As human performance takes center stage, are traditional productivity metrics enough?
In an era of human-centered work, new sources of data and artificial intelligence can help organizations shift from measuring employee productivity to measuring human performance.
Sue Cantrell
Julie Duda
Corrie Commisso
Kraig Eaton
John Guziak
When Japanese tech company Hitachi set out to improve organizational productivity and efficiency several years ago, it decided to experiment with an unconventional approach. This approach didn’t involve seeking ways to squeeze more work out of working hours or reinventing processes to shave minutes or seconds from production processes. It didn’t push workers to produce more with less, and it didn’t require leaders to double down on monitoring every movement of their workforce in search of workers who weren’t carrying their weight.
Instead, Hitachi focused on tracking a single, unexpected metric: worker happiness.
Using wearables and an accompanying mobile app, Hitachi offered participating workers artificial intelligence–based suggestions for increasing feelings of happiness throughout the day by boosting psychological capital (self-confidence and motivation), psychological safety, and alignment with management objectives.1
The early results were stunning. Workers’ psychological capital rose by 33%—a particularly meaningful improvement, given that increased psychological capital results in increased worker engagement, greater job satisfaction, and lower turnover intention and burnout.2 Profits increased 10%. Sales per hour at call centers increased 34%, and retail sales increased 15%.3 What’s more, the majority of participants said they were “happy”4—just one indication that the key to unlocking organizational performance in a rapidly evolving era of work may no longer be tied to traditional productivity metrics.
Hitachi’s focus on measuring and building worker happiness represents a shift away from traditional efforts of gauging and improving worker performance, which tend to focus on activity-centric productivity metrics such as hours worked, time on task, product produced, and revenue per employee. These traditional ways of measuring worker performance as a series of outputs solely reflect the perspective of the organization. New approaches, by contrast, can and should consider the worker as a human being, with a more nuanced perspective on how they contribute to the organization.
Making the leap from knowing to doing (figure 1) is important for organizations that want to thrive in a work environment that is becoming increasingly human. The once clear line that linked individual worker activity (for example, hours worked or calls completed) to tangible outcomes (customer satisfaction or commercial potential of research and development projects) is now blurred, replaced by a complex network of collaborations and a demand for sophisticated skills that aren’t easily observed by traditional productivity metrics. Even in front-line, logistics, and manufacturing environments where traditional metrics like minutes per call or widgets produced may seem most applicable, technology and AI are being increasingly used to automate such tasks. The workforce can then be free to undertake complex problem-solving that requires skills that are less technical and more abstract, such as creativity, critical thinking, and collaboration. In agriculture, for example, autonomous drones can be used to plant seeds, apply fertilizers and pesticides, and check for pests or environmental damage.5 Workers would then be able to spend time learning new skills that can enable them to manage the technology, optimize processes, deal with exceptions, or develop sustainable strategies for crop health and maintenance.
The once clear line that linked individual worker activity to tangible outcomes is now blurred, replaced by a complex network of collaborations and a demand for sophisticated skills that aren’t easily observed by traditional productivity metrics.
At the same time, some organizations are looking beyond traditional metrics such as revenues and profits to consider how they can create shared value—outcomes that benefit individual workers, teams and groups, the organization, and society as a whole. The organizations that successfully navigate this new environment will likely be the ones who make the shift from old methods of understanding productivity to embracing a new paradigm of human performance.
Signals your organization should consider prioritizing human performance metrics
- Your organization primarily measures work output metrics rather than the broader organizational outcomes you’re driving toward.
- Your leaders are overwhelmed by the amount of data available to them and want to focus on measuring what really matters.
- Traditional productivity is relatively flat despite your investments in technology.
- Your workers are engaged in “productivity theater,” in which they do tasks to make themselves appear busy and show that they are being productive.
- Your workers are burned out because of the perception—or the reality—of constant activity monitoring.
Rethinking traditional productivity metrics
Leaders across industries are beginning to recognize the limitations of legacy productivity metrics in the current work environment: Seventy-four percent of respondents in Deloitte’s 2024 Global Human Capital Trends survey said it’s very or critically important to seek better ways to measure worker performance and value beyond traditional productivity. But change has been slow. Only 17% of respondents said their organization is very or extremely effective at evaluating the value created by individual workers in their organization, beyond tracking of activities or outputs.
With new digital technologies providing access to more work and workforce data than ever before, it may seem that shifting to a new system of measurement would be easy to do. Organizations’ ability to track the outcomes of human performance and understand what drives it is supported by exponential growth in their ability to collect, measure, and analyze this data—and, with the help of machine learning or human judgment, convert the data into actionable suggestions. The resources at their disposal for this kind of data collection and analysis include the following:
- Workplace tools and technologies, such as email, collaboration platforms, social tools, and shared calendars, generate passive data that can offer real-time insights into how people and organizational systems are working. A large global oil and energy company analyzed anonymized collaboration data (email, calendar, and conferencing and chat data) to understand how teams in its 500-person corporate law department were collaborating. Aiming to better develop and retain talent, the organization used the findings to redesign the workplace, which resulted in more collaboration.6
- Organizational network analysis can be used to measure connections and collaboration between people across an organization. As part of its efforts to promote more women, a global financial services organization used organizational network analysis to understand the relationship between the size and quality of women employees’ internal and external networks and their chances of being promoted.7
- Sensors and connected devices, such as wearables, badging scans, neurotechnology, biometric sensing tools, extended reality headsets, and precision location-tracking technologies, can generate data on worker behaviors and interactions. For example, when a Finnish railway company shifted to hybrid work and wanted to optimize its physical space more effectively, it used occupancy sensors to detect workers’ movements and use of its spaces. This data helped the organization reduce real estate cost by downsizing building space from five floors to two, while making sure workers were able to move about easily and access critical workplace assets.8
- AI-enabled voice or audio analytics generated from worker interactions with machines and AI systems, such as algorithms that assess code quality or the emotional tone of call center interactions, can offer valuable insights for evaluating various aspects of business operations. At MetLife, where customer service agents field an average of 700 calls a week, AI coaching has helped agents have more “human” conversations, which has increased customer satisfaction by 13%.9
While some organizations are moving ahead, what potential challenges may be keeping others from expanding their view of performance beyond traditional productivity?
Pressure from external stakeholders. Despite their desire to find better ways to measure human performance, senior leaders are currently under pressure from external stakeholders to demonstrate improved productivity and efficiency amid high inflation, shrinking profit margins, and the looming threat of economic recessions.10 As a result, they may become focused on achieving short-term, bottom-line results instead of desired human outcomes (for example, improved worker well-being) that are less tangible.
Uncertainty about what to measure. More data doesn’t automatically equate to better results. Many organizations may find themselves lost in an ocean of data as their ability to collect data outpaces their ability to analyze and act on it. As a result, they may end up with too much data and too little insight, leaving leaders unsure about what metrics are most important and which actions are truly driving performance.
Productivity paranoia. During the COVID-19 pandemic, many organizations were quick to adopt new worker-monitoring tools that tracked keystrokes, mouse activity, and more to gain visibility into who was working on what and for how long—the same productivity standards they’d always tracked. But new ways of working require new metrics. Now, some organizations are finding themselves at odds with workers over this increased monitoring. Productivity paranoia—a concern that remote workers aren’t being productive11—may lead to a surveillance state and a breakdown of trust, instead of important conversations about what effective performance looks like in today’s work environment.
Lack of visibility into outcomes. Many organizations are still focused on measuring worker inputs and outputs rather than outcomes. As organizations begin to measure human performance, they can begin tracking two areas: business outcomes that create value for the organization and human sustainability, or human outcomes (both of which may vary by workforce).
One way forward requires a fundamental rethinking of what measures matter in a workplace being transformed by rapid advances in technology and shifting priorities. If leaders want to realize the human potential in their organizations and enable innovation, the focus should shift from only productivity to a broader view of performance.
A new equation for human performance
The flood of possibilities unleashed by the unprecedented volume of work and workforce data now available to organizations raises an important question: If traditional productivity metrics are becoming less relevant in the workplace, what should organizations be measuring to meaningfully assess human performance and how should these new metrics be operationalized?
The new math involves a balance of business and human sustainability—creating shared, mutually reinforcing outcomes for both the organization and the worker. Business outcomes define the quality, value, or result of work, and how it creates value for the organization. Human sustainability defines the degree to which an organization creates value for people as human beings, leaving them with greater health and well-being; stronger skills and employability; good jobs with sustainable wages; opportunities for advancement; and greater belonging, equity, and purpose.
After all, organizations essentially compete in two industries: the industry it works in and the industry of talent management. Leaders should leverage the connection between human and business outcomes to increase the likelihood of success in both these industries (figure 2).
When an organization uses the data it collects about its workforce to benefit everyone—individual workers, teams and groups, the organization, and society as a whole—it creates shared value. The value created at each level can flow between them, reinforcing and amplifying the value created at other levels.
In the example of Hitachi’s experiment with improving worker happiness, it’s easy to see how creating value at the individual worker level led to value at the enterprise level—increasing both revenues and profits. This is not a zero-sum game: Organizational initiatives that were originally designed to achieve benefits like higher cost savings or improved quality can also help amplify worker satisfaction and performance. For example, a major energy organization recently used workplace badge data to analyze where and how different groups were interacting while planning an office relocation. It found that, as cross-functional teams became more dispersed, they had fewer informal interactions and instead relied too heavily on occasional, formal meetings. The organization used this finding to plan the location of team members during relocation to create more informal connection opportunities, boosting team belonging and workflow efficiency by 5.3%.12
Also consider how this shared value dynamic played out at a large automotive supplier, which deployed AI-powered video analytics to increase its visibility into factory operations. Analysis showed that the configuration of physical stations on the line was slowing down operations and creating fatigue for workers. The organization used these findings to reconfigure the stations, decreasing both idle time and overall production time. The analytics helped the organization make informed decisions that directly impacted worker well-being, while also improving areas such as capacity planning, quality improvement, workforce management, and process engineering—and the plant’s operation product manager also noted improved happiness, health, and productivity in line workers.13
Organizations have a window of opportunity to capture human performance metrics
Despite many examples of work and workplace data being used to drive improved human performance in organizations, the prevailing narrative tends to pit workers and organizations against each other. When it comes to the collection and use of work and workforce data, the typical assumption is often that workers are uniformly opposed to any type of monitoring and executives want to track every metric available, no matter how intrusive. However, Deloitte’s research into the quantified organization suggests that this isn’t necessarily the case: Workers and executives have surprisingly similar views about how work and workplace data can improve outcomes in ways that benefit the organization as well as the workforce.14
The Quantified Organization
Deloitte’s Quantified Organization research delves into what it means for organizations to take a strategic approach to measuring what they should, not just what they can. Through in-depth interviews conducted with senior global business executives, global surveys of 2,000 workers and leaders, and an analysis of more than 50 case studies and 30 distinct use cases, the quantified organization series of research reports highlight how new data sources and AI tools, responsibly used, can create shared value for workers, organizations, and greater society.
For example, workers and leaders largely agree that new sources of data have positively impacted both business and worker outcomes (figure 3).15
In addition, they’re largely in agreement about what sources of data an organization should collect—and which to avoid. For instance, more than three-quarters in each group are comfortable with collecting data from employee emails and calendars. But other data sources, including location-tracking technologies or the review of external sites such as social media and personal emails, give both groups pause.16
This fundamental alignment may point to a critical window of opportunity for leaders to unlock the potential of work and workforce data in measuring human performance. While our recent Quantified Organization research shows a relatively high level of worker trust in their organization’s data collection efforts, it also shows that trust is tenuous: Workers are less confident than leaders that their organizations are using data in a responsible way (70% vs. 93%).17
Transparently communicating how and why data is being collected and used, along with giving workers the option to opt in or out, is important here: It may be hard to imagine a scenario in which workers would object to the use of location-tracking technologies, specifically for safety purposes, such as disabling equipment when someone is standing in a dangerous spot. However, unless leaders continue to invest in building worker trust and creating shared value through their data collection efforts, the window of opportunity may close before organizations can realize the value.
While workers and organizations appear to be more aligned on the use of work and workforce data than one might expect, the use of this data is still complex. When implementing new metrics and using newly available workforce data to capture human performance, organizations should carefully consider what to make transparent, to whom, and how—considering critical factors such as worker consent, providing benefits to the worker, and other responsible data collection practices. These efforts are essential, given the potential payoff: A predictive outcomes analysis of our quantified organization survey data suggests that trust in an organization’s approach to data management raises the probability of improved business growth by roughly 50%.18
Laying the groundwork for a human performance focus
The shift toward using work and workforce data to measure human performance is still in its infancy, as organizations are still determining which metrics are best suited to their industry and their organization’s specific needs. A majority (53%) of respondents agreed that their organization is in the early phase of the journey toward identifying better ways to measure worker performance and value beyond traditional productivity. Just 8% said their organization is leading in this area. But there are steps organizations can take now to lay the foundation for a shift toward human performance metrics.
- Cocreate metrics and solutions with workers. Organizations can build trust in their use of worker data by providing workers with opportunities to provide input into which human performance metrics should be prioritized, as well as opportunities to respond to insights the data may reveal. Consider an example of what this kind of partnership could look like: An oil and gas company used wall-mounted cameras to observe workers and assets at a maintenance and manufacturing facility, and AI turned the aggregated, anonymized video data into insights on patterns of productivity. Workers were involved from the start, choosing to opt-in for data collection, viewing the results of the AI analysis, and collaboratively engaging in problem-solving on how to use the data to improve their experience and results. One set of data insights led employees to modify rest areas and take more frequent breaks to minimize fatigue—decisions that also improved their productivity.19
- Measure what you should, not just what you can. The human performance metrics that matter most to an organization will vary based on industry, geography, workforce, and how the organization currently operates, and will likely require some experimentation to find the right balance of business and human sustainability outcomes. For example, in a call center, productivity is typically measured by things like the amount of time per call or the number of sales made. But when human performance becomes the primary focus, metrics like customer satisfaction, retention, and upselling may give a call center manager a better picture of how their workers are performing. Organizations should continue to focus on the “why” of their data collection efforts, asking themselves: Just because it can be measured, does it really need to be—and if so, why? For instance, metrics in logistics that focus on safety or worker fatigue may not necessarily be the wrong measures but can become more human-centric when they are measured with the intent to improve conditions for workers. Deloitte’s Quantified Organization research revealed that a lack of predetermined strategic goals for using workforce data was related to workers’ lack of trust in the organization’s intentions to collect and use that data for their benefit.20 Creating clear goals for data collection and use that are directly aligned to organizational strategy and objectives can go a long way toward earning and reinforcing worker trust.
- Implement these practices in your performance management approach. Traditional performance management can be a challenging process if there are unclear or unrealistic expectations for workers and opportunities for errors in human judgment. For example, performance reviews that happen only once a year may lead to recency bias, where only a worker’s most recent activities are included in an evaluation. As organizations make the shift toward human performance, an organization’s approach to performance should evolve from management to development. AI tools are poised to help leaders redefine—not just augment—performance. Not only can these tools collect unbiased data to foster fact-based performance reviews, but generative AI tools may be able to play a key role in summarizing and synthesizing multiple sources of data. When leaders are clear with workers about how AI is used in performance reviews, this kind of data-driven system can help maintain transparency and build trust. In addition, AI can act as an additional coach for workers, offering personalized feedback based on their established performance outcomes.
- Integrate new metrics into the processes of other areas of the talent life cycle. As organizations transition to the use of human performance metrics, they should carefully consider how best to leverage this data to better the work, and the experience, of individual workers. Organizations should consider which human drivers to focus on, then calibrate how team leaders discuss those metrics with workers and teams. This process begins with experimentation as organizations and teams uncover which metrics, communicated in which context, create the human and business outcomes they seek.
- Establish responsible data and AI practices. Responsible data practices give workers input on how their individual, personal data is shared across an organization and help organizations comply with the evolving global regulatory requirements around data use. Such practices may include facilitating increased visibility into which type of data is collected and why, respecting privacy and data integrity concerns, and seeking worker consent whenever possible or required. Aggregating and anonymizing data, for example, can help maintain worker privacy. While AI can be a valuable tool for assessing and improving human performance metrics, it can also damage an organization’s reputation and performance if it is not used appropriately. For this reason, organizations should rely on a multidimensional ethical framework to manage AI’s potential risks and rewards.21
- Plan now to address tensions around the use of emerging technologies. While our Quantified Organization research showed that workers are relatively comfortable with data collection from known technologies like email, calendars, and other traditional technologies, they are far less comfortable when it comes to data captured from emerging technologies like wearables and XR headsets.22 Still, a majority of leaders said they expect to implement the use of these technologies for data collection in the coming years (figure 4). This could put leaders and workers at odds and threaten organizational trust. Leaders should plan now for how they will work to bridge this gap, being mindful of worker concerns around privacy and reinforcing the line between professional and personal data collection.
Human performance: An evolving approach to strengthening workers and organizations
It is still early days for channeling the flood of available work and workforce data into meaningful measures of human performance. But the time to act is now. Forward-thinking organizations can cocreate their human performance metrics and the data policies and practices that can measure or identify ways to drive these metrics with workers in real time, fostering trust throughout the process. Failing to do so, whether by imposing policies and practices from the top or continuing to rely on outmoded measures of worker performance, can create potential challenges in talent attraction and retention, unintended consequences to well-being and mental health that productivity paranoia may create, and a potentially disastrous misunderstanding of what factors actually drive the organization’s value creation.
The alternative is far more appealing. As organizations begin threading human performance throughout their practices, they can strengthen business outcomes and make a positive impact on everyone the organization touches.
Research methodology
Deloitte’s 2024 Global Human Capital Trends survey polled 14,000 business and human resources leaders across many industries and sectors in 95 countries. In addition to the broad, global survey that provides the foundational data for the Global Human Capital Trends report, Deloitte supplemented its research this year with worker- and executive-specific surveys to represent the workforce perspective and uncover where there may be gaps between leader perception and worker realities. The executive survey was done in collaboration with Oxford Economics to survey 1,000 global executives and board leaders in order to understand their perspectives on emerging human capital issues. The survey data is complemented by over a dozen interviews with executives from some of today’s leading organizations. These insights helped shape the trends in this report.