Bias toward action Move from discussion to action as quickly as possible

Article Sections

​In a rapidly changing world, workgroups don't have time to react to developments; members need to increase decision-making velocity without cumbersome approval processes, taking actions quickly and learning from each one. The key: choosing where to act to get the best impact on the outcome over time.

In this era of rapid change and constant disruption, how can organizations not only improve performance but accelerate it? In the article Beyond process, we introduced a new approach to accelerating performance—one that drives improvement through the adoption of nine high-level practices by frontline workgroups. These practices are more powerful together, and they are not sequential; not all groups will focus on all of the practices, and the practices will look different in each group. Together, they are critical for supporting the continuous learning and action that workgroups need to achieve greater and greater impact.

For an overview, see our introductory article, Moving from best to better and better.

Introduction: Reactivity is not enough

The front line is, of course, where most problems or opportunities first appear—and where people find themselves crafting strategies and taking actions to address them. Such moves usually need to happen at top speed, since the window of time to address the issue at hand is often short—too short to accommodate exhaustive analysis, planning, and approval processes. It’s no surprise that many organizations look to speed reactions and solve problems more quickly. But workgroups aiming to accelerate performance improvement should adopt a different mind-set: They should act rather than react.


​Interactive: Explore the nine practices 

Visit the full collection

Listen to the podcast

Subscribe to receive updates from the Center for the Edge

Create a custom PDF or download the full collection

In a rapidly changing world, workgroups take a real risk in reacting to whatever is happening at the moment. Reactivity tends to breed modest, incremental improvement at best; at worst, it tends to lock workgroups into their approaches of the past. Groups need to respond quickly to whatever they are confronting—and respond in ways that can move them toward achieving higher impact.

Action is a means of targeted and rapid learning that is an important element of accelerating performance. It is a different type of learning than training or sharing existing knowledge. Taking action to engage with a possible solution uncovers a problem’s conditions and requirements as well as the capabilities and limitations of our resources. This information informs the next action and ultimately creates new knowledge that can be built into a better approach. Until the new knowledge is embodied in action, the workgroup is unlikely to learn from it.

A group can learn faster how to achieve higher levels of performance by taking more of the types of actions that create new knowledge and matter to the outcome. Balancing the value of fast feedback with the longer-range goal to significantly improve an outcome, a group can avoid the reactive incremental loop and pursue truly impactful learning. This can shape how members will think about what actions to take and which actions and opportunities to pass by.

Further, in a world where what is true today about a given issue may not be true tomorrow, a bias toward action could orient the workgroup to look beyond compliance and the status quo. It can help propel a workgroup past the paralysis brought on by uncertainty and prompt the group to keep testing assumptions and developing new approaches to improve performance regardless of inertia or roadblocks in the larger organization. A bias toward action can also help clarify the overwhelming noise that many workgroups sometimes encounter.

To accelerate performance improvement, workgroups should increase decision-making velocity, taking reasonable and fluid actions—whether that is first responders breaking down a door or a product designer posting a mock-up on a platform—without cumbersome decision-making and approval processes. Groups should be able to take action—small moves, smartly made—over and over and over, to keep testing conditions and assumptions and pushing boundaries to reach higher levels of performance.

Too much planning or approval-seeking without action can defuse momentum, squelch passion, and delay the learning and refinement needed to progress. If workgroups are too slow to try things outside the status quo, they may miss valuable opportunities. For example, Polaroid was slow to act when radical changes were occurring in the imaging/camera industry even though the company had the technical capabilities to pursue a new approach.1 Accelerating performance improvement, then, doesn’t necessarily come from response and reaction, no matter how fast, but from choosing where to act to get the best impact on the outcome over time.

You know you need this practice when:

  • We are slow at putting our ideas into action; we focus on process (for example, stage gates) more than the results we are trying to achieve
  • There is no time to tinker, prototyping new ideas feels high risk, and failure is frowned upon
  • Many people can say no, while no one can clearly say go
  • We can’t get the right people in the room to make decisions—or, worse, we have to get everyone in the room
  • The outcomes around which we align cater to the lowest common denominator; we could be missing the opportunities that would have significant impact
  • We feel as though we have too few resources to achieve the impact we want
  • There are no consequences for not improving performance over time

The bias toward action practice: What it is

In a fast-paced and unpredictable business environment, not all actions are equal. Action matters when it leads to new actions that can ultimately deliver higher impact. Bias toward action is about acting quickly to learn faster, but it’s also about choosing where to act; deciding what will drive the most useful learning. It is a balancing act: Groups need to get into action sooner—and also take every possible moment before action to get the most out of it. A strong sense of where the workgroup is aiming, what performance metrics matter most, and what the workgroup doesn’t yet know make bias toward action possible.

Effective action to accelerate performance improvement is typically characterized by:

  • Timing. Be explicit about the downsides of waiting to act. Knowing when to act can be as important as knowing what to do.
  • Leverage. Leverage others’ capabilities to learn as fast as possible and focus on what has not been done before.
  • The unknown and unpredictable. Aim for actions that haven’t been taken before, whose effect is unknown, rather than variants designed to confirm a hypothesis. There shouldn’t be a designated result that, if it doesn’t turn out that way, the action is a “failure.”
  • Improvisation. Improvise as you go. Look for ways to tinker with the approach, and incorporate feedback to build on—and build in.
  • Short feedback loops. Take action that elicits useful feedback faster. Look for ways to get feedback earlier from actions that take longer.
  • Planning. Take time to understand and manage risk in advance. Plan for how the workgroup will gather feedback and learn from actions, and consider whether the action can be made more productive.

. . . and what it isn’t

Agile. The way many organizations have interpreted and implemented Agile, it is almost exclusively focused on speed-to-market and the ability to respond more quickly. While organizations can find this incredibly useful, workgroups aiming to accelerate performance improvement should focus on the actions that will help them learn faster how to reach higher and higher levels of performance.

Acting for the sake of action. Without a clear direction and a desired impact to help guide and prioritize possible actions, groups can spread themselves too thin in a misguided belief that more action is inherently better than less. Of all the actions on the table, choose those most likely to have an impact on the performance that matters most.

Acting recklessly. It’s anything but. All failures are not created equal, and those resulting from inattention or lack of effort or competence should have consequences.2

Intersections with the other eight practices

Without action, the workgroup’s efforts create little value.

  • Prioritize performance trajectory. Out of all the possible actions, which is most likely to have an impact on the performance metrics that matter most to the outcome?
  • Maximize the potential for friction. A diverse and passionate workgroup can shape and build the actions that will have the most impact on performance.
  • Seek new contexts. Workgroups can draw inspiration and insight from other contexts about what assumptions they should test and what types of actions they can take next.
  • Cultivate friction. If the point of action is to generate the most impactful learning, groups should constantly question assumptions and look for opportunities to build on and improve other actions.
  • Eliminate unproductive friction. Workgroups can act more effectively and learn from actions if members do not fear judgment and repercussions and expect actions to be made in support of the workgroup’s learning.
  • Reflect more to learn faster. Action generates the raw material for reflection. What action can give us the quickest feedback about how to improve the outcome the most?
  • Frame a powerful question. The powerful question can help overcome old assumptions and build more of a creative set of conditions for action.
  • Commit to a shared outcome. Action has implications for bridging all of the workgroup’s practices into impact on the shared outcome.

Putting the practice into play

Workgroups aiming to accelerate their impact should act relatively quickly. But they should also act deliberately, to avoid getting trapped into reacting to the moment rather than choosing the actions that have the most potential to propel the workgroup toward its long-term objectives. The practice of biasing toward action, then, is a balancing act between speed and impact. It requires prudence and planning, as well as a nuanced understanding of risk to make it more manageable and place it in the context of other risks and rewards. Rethinking and reframing risk can make taking action more compelling. At the same time, planning actions to be less burdensome, more productive for learning, and designed to accommodate improvisation in the moment can further encourage workgroups to act. In fact, part of the practice is knowing when not to act—and being focused on exploiting the limited time available to make the next action, and the one after that, have as much impact as possible.3

Reframe risk

In a fast-moving environment, inaction is one of the greatest risks that workgroups face. Conceptually, we know that inaction means sticking with the status quo, which means, at best, diminishing returns and a shallow line of incremental improvement. At the outset, though, it may be hard to appreciate the opportunity missed or gauge the cost of a chance to learn passed up. To begin reframing the notion of risk, groups should make the risks of inaction part of the conversation. What are we risking by doing nothing? What is the potential impact of what we might learn? What is the cost of continuing without this learning? The risk of doing nothing is missing the opportunity to jump from a shallow linear curve to an accelerating trajectory of performance improvement: Where could we be in six months, in a year, in 10 years relative to today, if we get on an accelerating trajectory?

The same phenomena—increasing rate of change and shifting expectations and demands—that are moving the action to frontline workgroups also significantly increase the risk of inaction, though few organizations have the tools or skills to really understand the impact of opportunities missed. A group can set a tone by deliberately focusing conversations on action and making the risk of inaction part of any conversation about risk. It can also be useful to draw on well-known examples of the changing dynamics in other domains to be explicit about the potential downsides of waiting (for approval, for clarity, for external pressure) relative to the potential of getting on a higher trajectory in such an environment. Consider, for example, the story of Amazon Web Services. Back in 2005, when a group at Amazon began working on the project, many likely questioned the investment—after all, what did it have to do with books? Yet within a decade, it had reached $10 billion in annual sales and was growing at a faster pace than Amazon’s e-commerce business.4

For workgroups, a large part of developing a bias toward action is to focus on what can be gained from taking an action and then maximize the upside potential. What impact or learning might an action have on the outcome? How could we tweak this action to increase the impact or gain even greater learning? Consider what is desirable, feasible, and viable—in that order. One way to maximize the upside is to focus on actions that haven’t been taken before, whose effect is unknown. These will likely have far greater learning potential than trying out variants designed to confirm a hypothesis. The action should generate information or create new knowledge rather than have a designated answer that is either right or the action “fails.”

Putting thought and planning toward action may help to direct the group’s efforts toward high-impact goals, but members will likely have different perspectives about which actions have the greatest potential and how exactly they should be taken. The goal should be to balance impact with the speed of getting feedback to drive learning. If a workgroup has divergent views, try to disagree and commit rather than force consensus. This practice is inspired by a phrase from Amazon CEO Jeff Bezos, who credits this concept with the productivity of the company’s teams: Feel free to challenge an idea or plan, but when the time comes to make a decision, everyone commits to executing it even if they disagree.5 For workgroups, it generally means knowing when the value from further discussion is less than the value of getting feedback from action—when it is time to make a decision and move on. The expectation that many actions will fail to generate anticipated results means that the commitment to any particular action may be limited. It would generate learning of one kind or another, and if it fails to have the expected impact, the group may well go back and execute the other option.

Developing this sense of open-endedness about a workgroup’s decision-making can help that group take action more easily. Members can question their own assumptions about a proposed action’s magnitude and finality. Decisions often feel weighty because we assume they are weighty; it’s always worth questioning. Consider making it a formal part of discussion to ask: How significant is this decision? What are the implications for regrouping and trying something else if this action doesn’t have the expected impact? This is somewhat of a paradox, since workgroups should be looking for actions to take that are significant in terms of potential for impact and learning, while also thinking about them as transitional and experimental. In rapidly changing conditions, this impermanence only increases; as one Southwest field tech put it: “What was ‘no’ yesterday might be ‘yes’ today.”6

Consider playing with assumptions and boundaries to make more decisions reversible. Reversible decisions can be made with less authority or consensus, creating a stopgap for workgroups that might get stuck in analysis paralysis. If an approach or decision fails, the group can quickly recover and try another option rather than live with the consequences for too long. In doing so, members would learn and move on to focus on learning about the biggest opportunities rather than trying to predict the future.

Making risks more manageable can also tip the scales toward action rather than deliberation. Simulate actions by creating sandcastles in environments that have a lower cost of tinkering and contain the ripple effect of experiments. Try using environments that aren’t dependent on core processes and IT, and leverage virtual tools to iterate quickly on specific actions that would benefit from tinkering. This practice could make it easier for workgroups to embrace productive friction because it would lower the stakes of any one challenge or decision—hey, it’s only sand. The group would develop an approach and then immediately build or test it. The immediacy of the action can generate rapid feedback for the group to take in and reflect on, shortcutting the need for long decision-making processes and cumbersome scheduling and buy-in. Having limited downstream effects, a group would have increased degrees of freedom to quickly test out and adjust its approach rather than trying to fix an airplane midflight.7

For example, FirstBuild—the open innovation unit of GE Appliances, now a Haier company8—uses a community of enthusiasts to test concepts for new products. After seeing which types of products generate enthusiasm, such as a “chewable ice maker,” FirstBuild might go back to the community with more detailed concepts. After narrowing the concept, the workgroup begins a more detailed design, going back to the community as needed. Once a design is ready to prototype, the group uses a crowdfunding site to test the market’s interest in the product as designed and priced. If the market is less interested than expected, the group can easily pull the product back and either kill it or tinker with features and pricing to take to market again. In Royal Caribbean Cruises’ Newbuilding & Innovation workgroup, members build sandcastles in virtual environments where they can simulate thousands of design solutions in a few hours, swapping out details that would otherwise have been costly or impossible to test in reality. For example, the group adjusted colors of panels, tested how much light structures provided at night versus day, and got to identify safety hazards invisible in blueprints. While members don’t get the benefit of guest feedback, the entire group can see the design impact of decisions almost immediately, making for richer reflection and discussion of what actions come next.9

Workgroups should be able to address the issues or opportunities they see unfolding in front of them when the chain of command is occupied with other concerns, or when there isn’t time to wait for more complete feedback or further instruction. An assumption of permission—go until “no,” both for the group actions and for individual members—is key for moving quickly and not getting hung up seeking permission, consensus, or buy-in from a wide array of possible stakeholders in advance. Asking for forgiveness rather than permission can help a group maintain momentum and spend its resources on activities that generate new knowledge rather than on navigating the organizational structure.This assumption of permission may conflict with an organization’s broader culture and make members uncomfortable. In order for this to work, members would have to trust each other to act in good faith in the interest of improving the shared outcome and to have a clear understanding of the need to prioritize actions that have the potential for greatest impact on the outcome. In increasingly dynamic environments, acting too slowly may be riskier than letting competent people exercise their judgment. Constraining decision-making authority could also constrain a group’s learning potential and may be unnecessary if the group’s objectives and priorities are clearly understood to guide decision-making. As Gen. Stanley McChrystal describes in Team of Teams (his book about the Joint Special Operations Command during the Iraq War), “I was connected to almost every decision of consequence. This was great for establishing holistic awareness but it also created a nightmare of paperwork and approvals. . . . The wait for my approval was not resulting in any better decisions, and our priority should be reaching the best possible decision that could be made in a timeframe that allowed it to be relevant. I communicated across the command my thought process on decisions like airstrikes and told them to make the call.”10

Act to learn

The most powerful learning for workgroups can be through action—getting out there and doing something—rather than sitting around a table and discussing. The more quickly the group gets to action, the sooner it can start learning how to accelerate performance improvement.

The relevant actions for a group often aren’t the type that require large investments and extensive planning. If the actions are, instead, a means of learning to improve the outcome, how can we formulate actions to maximize the potential impact on the outcome and also have shorter feedback loops? One way could be to divide complex actions with long feedback loops into a series of assumptions to test. Consider formulating the most impactful actions into a series of small moves with interim milestones designed to elicit new information and create new knowledge. Try to stage your moves to focus on getting the actionable information or feedback that is important to the next step as quickly as possible without losing sight of the larger action. Consider what information or knowledge the workgroup may be missing and what feedback would be sufficient to inform further action. These actions can be viewed as interwoven experiments in a larger experiment that can lead to better solutions and outcomes in the future.11

A minimum viable approach (MVA) can help minimize effort, maximize momentum by helping groups quickly identify what works and what should be discarded. MVA is frequently used in product development to deploy a product in the market sooner; in the context of workgroups, it has a wider aperture. The group would focus on identifying the barest approach or action that can lead to the next iteration, accelerating the rate of learning and encouraging members to test ideas outside their comfort zone or established approaches with minimal investment of time or resources.12 MVA may not be appropriate for all situations (for instance, space exploration or surgery), and scaling a solution may eventually require greater organizational support.

Workgroups can lower the barriers to action, increase the diversity of perspectives, and reduce risk by leveraging (capabilities, expertise, resources) to learn from outside the group. In fact, setting constraints—time, budget, technical—can prompt more creativity and also focus a workgroup where it is most likely to create value. If someone does it better, let her do it for you. Many work products are openly available and can reduce the cost, time, and effort required to act. Doing this well may require emphasizing rapid appropriation and reusing knowledge from other contexts and tight feedback loops so that participants can rapidly build on the contributions of others. While orchestrating others and using existing third-party tools has costs, mobilizing others can cultivate allies, build relationships, and allow a workgroup to focus on what it does best. The more rapidly a group learns from others, the richer the overarching set of possibilities—both the nature of the opportunity and the journey needed to achieve it.

Keeping in mind that the reason for adopting “minimum viable” practices is to accelerate the rate of learning, workgroups can accelerate decision-making as a proxy for whether they are making progress toward creating an environment where more learning happens faster. A “good” decision made too late for the opportunity or challenge can prove worse than an imperfect decision made in the moment. Part of this practice is to get more comfortable with acting on less information. Another is to get more creative at identifying proxies for the information you need. A third aspect is to make use of the immediacy and transparency of technological tools to get input much more rapidly than hiding behind established decision-making processes. For example, e-commerce luggage start-up Away attributes the use of Slack to “making decisions in a day that used to take weeks or months.”13

Consider how the Joint Special Operations Task Force in Iraq went from 10–12 highly planned monthly raids to more than 300 monthly raids by learning how to take action within, sometimes, minutes of receiving actionable intelligence information. The raids may have had more unknowns, but they were also more successful at capturing targets and additional intelligence information because they were acting on information that hadn’t gone stale.

Another example comes from Southwest Airlines, where the Baker workgroup has developed a tool to see decisions’ direct impact on network operations. At one point, the group wanted to add a graphical dashboard that would show where Southwest was long or short on airplanes. Rather than take weeks to build the functionality, the superintendents of dispatch sat next to a developer in the workgroup, working together to create a summary table built into the system in only a few hours. They began getting feedback immediately as their colleagues began using it.14

If you know in advance that it’s going to work, it’s not an experiment.

—Jeff Bezos15

Jazz it up

There is no such thing as perfect experimentation or efficient innovation. Taking action early and often can only produce so much learning if workgroups don’t also bring a spirit of play and possibility to their work. Try to take off the guardrails and embrace the messiness of rework and deviation. The point is not to discourage mistakes but to encourage recognizing mistakes, ineffective approaches, and invalid assumptions, and use them as inputs into better approaches, sooner. To really draw on its performance improvement potential, a diverse workgroup should embrace the vital role of improvisation, failure, and the unexpected in creating new knowledge that can lead to better and better outcomes over time. Similarly, leveraging capabilities from the outside is nothing more than outsourcing if the group uses those capabilities in predetermined, already-established ways.

With any action, look for what’s not being done. We tend to focus on the urgent or the easy—because it’s right there—but what is most important to the outcome may be neither urgent nor close at hand. Venture into a territory where your efforts can expose or create new knowledge rather than iterating on well-worn ground where the insights are incremental. The more unexpected the outcome, the more potential for valuable learning. If a workgroup is generating few surprising outcomes, it may not be pushing the boundaries that would lead to a new level of performance.

Improvisation is a skill that defies documentation, codification, and outside control. It can be misconstrued as chaotic, with individuals just winging it. In fact, for workgroups, similar to jazz ensembles, the quality of improvisation could depend in part on the foundational skills and talents each member brings, and in part on the quality of listening and riffing on what others are doing—and what has already been done—to make each additional move additive and constructive. Expand the potential for improvisation by relaxing organizational and operational constraints that get in the way. Royal Caribbean, for example, creates the space for improvisation by building change orders into the plan so that the company is prepared, structurally and mentally, to benefit from the interactions of the diversity of backgrounds brought together in the design workgroups. This also seems to set the expectation that members could build off of each other.

Antibodies at work

  • We need consensus. If we get everyone on board, we’re full steam ahead.
  • This is it—take your shot.
  • Failure is not an option: Screw this up, and you’d better look for another job.
  • We don’t move until we’re sure.

Celebrate the “fast failures” as opportunities to practice improvising in the moment. This would keep the focus on problem-solving, incorporating new information, and creating new knowledge. Although failing fast has become almost a cliché when talking about innovation, the key is often to keep improving the group’s ability to find the easiest, fastest ways to generate discrete and actionable feedback. Workgroups’ limited size and the shorter time frames within which they work typically demand that these practices more closely resemble tinkering than iteration, and small, rapid adjustments rather than formal revisions.

Failing can lead to unexpected outcomes. Build on mistakes. Rather than start over after a failure or, worse, hiding it, consider incorporating failures and the learning from them into the next action. Starting with a clean slate loses the learning. Workgroup members may struggle with recognizing the value of what they are learning from unexpected outcomes, and what is relevant may come to light only through discussion and additional viewpoints. Consider the well-known example of 3M and the Post-it note. One of the company’s most widely sold products, the Post-it resulted from a “defective” new adhesive that was insufficiently sticky to hold papers together; sheets could just be peeled right off. It was only after consideration that the workgroup recognized the potential for an alternate use.16

Questions for reflection

  • What has prevented us from accomplishing our goal?
  • What do we do best that will make a difference, and what can we rely on others to do?
  • To what extent is the current process holding us back from achieving more?
  • To what extent could we make decisions faster with less information and achieve more as a result?
  • Are the types of decisions we make final and unchangeable? Do they feel that way?