Clear on Direction, Local on Method

“We’ve set the direction. Decide the how locally.”

Hear this sentence in an organization and some people feel they have been given authority, while others feel more lost than before. The same words produce different reactions because of what was shared before they were spoken. If goals and criteria for judgment have been agreed, the sentence is a starting point for execution. If it is unclear even what needs to be achieved, the front line inherits the job of interpreting the direction first.

I want to think about this difference through the lens of strategic ambiguity. There is value in choosing not to specify everything in an organization. What matters is what you make clear and what you leave open. The space a strategy leaves empty should be space in which the front line can build an answer.

How much should a strategy answer?

How specific does a strategy need to be in order to be executed? If headquarters believes it must decide customer-facing wording, regional sales methods and the order of execution, the strategy document grows without end. If it believes a broad direction is enough, each unit may move toward a different goal.

Headquarters and the front line hold different information. Headquarters is better placed to see resource allocation across businesses and long-term priorities. The front line is closer to customer reactions, competitors’ moves and local business practices.

An early statement of this problem is the economist Friedrich Hayek’s 1945 paper “The Use of Knowledge in Society.” It is about market economies, but it also helps in thinking about organizations. Hayek argued that decentralization is needed if knowledge of the particular circumstances of time and place is to be used promptly. In the same paper he notes that the “man on the spot” cannot decide on the basis of knowledge of his immediate surroundings alone. In a market, prices carry word of what lies beyond them.

Read in an organizational setting, the question shifts a little. When discussing how specific a strategy should be, I think we should ask how much detail to write, and also where each decision should be made and how the front line will learn what it cannot see.

Take customer retention. Whether to prioritize it may be a company-level choice. But execution depends on whether customers leave because of late deliveries, difficulty using the product or the quality of support. A shared goal does not require a shared solution.

“Local” in this essay is not limited to overseas subsidiaries or regional organizations. It includes units such as sales, customer support and product teams, which encounter problems up close and carry out the work.

Where should ambiguity remain?

First, the term needs a clear scope. Strategic ambiguity comes from research on organizational communication. In his 1984 paper “Ambiguity as Strategy in Organizational Communication,” Eric Eisenberg argued that clarity is not a self-evident standard for communication in organizations. Members facing conflicting goals may use ambiguity on purpose, and that ambiguity can promote “unified diversity” among people who interpret things differently.

The same paper points out that ambiguity can also preserve privileged positions. This deserves attention. An ambiguous sentence is convenient for a leader who has put off a decision and for one who would rather avoid responsibility. From the words alone, you cannot tell whether “let the front line decide how” is delegation or a transfer of blame.

So this essay does not equate strategic ambiguity with local autonomy. The way of operating proposed here is closer to clear strategic intent with decentralized execution. I keep the term because the question it raises is still useful: how much of a strategy should be left undecided? I read it as room for judgment, deliberately designed. That is my own view, connecting related arguments, and not an academic definition.

Settle together

• The result to achieve
• Priorities when goals conflict
• Boundaries not to cross
• The scope of local decisions

Leave open to the front line

• Diagnosing the cause
• Choosing the method
• Sequence and pace
• Designing experiments

My own framing. When the left side is empty, the freedom on the right easily turns into confusion.

Put the open space in the wrong place and problems follow. If methods are delegated while success remains undefined, people have no standard for evaluating their own choices. If executives reveal different expectations only after the work is done, people may wait for approval next time instead of deciding for themselves.

Three gaps between plans and results

One of the most systematic treatments of this problem is Stephen Bungay’s The Art of Action. Starting from Moltke and Clausewitz in nineteenth-century Prussia, Bungay identifies three gaps between plans, actions and results.

GapCommon responseBungay’s proposal

KnowledgeBetween plans and resultsMore detailed information and plansDo not direct in more detail than the situation allows

AlignmentBetween plans and actionsMore instructionsCommunicate intent as clearly as possible

EffectsBetween actions and resultsTighter controlGive freedom of action within the bounds of intent

Adapted from David K. Hurst’s review “Closing Implementation Gaps” in strategy+business.

Bungay’s point is interesting because the common responses look like common sense. When results fall short of expectations, it feels natural to investigate more, instruct more and check more. Bungay argues that these responses actually damage people’s ability to act toward a shared goal. His alternative is “directed opportunism,” which carries Prussian mission command (Auftragstaktik) into the language of business. The emphasis moves from “plan and implement” to “do and adapt.”

The same review also notes the limits. It is less clear how this works in actual management, and taking root seems to require considerable training. Military doctrine cannot simply be carried over into a company, and that is worth keeping in mind while reading.

Intent and rules, a language for judgment

Communicating intent does not by itself enable the front line to judge. In the HBR article “How Leaders Can Let Go Without Losing Control,” Mark Bonchek observes that organizations have goals and rules but rarely have principles that guide judgment. He argues that intent and principles together deliver both alignment and autonomy, and he distinguishes values, goals and principles.

ValueFrugalityWhat matters

GoalSave moneyWhat you want to see

PrincipleSpend others’ money like your ownHow to decide in the moment

Applied to my question, the executive’s job is to create a language in which the front line can judge. “Care about customers” does little to resolve a conflict among price, service and cost. Suppose instead the rule is “put keeping customers using the product ahead of one-off sales.” Now people can debate what to choose first in a specific situation. This example is my own illustration and does not come from the article.

Kathleen Eisenhardt and Donald Sull’s “Strategy as Simple Rules” offers another clue. The authors propose a few simple rules for capturing opportunities in fast-changing environments, and they distinguish five kinds: how-to, boundary, priority, timing and exit.

Applied to execution, this means sharing the criteria for important choices without scripting every action in advance: which customer problems to focus on, how wide a range of experiments to allow, when to stop and reassess. Within those criteria, the front line can explore several paths.

Connecting the two arguments, I would put it this way. Purpose sets the direction of action, principles set the criteria for choice, and local judgment produces the specific solution.

How far can authority go?

Once direction and principles are shared, the next question is the scope of authority. Two ideas help here.

The first is the experience of David Marquet, former commander of the US Navy submarine USS Santa Fe. In Turn the Ship Around! he describes having his officers say “I intend to…” rather than wait for orders. By his account, changing the form of speech led officers to think from one level up. He holds that control can be passed down only together with competence and clarity, because authority without competence leads to chaos.

The second is the distinction between decisions in Jeff Bezos’s 2015 letter to Amazon shareholders. Decisions that are hard to reverse should be made carefully, while reversible ones should be made quickly by high-judgment individuals or small groups. He warns that as organizations grow, they tend to apply heavy processes to every decision, which leads to slowness and excessive risk aversion.

Add a second axis, the effect on other units, and you get a table that is useful for discussing the scope of authority.

Stays within the teamAffects other teams Easy to reverse

Front line decidesShare results and lessons

Front line decides, with noticeAlign timing with affected teams

Hard to reverse

Front line proposes, joint review“I intend to…”

Decide with executivesJudge against company priorities

My own example, adding scope of impact to Bezos’s distinction. The lines will differ by organization.

Five agreements behind “decide locally”

To put the discussion into practice, I propose confirming five things. These are not an official framework from HBR or the books above. They are my own application.

1. ResultsWhat change are we trying to make?Separate expected results such as revenue, retention and user experience, and set the period and criteria for judging performance.
2. PrioritiesWhat comes first when goals conflict?Agree where the weight falls between growth and profitability, or speed and quality.
3. AuthorityHow far can the front line decide on its own?Determine which of budget, price, staffing and customer commitments it can change.
4. BoundariesWhat must hold even during execution?Check specifics such as customer contracts, product quality and dependencies on other units.
5. ReviewWhen do we stop or revisit the direction?Allow for review when a strategy’s premises change as well as when performance is low.

These agreements are not completed by a single instruction. The front line has to apply executives’ criteria to real cases, and the two sides have to work through the places where interpretations differ. Using the same words is different from having the same criteria for judgment. There is a simple test: take one hard decision from last quarter, have executives and the front line each write down how they would decide, and compare.

One direction, different solutions

Here is a hypothetical case for illustration. Suppose a company makes improving customer retention its priority for next quarter. Headquarters shares a common metric and measurement standard, a budget for experiments and principles for notifying customers of changes.

RegionCause found locallySolution chosen locally

ALate deliveries discourage repeat purchasesChange logistics partners and delivery notices

BHard initial setup leads customers to stop using the productStrengthen first-use guidance and training

CInquiries go unresolvedAdjust agents’ authority and escalation paths

Shared result · Better customer retention

The three regions act differently, yet all aim at the same result. Headquarters has no need to conclude that consistency is lacking just because methods differ. It can look instead at which customer problem each method targets and how each hypothesis will be tested.

If each team measures results in the way most favorable to itself, however, comparison and learning become difficult. Other explanations, such as customer mix or seasonality, need examining along with the before-and-after change. A shared metric is where the conversation starts, and understanding what an action achieved also takes local context.

Autonomy depends on connections across the organization

In “Why Strategy Execution Unravels—and What to Do About It,” Donald Sull, Rebecca Homkes and Charles Sull test common beliefs about execution against survey data. What struck me first was the difference between vertical and horizontal relationships.

I see this as a key condition for local autonomy. One team’s decision can change another team’s work. If sales promises a new service level, operations needs the resources. If the product team changes a feature, customer support has to change its guidance. The freedom to decide locally has to include how to negotiate choices that affect other units.

In Team of Teams, Stanley McChrystal addresses a similar problem, drawing on his command of special operations forces in Iraq and Afghanistan. He shared information transparently across the organization to build “shared consciousness,” and on that basis field teams made decisions in real time. The leader’s role shifts toward “Eyes On, Hands Off.”

In corporate terms, Netflix’s culture memo is a well-known statement. It asks managers to practice “context not control” and describes an approach that is “highly aligned and loosely coupled.” The same memo adds that this should not be confused with hands-off management.

What gets shared has to go beyond results reports. What problem did you see, what did you assume, why did you choose this method and what did you actually learn? Passing these along helps other teams avoid copying only the surface of a success.

Executives must respond to what the front line learns as well. When evidence accumulates that the premises of the original direction were wrong, the strategy needs another look. Even with ample freedom of execution, an organization that ignores information from the front line can move faster in the wrong direction.

The leader’s job is to create the conditions for judgment

Leaders still have work to do after handing methods to the front line. They need to sort out where goals conflict, match resources and authority to the work, and coordinate problems that units cannot resolve among themselves. When the front line chooses a different method on the basis of evidence, leaders should be willing to examine that judgment and learn from it.

How accountability works matters too. Trying something reasonable within the agreed scope and falling short is one thing. Ignoring agreed boundaries or hiding problems is another. The two need to be looked at separately, so that people understand when to act on their own and when to ask for help.

“We’ve set the direction. Decide the how locally.”

For this sentence to become how an organization works, concrete commitments must follow it. We should be able to say what result we are after, which choices come first, how far people can decide and when we will reassess together.

A good strategy, as I see it, makes clear the direction we must travel together and leaves room for the front line to build the answer. When that room is filled with authority and resources, criteria for judgment and learning, “decide locally” gains the power to actually move an organization.

Further reading

Sources and scope