As growth slows and cost pressures mount, corporate attention is turning back to productivity. Productivity today, however, cannot be improved simply by extending working hours or reducing headcount. An era has begun in which companies must redesign the entire structure of work, from meetings, reporting, organizational structures, and performance evaluations to the use of AI.

***

Productivity has returned to the center of corporate management. For a time, companies viewed growth primarily through the lenses of market expansion, investment, and talent acquisition. They believed that finding new customers, expanding into new business areas, and hiring more people would naturally increase both revenue and organizational influence. Abundant liquidity, low interest rates, and rapidly growing digital markets supported these expansion strategies.

Management practices designed around the assumptions of low interest rates and high growth no longer function in the same way. Interest rates, raw material prices, wages, and energy costs have risen, while geopolitical conflict and supply chain instability have become permanent features of the business environment. Consumers have become more cautious about spending, and companies in mature markets can rarely expect rapid growth without taking customers from competitors. Revenue no longer rises as it once did, but labor expenses, system operating costs, marketing budgets, and financing costs continue to accumulate. Companies have therefore begun moving away from growth driven by ever-greater resource inputs and asking how effectively they are using the resources they already possess.

The renewed focus on productivity is not merely a passing trend. OECD productivity indicators show that labor productivity growth in many member countries has slowed noticeably compared with the early 2000s. Productivity is still rising, but the pace of improvement has weakened, while disparities among countries and industries have widened. Although digital technology, automation, cloud computing, and artificial intelligence have spread rapidly, economy-wide productivity has not increased as much as expected. Companies have adopted large amounts of technology without necessarily changing the operating systems needed to convert that technology into performance.

Productivity is commonly understood as the amount of work one person completes within a given period. Corporate productivity, however, is far more complex. It must consider how much added value is created from the same labor and capital, as well as how closely work is connected to customer satisfaction, revenue, quality, and innovation. A person who produces ten reports is not necessarily more productive than someone who produces two. If those two reports lead to important decisions, the value of the output matters more than its quantity.

Productivity in an era of slow growth is less a question of “how much more” than of “for what purpose.” If an organization is busy but fails to produce results, it should examine the structure of work before questioning employee attitudes. If employees work hard but decisions remain slow, the problem may lie in approval procedures and the distribution of authority. If meetings continue to multiply without improving collaboration, the purpose of those meetings or the composition of their participants may be flawed. If digital tools have increased while employees feel even more pressed for time, the organization must determine whether technology has reduced work or merely created additional tasks.

The recent productivity obsession resembles the scientific management practices of the past, but it is also fundamentally different. In the past, the core objective was to divide and standardize work to increase output per unit of time. Tasks were broken into smaller parts, like work on a factory conveyor belt, and the fastest method was identified and repeated. Today, knowledge work and service industries account for a much larger share of the economy. Performance depends on activities that are difficult to translate immediately into numbers, including judgment, creativity, customer relationships, and collaboration across departments. Simply increasing speed is no longer sufficient. Moving faster in the wrong direction only increases waste.

The return of productivity should therefore be understood not as a return to greater work intensity, but as a return to management design. Extending working hours and raising performance targets may be the most visible responses, but they are difficult to sustain. What the low-growth era requires is not an organization that pushes people harder, but one that identifies unnecessary work and concentrates time and judgment on what matters.

The greatest paradox of the modern workplace is that everyone is busy while important work continues to be delayed. Calendars are filled with meetings, and messenger notifications never stop. Employees process countless emails and documents but feel that they have too little time to examine customer problems deeply or develop new ideas. The amount of work being done has become disconnected from the amount of value being created.

Digital collaboration tools have substantially reduced the spatial limitations of work. Documents can be shared in real time, and meetings can be held instantly with people in distant locations. Yet as the cost of communication has fallen, the volume of communication has exploded. Information that once might not have been shared at all is now circulated, while content that only a few people need to know is distributed throughout the organization. Because anyone can easily arrange a meeting or send a message, little consideration is given to the cost imposed on other people’s concentration.

The term “infinite workday,” introduced by Microsoft through its analysis of workplace data, captures this situation well. Work no longer ends within defined working hours. It begins with checking email early in the morning and continues through late-night messages and meetings. Workplace tools have removed the barriers of time and location, but they have also erased the boundary between work and rest. Greater connectivity does not increase productivity at the same rate. More communication may support collaboration, but it also fragments concentration and intensifies decision fatigue.

The problem is that organizations easily mistake visible activity for performance. The number of meetings attended, the speed of email responses, and the volume of documents produced are easy to measure. In contrast, it is difficult to measure time spent thinking deeply about a complex problem, discovering an unspoken customer need, or transferring valuable knowledge to a colleague. When easily measurable indicators are incorporated into performance evaluations, employees begin managing visibility rather than value. The ability to appear busy can displace the ability to complete genuinely important work.

Reporting creates another productivity illusion. Reports are necessary to share information and support responsible decision-making. As organizations become more fearful of uncertainty, however, the number of reporting stages and required formats increases. Employees spend more time explaining the status of work than actually advancing it. Senior managers receive information but postpone clear decisions and request additional materials. Reports become longer and more polished, while decisions become slower.

The same applies to meetings. If it is unclear whether a meeting is intended to gather opinions, share information, or make a final decision, participants enter with different expectations. Discussions become unnecessarily long in meetings where information alone would have been sufficient, while decision-makers defer conclusions in meetings where decisions are essential. As the number of participants increases, speaking time per person declines, even as the number of interests requiring coordination rises. A large number of meetings is not necessarily evidence of active collaboration. It may instead signal that ordinary roles and decision rights are poorly defined.

Internal approval procedures are another major source of lost productivity. Approval systems are designed to reduce risk, but when too many stages are added, they create structures in which no one accepts responsibility for the final outcome. Employees say they acted according to instructions, middle managers say senior executives approved the work, and final decision-makers argue that they were not given sufficient information. Many people may have reviewed the work, yet no one may have exercised deep judgment.

The cause of work overload does not lie solely in the time-management abilities of individuals. Overload arises when an organization cannot decide what it will not do. New projects are easily added, but existing work rarely disappears. Every shift in executive attention produces new reports and committees, while previous systems remain in place because they were never formally abolished. Employees are expected to carry out yesterday’s priorities and today’s priorities at the same time.

Productivity innovation therefore cannot begin by instructing employees to work more efficiently. Organizations must first reduce the confusion, duplication, and unclear decision structures they have created. They must protect individual concentration, assign clear purposes to meetings and reports, and establish mechanisms for eliminating work at the organizational level. Productivity depends more heavily on an organization’s ability to choose than on an individual’s diligence.

When productivity pressure rises, reducing headcount is often the first and easiest option for companies. If the number of employees falls while revenue remains unchanged, measures such as revenue and operating profit per employee can improve quickly. The effect also appears immediately on financial statements. If headcount reduction is not accompanied by a reduction in work, however, the burden simply shifts to the employees who remain.

If the number of employees falls but meetings, reports, approval procedures, and projects remain unchanged, each person must take on more work. In the early stages, the organization may continue operating through heightened tension and additional effort. Over time, fatigue and errors accumulate, while customer service and quality control deteriorate. Key talent leaves, and remaining employees avoid experimentation. Short-term cost savings can damage the organization’s long-term productive capacity.

The first question in productivity improvement should not be “Whom should we eliminate?” but “What work should we eliminate?” Companies must identify tasks that continue only through habit after their original purpose has disappeared, duplicate materials produced separately by multiple departments, data that are entered but never used, and meetings that repeat without producing decisions. When work disappears, both time and costs decline. When only people disappear, the work is transferred to those who remain.

Work redesign also differs from simply accelerating current procedures. Even if an unnecessary ten-stage process is automated and completed ten times faster, productivity has not truly improved if the process itself serves no purpose. Before improving efficiency, the organization must confirm why the work exists. If a task cannot be shown to provide value to customers or managerial decision-making, it should be reduced or eliminated.

When redesigning work, organizations should examine the individual tasks that make up a job rather than attempting to automate or eliminate the entire occupation at once. A single job may combine repetitive data entry, rule-based review, coordination with stakeholders, judgment in exceptional situations, emotional communication with customers, and the design of new solutions. Some of these activities are well suited to automation, while others require human experience and responsibility.

The International Labour Organization has also emphasized the transformation of tasks rather than the wholesale replacement of occupations in its analysis of generative AI. Even when a job is highly exposed to AI, the entire occupation does not necessarily disappear immediately. As activities such as document preparation and information organization are automated, the composition and relative importance of the work performed by humans may change. The central challenge of productivity innovation is therefore not to place humans in competition with technology, but to design a new division of responsibilities between them.

Consider the introduction of AI into customer service. Technology can handle simple inquiries and repetitive instructions, but people remain necessary for complex complaints, exceptional circumstances, and problems involving customer emotions. If a company simply reduces the number of service agents, the customer experience may deteriorate. If AI instead organizes relevant information for agents and allows them to concentrate on judgment and relationship-building, both response speed and service quality can improve.

Managerial work also requires redesign. In many organizations, managers spend most of their time attending meetings, compiling reports, and relaying requests from senior management. Their primary responsibilities?clarifying team goals, removing obstacles, and strengthening employee judgment?are pushed aside. If AI and automation reduce information collection, scheduling, and standardized reporting, managers can devote more time to coaching, setting priorities, and resolving conflicts across departments.

Organizational structures should likewise be examined according to the flow of work. If a single customer request must pass through several departments, the total processing time may remain long even when each department works quickly. Department-level productivity may appear high while productivity as experienced by the customer remains low. Organizations must redesign work according to the end-to-end flow of value rather than optimizing individual departments in isolation.

Work redesign also requires rules for stopping. Many companies have procedures for starting new work but none for ending existing work. If the approval of a new project requires a simultaneous decision about which existing task will be discontinued, priorities become clearer. Reports, indicators, and committees that have not been used for a specified period can also be reviewed automatically. The ability to stop is as important to productivity as the ability to begin.

Generative AI is the strongest force reigniting the productivity debate. Expectations have spread that the time required to draft documents, summarize materials, generate code, and classify customer inquiries can be reduced dramatically. Companies hope to use AI to create more output with the same workforce or to maintain existing services at a lower cost.

AI does, in fact, excel at accelerating certain tasks. It can provide immediate assistance in language- and information-intensive activities such as search, organization, translation, drafting, and the production of documents in repetitive formats. It can also offer basic frameworks to less experienced employees while reducing the time skilled workers spend on routine tasks.

Yet an individual completing a task faster with AI is not the same as an organization becoming more productive. Even if one employee produces a report draft twice as quickly, the final decision may not be made much faster if the review and approval stages remain unchanged. Because AI makes documents easier to produce, the number of reports and proposals may actually rise, increasing the burden of review. Faster production can create a new bottleneck in the form of excessive output.

The cost of verifying AI-generated work must also be considered. Even when the language sounds natural and the format appears complete, factual errors may remain or important context may be missing. If experienced employees must verify the material again to detect mistakes, some of the time saved will be consumed by validation. If errors reach customers or the market, the company may face the much greater costs of quality failures and damaged trust.

AI productivity therefore cannot be calculated solely by asking how quickly something was produced. The equation must include production time, verification costs, the likelihood of error, and the value created by the final result. If drafting time is cut in half but editing and verification time doubles, real productivity has not improved. Conversely, even if the number of outputs remains unchanged, productivity may have risen if customer responses become more accurate and decisions improve.

This is why AI adoption must expand from the level of individual tools to the level of complete workflows. Providing employees with AI accounts and encouraging them to use the technology does not change the organization’s operating system. Companies must define the stages at which AI will collect information, who will validate its output, under what conditions people will intervene, and who will hold final responsibility. What is needed is not merely a set of technology usage guidelines, but a new operating model for work.

The criteria for selecting AI-appropriate work must also be clear. Tasks that are high in volume, repetitive, and relatively easy to verify offer strong potential for automation. In contrast, work that occurs infrequently, varies substantially by situation, and carries a high cost of error requires human review and responsibility. In areas such as healthcare, finance, law, and human resources, where a single error can significantly affect individual rights and organizational trust, verifiability and accountability must take precedence over speed.

How the time saved by AI is used is another important productivity variable. If the saved time is filled with new meetings and reports, work intensity may simply increase. If it is used to understand customers more deeply, help employees acquire new skills, and solve complex problems, the organization’s long-term capabilities can grow. The value of AI lies not only in reducing time, but in moving human time toward more valuable activities.

Productivity in the age of artificial intelligence should be assessed not by how much work one person completes, but by how well people and technology make decisions together. AI can produce answers rapidly, but it does not automatically determine goals, reconcile conflicting interests, or assume responsibility for outcomes. As technology becomes more capable, people need stronger skills in framing questions, judging results, and accepting responsibility.

Productivity improvement requires measurement. An organization that does not measure cannot easily determine where time is being wasted or what has improved. If productivity is reduced to overly simple numbers, however, an organization may improve its indicators while damaging its real performance.

If call center employees are evaluated only on shorter call times, they can handle more customers per hour. But if customer problems are not fully resolved, the same customers may need to call again. The efficiency of the first call improves, while the total customer experience and overall processing costs deteriorate. If salespeople are judged only on short-term revenue, they may sell unnecessary products or rely excessively on discounts. Current figures improve, but long-term customer relationships and profitability weaken.

Measurement is even more difficult in knowledge work. New ideas and strategic judgment cannot be evaluated easily through input hours or the number of outputs. A failed experiment can produce valuable knowledge, while research that creates no immediate revenue may become a core capability several years later. When productivity is connected only to short-term results, employees choose safe work with a high probability of success and avoid uncertain but potentially valuable initiatives.

The value of collaboration is also poorly reflected in individual performance indicators. Employees who help others, share information, and train junior colleagues may produce fewer short-term outputs of their own. If evaluations focus exclusively on individual performance, withholding knowledge becomes more advantageous. Organizational productivity declines even as individual indicators improve.

As remote and hybrid work expanded, some organizations attempted to use observable activities such as screen time, message volume, and keyboard movement as proxies for productivity. Remaining online for long periods, however, does not mean that important work is being completed. Such surveillance signals distrust and encourages employees to manage activity records rather than perform meaningful work. A system introduced to verify productivity may instead reduce it.

Good productivity indicators must consider quantity and quality, the short term and the long term, and individuals and organizations at the same time. Organizations should examine not only processing volume but also error rates and customer satisfaction, and they should assess short-term sales alongside repeat purchase or retention rates. Individual output should be considered together with contributions to team goals and knowledge sharing. It is safer to examine several indicators that exist in tension with one another than to compress everything into a single number.

The purpose of measurement should also be learning rather than control. When indicators are used only as a basis for punishment, employees defend the numbers and conceal unfavorable information. When they are used to discover bottlenecks and improve work, problems can be exposed early. Productivity data should not be a tool for identifying who is lazy. It should be a map showing where work is becoming obstructed.

Organizations must also recognize the value of unmeasured space. Time spent exploring information that appears unrelated to immediate responsibilities, speaking with people in other departments, or considering new possibilities may look like waste according to short-term indicators. Innovation, however, does not emerge only from repeating established processes more quickly. It also arises from viewing old problems differently and combining previously disconnected ideas.

When productivity management becomes excessive, every hour is tied to a target or indicator. Employees concentrate on work that can be evaluated immediately rather than on experimentation and learning, while managers avoid approving initiatives that might fail. The organization becomes efficient at repeating the present but loses the ability to prepare for the future. Efficiency and innovation are not opposites, but maximizing efficiency alone can eliminate the space required for innovation.

Managers play a major role in productivity innovation because they must resolve work overload and competing priorities. Yet many managers interpret productivity as an employee attitude problem. When work is delayed, they assume that employees lack urgency and require more frequent reports. Employees then spend more time explaining progress than advancing the work itself.

A productive manager does not merely add work but selects it. Managers must translate organizational goals into concrete priorities, block requests that do not support those goals, and define the scope of authority so that employees can exercise independent judgment. When every issue requires a managerial decision, consistency may improve, but decision-making slows and the manager becomes a bottleneck.

Delegation does not simply mean telling employees to handle matters on their own. Managers must explain the desired outcome, the standards that must be upheld, and the acceptable level of risk. The clearer the boundaries of the work, the more confidently employees can make decisions without waiting for approval at every stage. Productivity depends less on the speed of instructions than on where judgment is allowed to occur.

Performance management also needs to shift from annual evaluation toward continuous adjustment. If goals established a year earlier remain fixed until the end of the year, they may no longer reflect changes in markets and customer needs. Employees continue pursuing objectives that have lost importance simply because those objectives remain tied to their evaluations. Goals should be reviewed at shorter intervals and formally revised when priorities change.

A good goal communicates not only what must be done but also what no longer needs to be done. When there are too many goals, the organization effectively has none. If everything is declared important, employees deal first with requests from the loudest person or tasks with the nearest deadline. Urgency captures resources that should be directed by strategic importance. The number of core goals should be limited, and existing goals should be adjusted whenever new ones are added.

Psychological safety is also important in productivity improvement. Waste persists when employees identify inefficiencies but feel unable to question established systems or decisions made by senior managers. People must be able to speak honestly about the possibility of failure and excessive workloads before problems can become visible. In organizations where only positive reports travel upward, executives cannot accurately assess real productivity.

Managers should examine the results and learning that employees produce rather than the amount of time they appear to spend working. At the same time, they should not demand results while ignoring constraints in the process. Systems, approval structures, and resource shortages that obstruct goal achievement must also be addressed. When performance is weak, it is the manager’s responsibility to examine not only individual ability and effort but also flaws in work design.

Productivity pressure is difficult to avoid in an era of slow growth. Companies must create greater value at the same cost, while governments and societies must find ways to maintain living standards amid population decline and labor shortages. Productivity improvement itself cannot be rejected. The central questions are what kind of productivity is being pursued and who bears the costs of achieving it.

Reducing headcount and increasing work intensity can produce rapid results. As fatigue, turnover, quality deterioration, and reduced learning opportunities accumulate, however, productivity declines again. Efficiency achieved by exhausting people is not sustainable. Sacrificing tomorrow’s capabilities to increase today’s output is closer to shifting costs than improving productivity.

Sustainable productivity begins by defining valuable work. Every department and occupation should be able to explain how its activities create change for customers and the organization. What matters is not the number of outputs, but the results those outputs produce. The purpose of writing a report is not the report itself but better decision-making. The purpose of holding a meeting is not the meeting itself but reaching necessary agreement and making decisions.

The second condition is the continuous ability to remove unnecessary work. A single organizational restructuring or cost-reduction campaign is not enough. Tasks and rules accumulate again over time. Whenever new systems and reports are introduced, existing work should be reviewed, while indicators and committees that are no longer used should be eliminated periodically. Responsibility should be assigned not only for creating work but also for ending it.

The third condition is designing the roles of people and technology together. Not every task should be handed over to technology simply because it can be automated. Speed, cost, quality, and responsibility must be considered together. AI should handle repetitive information processing, while people concentrate on judgment, relationships, exceptional situations, and accountability. Structures that expand human capabilities are more favorable to long-term performance than those designed merely to displace people.

The fourth condition is recognizing concentration and recovery as parts of work. A knowledge worker’s attention is not an unlimited resource. Constant notifications, meetings, and frequent task switching consume mental energy. Organizations must create meeting-free focus periods, establish times when responses are not expected, and design working environments with clear priorities. Rest is not the opposite of productivity. It is a condition that enables sustained judgment and creativity.

The fifth condition is distributing the benefits of productivity improvement fairly. If technology and work redesign create greater performance but all the benefits go to the company while employees receive only more work, the transformation will struggle to earn trust. Saved time should be used for learning and career development, while improved performance should lead to better compensation and working conditions. Employees participate more actively in change when they also benefit from productivity innovation.

Productivity is both a number and an organizational philosophy. A company’s approach to productivity reveals what it recognizes as valuable, whether it views people merely as costs or as sources of capability, and whether it uses technology for surveillance and control or to support judgment and creativity.

Companies in an era of slow growth must make choices with less room for error than before. There is a limit to maintaining every task while attempting only to increase speed. Organizations must distinguish important work from unimportant work, divide responsibilities between people and technology, and design a balance between short-term performance and long-term capabilities.

The true purpose of productivity is not to make people busier. It is to enable organizations to create greater value for customers and society while spending less time wandering, waiting, and repeating work. If the productivity obsession leads only to increased work intensity, organizations will become exhausted rapidly. If it instead becomes an opportunity to reconsider the purpose and structure of work, the pressure of slow growth can become a force for replacing outdated operating practices.

The standard separating productive organizations from the rest will no longer be how long employees work. Competitiveness will depend on how decisively organizations eliminate low-value tasks and how effectively they invest saved time and technology in important problems. Productivity must be redefined not as a technique for demanding more work, but as the managerial ability to choose better work.

OECD, June 2026, OECD, OECD Compendium of Productivity Indicators 2026

Microsoft WorkLab, June 2025, Microsoft, Breaking Down the Infinite Workday

International Labour Organization, May 2025, Gmyrek, P., et al., Generative AI and Jobs: A Refined Global Index of Occupational Exposure

International Monetary Fund, January 2026, Georgieva, K., New Skills and AI Are Reshaping the Future of Work

Microsoft WorkLab, April 2025, Microsoft, 2025: The Year the Frontier Firm Is Born