9–5 lives a boost as well. The big question is whether we’re ready to give our employers access to this information, unleashing details about where we are, what we do, and with whom we do it. The growth of disciplines such as organizational network analysis and people analytics clearly shows that employers are interested, and every day they look at new and innovative ways to help us quantify what we do at work.

As the market for the quantified employee grows, fed by a focus on productivity, fitness, wellness, and improving the work environment, two key questions come to the fore:

• In what interesting ways can employers use employee-driven IoT technology to solve problems and improve work processes?

• How can employers overcome resistance and persuade workers to willingly become quantified employees?

The IoT's flexibility

For most, the term Internet of Things calls up visions of consumer products such as connected thermostats or smart washing machines, or of companies installing data-capturing sensors to monitor unmanned facilities. Underlying any of these seemingly simple gadgets is an incredible network of sensors, communication technologies, and analytic power. The IoT is a technology architecture connecting the technologies together to perform actions, a way of stitching together many different types of technologies in a specific way in order to do something new. Regardless of the specific technologies in any particular IoT application, the architecture—the way those technologies are connected—is described by the Information Value Loop.

The fact that no particular device or piece of technology defines the IoT makes it remarkably flexible. IoT technology has applications for consumer wearables and in the home—and in every industry and sector, both visible and behind the scenes. That wide range of uses is driving serious growth projections: Research suggests that manufacturers shipped 1 billion IoT devices in 2015, and analysts predict that the market will grow over 3,000 percent in four years. Already, organizations are placing IoT-based devices far and wide: cameras that watch traffic, stress gauges that monitor bridges, thermostats that monitor and regulate temperature—almost everywhere but in the office.

Reflecting executives’ increasing recognition of IoT technology’s demonstrated value, analysts estimate that almost 60 percent of the market will end up in corporate or business applications. For example, IoT sensors placed throughout a factory can determine when machines require maintenance or alert plant managers if temperature or humidity levels are too high for sensitive processes such as painting or mixing ingredients. One chemicals company struggled with unplanned downtime due to multiple failures of process equipment, occurring 90+ times per year, hurting production, driving up overtime labor costs, and frustrating workers. Once the company installed IoT sensors—along with predictive data analytical models—it reduced equipment downtime by 80 percent.

In business as well as consumer applications, increasing efficiency is only one angle for IoT technology, and it seems a natural next step to look to instrument and augment human workers themselves. After all, plenty of employees already wear smartwatches and carry smartphones—and, when in the office, don lanyards with ID badges that open doors and allow access.

The rise of people analytics

As the quantified self enters the workplace, leaders must contend with a deluge of data, determining which information is most relevant and useful—and what data-driven responses are in workers’ best interests as well as management’s. Increasingly, organizations are practicing people analytics.

Of course, analyzing worker performance is hardly new—companies have looked to measure employees for more than 200 years, long before the advent of HR departments and industrial psychologists. In the late 19th century, mechanical engineer Frederick Winslow Taylor set out to measure the movement and behavior of ironworkers in steel mills, aiming to figure out what would increase organizational productivity. As an industrial engineer, Taylor thought about people primarily as machines and therefore measured factors such as the amount of weight they could carry and what path they should follow to optimize throughput at the plant. He concluded, for example, that for maximum efficiency, grown men should carry exactly 65 pounds of pig iron: More weight causes fatigue and accidents, while less wastes time. Taylor’s work typified business thinking in the industrial age, when few managers took a holistic view of their employees.

Taylorism fell out of favor by the 1970s—about the time when computers began taking over the workplace, aiding payroll and HR management systems in particular. HR departments started warehousing data, and many large companies began looking at the relationship between pay and turnover, patterns of progression for high-performing leaders, and other measures of organizational success. Computing power and sophistication have, of course, exploded since the early days, and HR is finally taking advantage: Adoption and maturity of people analytics have nearly doubled this year alone.

HR strategists originally conceived people analytics as a way to correlate information such as employee engagement, performance ratings, and other work-related activities, and IoT technology is fundamentally changing its practice. By gathering data about workplace activities that were previously invisible to both managers and employees alike, companies are able to use an ever-widening range of information to help make the business run better.

Will IoT technology deliver value to employees?

If IoT applications have the power to fundamentally alter how organizations measure and improve themselves, they also bring challenges that companies may not yet be equipped to handle.

Experience shows that workers worry that their employers may use personal monitoring against them. Research shows that almost 75 percent of employees believe their employer is capturing data about them without their knowledge, so people are wary and a little worried. Consider the UPS delivery system discussed earlier: Yes, tracking has increased efficiency, saved time and money, and reduced environmental impacts, but for many drivers, the constant reminders from managers to brake less often and avoid idling or reversing the truck felt like Big Brother surveillance, ultimately reducing employee engagement. While the drivers’ contract prohibits managers from disciplining drivers based on telematics data alone, monitoring inevitably generates frustration and stress.

This situation is hardly unique to delivery drivers. In every industry, if employees feel that a new technology or management system provides no real benefit to them, they may avoid using it or even actively undercut its adoption. The challenge, then, is to design workplace IoT applications to offer employees obvious, tangible value—even though developers’ ultimate aim may be to improve performance efficiency. Instrumenting workers is fundamentally different from attaching sensors to machines.

Consider sensor-data driven auto insurance rates. Insurers offer dongles that plug into a car’s onboard data port and report back data on speed and braking velocity; this data helps set rates based on individuals’ real driving performance rather than relying on generic actuarial tables. The benefits of a more accurate valuation of risk are clear to the insurer, less so to the insured. While some particularly cautious drivers might expect a reduction in rates, others will likely suffer a corresponding increase, and since no driver can be certain of paying less, most consumers will see no clear benefit from adopting telematics-based insurance. In fact, because people generally overvalue potential losses, there may even be a slight disincentive, which may help explain why 47 percent of the US driving population is skeptical of this type of insurance under any circumstances—and why insurers looking to accumulate more data may have to entice customers with discounts, at least at first. (For much more on auto insurance and IoT technology, see our article Opting in: Using connectivity to drive differentiation.)

The only segments proactively taking advantage of telematics-based insurance are those that do see some benefit from it: In the United Kingdom, teen drivers, with insurance rates nearly three times the average, have plenty of incentive to adopt the technology—so long as they do in fact drive responsibly.26 (For more detail on how perceptions of value impact technology adoption, see our article Power struggle27 or listen to the podcast.28) To effectively use IoT technology—not only in the workplace but everywhere that requires user buy-in and compliance—it is important to make clear the benefits of that technology to users, whether customers or employees.

And framing an IoT-aided efficiency gain as an employee benefit doesn’t have to be difficult: If instrumenting UPS delivery drivers generates data that reduces their driving time by even one minute per day, that translates into a $14.5 million annual savings.29 If a company were to pass on even a portion of such savings to employees as a bonus or salary increase, it would likely help align drivers’ motivations with organizational priorities, making monitoring a little less ominous.

And value to the employee hardly needs to be limited to financial rewards. There’s obvious, visible value in making a job easier, faster, or safer. Take firefighters, for example: By instrumenting each firefighter, not only can on-scene commanders more effectively deploy people at a fire—an IoT-driven system can automatically, instantly alert individuals to conditions those people might not sense or to dangers such as an impending building collapse. As firefighters approach a crash involving an IoT-connected car, the vehicle’s sensors could communicate with the firefighters’, helping make real-time decisions about rescue procedures. Very few of us work just for money, and IoT applications can offer very real opportunities for improving work both in the office and in the field.

Designing a quantified work environment

In or out of the workplace, data fuels the IoT, and the value of the technology and applications depends on that data being useful, valid, and plentiful. Too little information—or the wrong information—can generate misleading and unhelpful analysis; too much data poses the drinking-from-a-firehose problem, leaving managers staring at columns of figures with little sense of how to prioritize or act.

With employee buy-in, IoT sensors can offer data analysts an embarrassment of riches: information as simple as how many daily steps a worker takes (and where and when) to complex sociometric measurements of her emotions and tone of voice in different workplace situations. Before equipping an entire workforce with smartwatches and connected ID badges, you should carefully consider your goals: What problem are you trying to solve? Begin with the business problems you want to solve first, and then decide what data you need.

Now: How to get the data you need? Again, many employees are reticent—justifiably so—about becoming quantified, giving their bosses unchecked IoT-aided access to their movements, meetings, and conversations. As U.C. Berkeley management professor Morten Hansen puts it, “The quantified self is perverse.” Promises of confidentiality notwithstanding, once potentially compromising data is in the system, can any employer guarantee that the information will never, ever be invoked during an annual review? Hansen recommends that employers aim for “less data, less but better feedback, focus and selective behavior change. Quality, not quantity.”

Consider one employee monitoring study, at a large American insurance firm, in which some employees were monitored while performing their daily job duties while others were not. When it came time for workers to review their own individual performance, 80 percent of those in the monitored group identified production quantity as the most significant factor in their self-evaluation, while 85 percent of the unmonitored employees singled out quality of customer service and teamwork. Focusing on metrics actually lowered service quality.

If the company had simply announced, “We are going to start a project to understand how we can help you improve productivity and service, by analyzing what types of support and activities create the most value,” employees might have responded differently. Indeed, they may have been willing to collaborate by suggesting particularly relevant IoT-driven metrics, based on their daily work experience.

Remember also that accumulating high-quality data often requires capturing information over an extended period of time. During certain times of the year, month, or season, employees’ responsibilities may radically shift—not to mention their behavior near the end of a stressful quarter. Daily workplace activity and performance during the summer may scarcely resemble those in the fall, meaning that you may need to collect data over as long a period as possible.

As professionals in the people analytics or employee engagement area know all too well, statistical validity and reliability are paramount—and an ongoing challenge. Suppose you collect data about employee travel and work behavior and come to a conclusion about what drives sales productivity or another business metric. If the data set is not statistically reliable and tested, you may discover only that “high-performing salespeople like to travel a lot”—which may or may not be valuable information. As always, correlation is not causation.

Another reason to carefully plan data collection: avoiding the common problem of multiple systems within an organization winding up with inconsistent, incorrect, and duplicated information. If you do decide, for example, to monitor employee location data or email metadata, it’s important to work with IT to ensure that you have a single, integrated source with solid privacy and security tools, along with strategies for data storage and distribution. Many HR organizations have not yet developed policies and procedures for data governance, which will likely be increasingly important as the current trickle of data becomes a flood. You should have a clear process for securing data, managing access, and holding people accountable for security and quality standards.


This article was originally posted by Deloitte and can be found here.

Feature New Record