McKinsey Quarterly

Adhocracy for an agile age

| Article

Even as companies from IBM to Caesars Entertainment to American Express succeed through advanced analytics and big data, a less visible side of the preoccupation with information may be having the opposite effect.

Academic studies show that information overload at the individual level leads to distractedness, confusion, and poor decision making.1 These problems beleaguer organizations, too, as we have seen from working with many large companies and through many interviews and workshops with senior executives in a range of sectors and geographies. Our experience reveals frequent cases of analysis paralysis (gathering more and more information rather than making a decision), endless debate, and a bias toward rational, scientific evidence at the expense of intuition or gut feel. These pathologies can have a deleterious impact on the functioning of companies. They can lessen the quality and speed of decision making and engender a sterile operating environment in which intuitive thinking is quashed. As a result, many companies end up standing still, even as the world around them is speeding up.

In short, the undeniable power of information brings the risk of becoming overly reliant on or even obsessed with it. What’s more,as the information age advances into an increasingly agile one, something important is changing: information is less of a scarce resource as it becomes ubiquitous and search costs plummet. In such a world, as Herbert Simon speculated more than 40 years ago, the scarce resource we have to manage is no longer information—it is attention.2 We believe that large companies today are poor at managing what might be called their return on attention. Particularly at the executive level, attention is fragmented; people are distracted; and even when the data are impeccable, decisions can be unduly delayed or just plain bad.

Clearly, not everyone has fallen into this trap. At some companies, executives understand both the power and the limits of information; they know that at times, getting the right answer is imperative but that at other times, being decisive and intuitive, and acting swiftly and experimenting, can work better. As Amazon’s Jeff Bezos says, “there are decisions that can be made by analysis. These are the best kinds of decisions! They’re fact-based decisions. . . . Unfortunately, there’s this whole other set of decisions that you can’t ultimately boil down to a math problem,”3 such as big bets on new businesses. Some of Bezos’s bets, like the Kindle and Amazon Web Services, have paid off; others, such as Amazon’s mobile phone, have not. But that hasn’t dissuaded the company from continued experimentation and action.

Clearly, there’s a need for balance—for a more nuanced understanding of when to dig deeper into the data, when to stimulate the kind of extended debate that can help eliminate hasty or biased decision making, and when to act fast.4 In our experience, many companies are more comfortable analyzing and debating than they are acting decisively and intuitively. Their default orientation toward more and better information binds and restricts their ability to move surely and quickly.

The purpose of this article is to suggest a set of capabilities—about how work is organized and people think—to complement the default orientation of companies and to help them manage their return on attention in a more systematic way. These capabilities, we suggest, are part of an organizational model—adhocracy—that differs from the bureaucratic and meritocratic organizational models currently in favor. By clarifying the pros and cons of these three models, and the conditions when each should be used, we aim to provide guidance on how to get the right balance between information and attention.

Three organizational models

The concept of adhocracy was first proposed several decades ago,5 essentially as a flexible and informal alternative to bureaucracy. Here we’re intending to redefine the concept in a way that further distinguishes it not only from bureaucracy but also from the meritocracy model of organization.

Adhocracy’s defining feature is that it privileges decisive (and often intuitive) action rather than formal authority or knowledge. For example, when bureaucracies face a difficult decision, the default is to defer to a senior colleague. In a meritocracy, the default is to collect more data, to debate vigorously, or both. The default in an adhocracy is to experiment—to try a course of action, receive feedback, make changes, and review progress.

Adhocracies are also likely to use more flexible forms of governance, so they can be created and closed down very quickly, according to the nature of the opportunity. By emphasizing experimentation, motivation, and urgency, adhocracy provides a necessary complement to progress in advanced analytics and in machine learning, which automates decisions previously made through more bureaucratic approaches.6 Specifically, we view adhocracy as an organizational model that maximizes a company’s return on attention, defined as the quantity of focused action taken divided by the time and effort spent analyzing the problem.7

The exhibit summarizes the differences among bureaucracies, meritocracies, and adhocracies. The right organizational model, though, often varies according to the business environment in which a company (or a part of it) competes. Bureaucracy still has merit in highly regulated and safety-first environments. Generally speaking, meritocracy works well in, for example, professional-service environments, universities, and science-based companies. Adhocracy is well aligned with the needs of start-ups and companies operating in fast-changing environments. The appropriate model also varies by function, with compliance more likely to be a bureaucracy, R&D a meritocracy, and sales an adhocracy.

That said, executives must carefully weigh their overall approach and the extent to which any of these models should hold sway. A professional-service firm, for example, might take an adhocratic approach to organizing its teams so as to exploit opportunities, even as its professional-development and strategic-planning groups use more meritocratic approaches. The selective application of all three models is a core executive task.

In the rest of this article, we describe adhocracy in more depth, explaining its key features and comparing its pros and cons with those of the bureaucratic and meritocratic models.

Three key features of adhocracy

An adhocracy can be readily observed in many organizational settings. For example, if you go to a hospital’s emergency room or an investment bank’s trading floor, the focus on getting things done rapidly is clear. Many companies have used “skunkworks” operations: small project teams that tackle a one-off problem outside the organization’s formal processes at an accelerated pace. Many small companies have adopted the lean start-up model, which emphasizes early prototyping and pivoting rapidly to new business models as circumstances change.8 In all these settings, informed, decisive action matters more than formal authority or knowledge.

Three key features distinguish adhocracy from bureaucracy and meritocracy. Each helps increase a company’s return on attention.

1. Coordinating activities around opportunities

Bureaucracies coordinate their activities through rules, procedures, and routines; meritocracies, through adjustments based on flows of information. In an adhocracy, by contrast, coordination coheres around discrete opportunities. For example, many companies have experimented with decentralized business units focused on specific customers or projects. One such company is the UK-based pharmaceutical firm GlaxoSmithKline, which has broken its drug-discovery operation into about 40 units that compete with one another for funding.9

Valve, the gaming company (based in Bellevue, Washington) that’s behind such best sellers as Half-Life and Counter-Strike, has developed an interesting version of adhocracy, though it doesn’t use the term. Valve claims to have no managers. Employees are encouraged to initiate new projects and to choose which of them to work on. Self-selected teams emerge spontaneously where the most exciting opportunities appear to be rather than according to a strategic plan or a product-development road map.10 As employee Michael Abrash noted in his blog, this approach is appropriate because “most of the value [in gaming] is now in the initial creative act. . . .What matters is being first and bootstrapping your product into a positive feedback spiral . . . Hierarchical management doesn’t help with that. . . .”11 Nor, we might add, would the meritocratic model, for speed to market could be sacrificed to disputes and debates.

A key feature of this form of coordination is the disbanding of project teams once activities are complete. Opportunities are ephemeral by nature, and work in an adhocracy should reflect this. Mundipharma is a fast-growing midsize player in the pharmaceutical industry. Traditional pharmaceutical companies make long-term commitments to specific therapy areas. Mundipharma organizes its business units around specific drug opportunities. When a business unit successfully launches a new drug, that unit continues to operate, but it shuts down if the launch fails, and the employees move over to other, more promising areas. As a result, the company is quite market focused—its model resembles that of a venture capitalist: it invests only when it sees a clear pathway toward a commercially viable drug.

In sum, adhocracy’s ability to coordinate workers around tangible external opportunities keeps them closer to the action and less inclined to spend time deliberating. That, in turn, generates a higher return on attention than the traditional coordination processes of bureaucracies or meritocracies do.

2. Making decisions through experimentation

In a bureaucracy, decisions are made through the hierarchy: superiors tell their subordinates what to do, and so on down the line. In a meritocracy, decisions are made through argument and discussion, and everyone is entitled to weigh in with a point of view. The decision-making model in an adhocracy, in contrast, is experimental, which means consciously cutting short internal deliberations and trying things out with customers to gain rapid feedback. While this concept has been around for many years, in our experience most large companies still fall back on formal stage-gate processes and committees rather than risk releasing unproven ideas in the market.

In 2012, Costa Coffee, the world’s second-largest coffee chain, developed an ambitious plan, codenamed Project Marlow, to transform its vending-machine offering by creating an entirely new self-serve coffee system that would engage all the five senses of the customer. Project Marlow was agreed on with a handshake in January 2012. The formal kick-off meeting, with 20 people, was held on April 19th. The beta version was delivered, on time and on budget, on September 20th of the same year. “The pace of work was uncomfortably high,” recalled project leader Eric Achtmann, “the team was small and world-class without exception, and decisions were made on a 24-hour cycle.”

One key principle of Project Marlow was to base decisions on what would move the effort forward. For example, potential partners that required a legal agreement to be in place before starting work were eschewed in favor of those prepared to get going on the basis of a handshake. Project Marlow’s ground rules included rapid decision making (less than 24 hours); a relentless focus on results, not activities; and a preference for asking forgiveness, not permission.

In one instance, technical problems with a key subsystem threatened to delay the project as a whole, so Achtmann created a parallel team to find a way around them. Decisions were made by the person closest to the action, whose proximity gave the company a better-informed and more instinctive understanding of what had to be done. “On a project like this,” observed Achtmann, “people are making decisions on the fly, very aggressively. But they are informed decisions. The purpose is always just to get to the next stage as effectively as possible; and once there, the next target becomes visible.”

This approach to decision making, by nature, has a strong intuitive component. While that can entail greater risk and doesn’t work out every time, it helps companies to avoid analysis paralysis—an increasingly costly pitfall in a world with more and more information but fewer and fewer clear answers. Adhocracy is extremely well suited to help generate such intuitions.

In a meritocracy, employees gather information to persuade their peers. In a bureaucracy, they pass information upward, often in a bid to influence budget allocations. Decisions are passed down from more senior levels, often in the form of budget distributions. Adhocracy keeps decision makers more deeply immersed in the flow of a project or a business rather than more removed from it. For individuals, this place in the flow is a powerful position for inspiring useful intuitions, although such a flow can be generated by various means, including the collective activity that takes place in online communities or internal social-media platforms.12

This experimental approach to decision making has already found its way into some management processes. Agile techniques, for example, have been shown to be a better way of developing software in many settings than the traditional waterfall model.13 But many other management processes—from budgeting to capital allocation to new-product development to project staffing—continue to be managed through the traditional bureaucratic or meritocratic models.

As the rapid experimentation characteristic of adhocracy gains sway, a company’s return on attention improves. The ratio’s numerator rises with the quantity and quality of quick, experimental decisions, even as less time is spent (or even available for) disputing data or managing upward in a hierarchy.

3. Motivating people through achievement and recognition

Generally speaking, bureaucracies motivate people primarily through extrinsic rewards—above all, money. Meritocracies and adhocracies both motivate them through achievement and recognition. But meritocracies also emphasize giving people interesting work and enabling them to achieve personal mastery in a field of expertise. In adhocracies, motivation centers on giving people a challenge and providing the resources and freedom they need to surmount it.

Consider again Costa Coffee’s Project Marlow: Eric Achtmann deliberately built an elite team and gave its members an almost impossible deadline. Of course, this doesn’t mean that an adhocracy can’t offer financial rewards as well; the Marlow team members all had a stake in the upside growth of the project—the sale of lots of machines—which ensured an alignment of material interests. Achtmann also spent a good deal of time working on team spirit, coupling demanding standards and grueling milestones with celebratory events every time a milestone was achieved. He also created a plaque that would be permanently mounted inside every production machine, with the names of all 38 key team members who made “an extraordinary and enduring contribution to Marlow, above and beyond the call of duty.”

Valve offers a slightly different proposition to its employees. Challenge is the starting point: the handbook for new employees says, “Valve has an incredibly unique way of doing things that will make this the greatest professional experience of your life, but it can take some getting used to.” Employees have very high levels of responsibility (“You have the power to green-light projects. You have the power to ship products.”), and the company emphasizes that hiring great colleagues is “your most important role.” There is also a significant extrinsic component to motivation at Valve—one that would not be out of place at GE: employees rate their peers, and a forced-ranking system gears discretionary pay toward those who contribute the most. Valve is a competitive and challenging place to work, and its founders believe that this makes it attractive to the most talented game developers.

Again, this approach to motivation helps overcome analysis paralysis and increases a company’s return on attention. The heroes are the people who make something happen—for example, completing a pilot project quickly and ahead of budget—rather than those who come up with the cleverest ideas (which would be celebrated in a meritocracy) or who oversee the biggest budgets (the mark of respect in a bureaucracy).

Choosing the right model

So which is the right model for your company? It is worth noting, first of all, that bureaucracy, meritocracy, and adhocracy are all about the relative emphasis on formal authority, knowledge, and action. When we ask executives which model they prefer, they typically say they want the benefits of all three: people who bring their knowledge and formal authority to bear, take decisive action, and act intuitively when necessary.

But we would argue that this is a hedge. Companies can’t put equal emphasis on all three dimensions at once; they have to make it clear which one takes precedence, at least in specific business environments and units. For example, if your company aspires to be focused and action oriented but continues to operate on traditional bureaucratic principles, don’t be surprised when things move more slowly than you expect. To get a better sense of your current overall organizing model—and which parts of your organization might be best suited for bureaucracy, meritocracy, or adhocracy—think about the descriptions in the exhibit, then download and try out our simple diagnostic (PDF–186KB).


The demands of the business world often change more quickly than the organizations where we work. Many companies are still moving from the traditional bureaucratic model, which has been around for 100 years, toward a more meritocratic one built around the primacy of information and knowledge. But our analysis suggests that for many companies, this isn’t enough. Meritocracy has its benefits, but we believe adhocracy will become increasingly important in the decades ahead. By understanding the benefits of all three management models, you will have a better chance of creating a style of working that positions your organization for future success.

Explore a career with us