When a business invests in tools, whether digital or not, it does so in the hope of becoming more efficient. The equation seems simple: by automating certain tasks, streamlining access to information, and digitizing processes, you free up time, reduce costs, and improve performance. However, despite years of effort in this direction, it is clear that the promise is not always kept, far from it. Productivity is stagnating, teams do not necessarily feel less overwhelmed, and the results do not always match the investments made.
In short:
- Digital investments do not always lead to noticeable productivity gains, as the benefits can spread or be diluted throughout the organization without translating into tangible performance.
- Productivity gains generated by digital tools can be absorbed by secondary tasks, adaptation, control, or coordination efforts, limiting their impact on employees.
- Some of the benefits of digitalization shift to other players, such as customers, who take on certain tasks, or technology providers, who capture a significant share of the value.
- Technological advances are redefining standards, making gains invisible as they become normal expectations, with no perceived improvement for employees.
- A systemic approach is needed to understand, measure, and guide productivity gains, taking into account their distribution within and outside the organization.
A well-known economic logic
This impression of a gap between the power of digital tools and their actual impact on performance is not new. It is part of a well-known economic phenomenon that is often overlooked in discussions about digital transformation: the diffusion effect of productivity gains.
In the 1950s, French economist Alfred Sauvy developed the trickle-down theory, which states that productivity gains in one sector do not remain confined to that sector. They lead to a transfer of resources (capital, labor, attention) to other activities and spread throughout the economy: lower prices for consumers, increased purchasing power, and the creation of new jobs elsewhere. In short, the benefits of productivity gains do not materialize solely or entirely where expected.
This principle was extended in the 1980s by Robert Solow, winner of the Nobel Prize in Economics, who highlighted a now famous paradox: You can see the computer age everywhere but in the productivity statistics. This sentence sums up what would later become known as the Solow paradox. He pointed out that massive investments in information technology did not generate, at least in the short term, clearly measurable macroeconomic productivity gains.
This paradox has been further explored in more recent work, notably by Erik Brynjolfsson, who showed that simply deploying technology is not enough. It is not technology itself that creates value, but the organizational, managerial, and cultural transformations that accompany it. In other words, productivity gains are neither automatic nor captive: they are the result of a process, they spread, and sometimes they are lost completely.
This is something to keep in mind for businesses that bear the entire burden of the effort and investment but will sometimes capture only a fraction of the gains.
Digital technology, a contemporary extension of the spillover effect
Digital technology is no exception to this logic. The tools we deploy today in organizations, whether artificial intelligence, automation, collaborative platforms, or digital workplace solutions, do generate productivity gains, but these gains are not confined to the business, which does not capture them in their entirety. They circulate, they sometimes escape, and understanding their trajectory also means understanding why digital transformation does not always deliver on its economic promises.
Gains that escape the business… and the employee
Within the business, we first observe that the expected gains in operational efficiency do not systematically translate into tangible performance. Tools enable us to do more, faster, or differently, but the time freed up is very rarely synonymous with a reduction in workload or value creation.
It is often absorbed by low-value-added tasks that multiply alongside digitalization, such as reporting, coordination, and managing ever-increasing information flows. It can also fuel an inflation of meetings, which remain the dominant (and rarely optimized) channel for communication and decision-making.
Another part of this gain evaporates in what could be called the “hidden cost of digital tools“. To use a tool effectively, you first have to learn how to use it, adapt to its ergonomics, understand its limitations, manage its notifications and updates, and navigate its logic, which can sometimes be very different from what you are used to. This learning curve is rarely taken into account, but it consumes a considerable amount of energy, especially for tools that are supposed to save time.
Finally, there is an increasing amount of monitoring. As automation increases, it becomes necessary to check what the technology produces: rereading AI-generated content, correcting classifications, ensuring consistency between systems, reconfiguring, testing, and validating. This verification work, which is often fragmented and little recognized, mobilizes a large amount of cognitive resources and generates a form of operational fatigue that is difficult to objectify.
In short, even when tools work properly, productivity does not automatically translate into value creation for employees. It is diluted in micro-tasks, friction, and constant adjustment efforts. And if nothing is done to rethink how they are actually used, they end up increasing the workload rather than improving performance.
A transfer to the customer
Furthermore, some of the gains generated do not remain within the business, nor are they captured as such by other players. Instead, they result in a shift in the workload, particularly to customers. For example, the automation of certain tasks (booking, tracking, problem solving) certainly improves the fluidity of the experience, but also relies on a transfer of workload. What was previously done by an agent is now done by the customer themselves, with no apparent reduction in cost or price. This is referred to as invisible outsourcing, where the user experience is partly based on shifting work downstream.
In a way, what initially appears to be a gain for the business can backfire with customers who are dissatisfied with having to do some of the work themselves in exchange for lower prices.
Capture by suppliers
Another form of capture is that of technology suppliers. Businesses that invest in digital tools not only create value for themselves, but also contribute to the economy of publishers, integrators, and platforms.
This logic is particularly visible in the current context of generative AI.
While many businesses are experimenting with these technologies, the potential gains are far from guaranteed for those who implement them, but they are already tangible for those who sell the infrastructure: Nvidia, Microsoft, Amazon, and OpenAI.
Nvidia alone embodies this phenomenon: as a GPU supplier, it captures a massive share of the value created by a still-fledgling ecosystem. We are talking here about gross margins that are significantly higher than those of companies developing downstream products for uses that are still uncertain.
This is sometimes referred to as the law of value leakage or, more commonly in Anglo-Saxon literature, the “value chain squeeze“. It has been popularized by a number of similar quotes about the gold rush (“During the gold rush, it wasn’t the gold miners who got rich, but the sellers of shovels and picks.?”) that has been attributed to many people, but it seems that it can most reliably be attributed to André Kostolany, a stock market columnist:
”In any gold rush, don’t invest in the gold miners, but in the sellers of shovels and picks.”
This illustrates what is sometimes called the “law of the shovel sellers” or the low-end economy: those who provide the infrastructure in a rush for innovation capture the rent, while downstream users or developers bear the risks, uncertainty, and most of the adaptation work.
The business model for SaaS tools is based on a similar logic, albeit less spectacular. The value generated by internal productivity gains is transformed into monthly subscriptions, paid options, integration costs, support services, and the tool becomes an indispensable infrastructure, but also a channel through which part of the profits escape, regardless of whether all the features are used or not, or whether or not a profit is made..
When progress becomes a requirement
Added to this is a more subtle but nonetheless significant phenomenon: productivity gains are changing expectations. What was once considered an improvement quickly becomes standard, then a requirement.
The norm shifts and progress becomes invisible because it is taken for granted. What was gained yesterday in terms of fluidity or speed becomes an obligation today to be constantly responsive. Employees do not see the gain but, on the contrary, accept the new pace.
Thinking about productivity from a systemic perspective
That is why it is essential, in any digital transformation initiative, to look beyond an internal, linear interpretation of the expected benefits. Productivity gains are not stable, localized data. They are the product of a chain of interactions and are redistributed throughout the organization, its ecosystem, and sometimes far beyond. This means that when measuring the impact of any productive transformation, we need to be extra careful about how things are used, how they’re managed, how their impact is measured, and also how we can track or even control how these gains are captured or lost.
Bottom line
The question is therefore not simply whether a tool works or is well adopted, but who benefits from the gains we generate and how we can ensure that they actually contribute to improving the organization’s performance.
Image credit: Image generated by artificial intelligence via ChatGPT (OpenAI)
Related articles:
Digital transformation: customer experience isn’t everything
The second machine age
Industry 4.0 : the dark side of the digital enterprise
Digital, talent and automation: what matters for executives in 2021 ?
Employee self-service: how far to go before you go too far
Productivity: what if quality was the new quantity?
