How Interdependent Teams Make Better Decisions Without Centralizing Control: What FAA Collaborative Decision Making and Peak OS Reveal

One of the hardest problems in a complex organization appears when everyone is making reasonable decisions from their own point of view.

Engineering protects technical integrity.

Sales protects the customer relationship.

Finance protects capital.

Product protects the roadmap.

Manufacturing protects production flow.

Program Management protects delivery commitments.

Each team may be acting rationally.

And the collection of rational local decisions can still produce a poor outcome for the company.

That creates an important leadership question:

How do we preserve functional autonomy without allowing teams to optimize their part of the system at the expense of the whole?

There is an unusually useful example in the way the Federal Aviation Administration approaches air-traffic-flow management.

The FAA's Collaborative Decision Making, or CDM, program is a joint government-and-industry initiative designed to improve air-traffic-flow management by increasing information exchange among stakeholders. Participants include the FAA, airlines, general aviation, private industry, and academia. The program traces its origins to a 1993 FAA/airline data-exchange experiment and was formally established in 1995.

The FAA describes CDM as an operating paradigm in which decisions are based on a shared, common view of the National Airspace System and awareness of the consequences those decisions may have for the larger system and its participants. Two central ideas sit underneath the program: better information can produce better decisions, and organizations need tools and procedures that allow them to respond as conditions change.

That second part is especially important.

Airlines do not stop being airlines.

The FAA does not become an airline.

Airports maintain their own responsibilities.

Different participants still have different goals, resources, constraints, and preferences.

The objective is not to centralize every decision into one organization.

It is to give the participants enough shared understanding of the system that local decisions can account for their effects on everyone else.

That creates a powerful parallel with Peak OS.

Peak was not developed from FAA Collaborative Decision Making. It emerged independently through more than two decades of work with hundreds of founders, CEOs, investors, and leadership teams.

But both environments reveal a similar organizational truth:

Complex systems perform better when specialized participants can retain local ownership while seeing enough of the whole to understand the consequences of their decisions.

For aerospace, defense, robotics, advanced manufacturing, autonomy, physical AI, and other frontier-tech companies, that may be one of the defining challenges of organizational execution.

Local Optimization Can Create System-Wide Problems

Imagine an airline deciding only from its own local perspective.

It wants its aircraft to depart when it prefers.

Another airline wants the same thing.

An airport has capacity constraints.

Weather changes available routes.

Airspace congestion affects what controllers can safely accommodate.

Each participant has valid interests.

But the National Airspace System is shared.

One participant's decision affects others.

That is why the FAA describes air-traffic-flow management as balancing demand against available system capacity, and why CDM centers decisions on shared information about conditions across the NAS rather than allowing each participant to operate from an isolated picture.

Companies experience a less visible version of the same problem.

Sales closes an important deal.

Locally, that is success.

But the commitment requires functionality Product had not prioritized.

Product adds it.

Engineering now has to move another milestone.

That delay affects Manufacturing.

Manufacturing changes its schedule.

The customer delivery changes.

Finance moves revenue timing.

What started as a good local Sales outcome changed five other parts of the system.

Nobody necessarily made a bad decision.

The problem was that the system consequence was not visible at the moment of the local decision.

This is why Peak's Symbiosis behavior matters.

In Peak Teams, Symbiosis describes an organization in which people and teams understand their contribution to the larger company, trust one another to fulfill their commitments, and work collectively toward shared Mission and organizational outcomes rather than operating as disconnected functions.

The standard for success becomes more demanding:

Did my function win?

is not enough.

The organization also has to ask:

Did the system win?

A Shared View Does Not Require Shared Priorities

This is one of the most sophisticated aspects of FAA CDM.

Different aviation stakeholders do not suddenly have identical priorities.

An airline has its own operational requirements.

Another airline has different ones.

The FAA is managing broader airspace demand and capacity.

Airports have local operating conditions.

Weather affects participants differently.

CDM does not eliminate those differences.

Instead, the FAA says stakeholders share information, values, and preferences to build a common pool of knowledge from which better system-level decisions can emerge.

This distinction matters in companies.

Alignment does not require Engineering and Sales to want exactly the same thing.

They should not.

Engineering should care deeply about technical reality.

Sales should care deeply about customers and revenue.

Finance should bring a capital perspective.

Manufacturing should bring production reality.

Those different perspectives are valuable.

The problem begins when each perspective operates from a different version of organizational reality.

Peak's planning architecture creates shared context through:

Mission.

Three-Year Vision.

One-Year Plan.

Company and team OKRs.

KPIs.

Roles and Responsibilities.

Those mechanisms do not eliminate functional expertise or priorities.

They create a common organizational reference point around which different teams can make decisions.

A company can therefore have different perspectives without having different realities.

That is an important distinction.

Better Information Helps Only If Everyone Is Looking at the Same Problem

One of the tools supporting FAA CDM illustrates the principle particularly clearly.

The FAA's Flight Schedule Monitor is designed to create common situational awareness among users and service providers in the National Airspace System. It presents airport and airspace demand and capacity information so participants can understand constraints and make collaborative traffic decisions. The FAA describes its significance simply: the system helps ensure that participants are looking at the same picture.

Companies have plenty of data.

The problem is often that different teams are looking at different pictures.

Sales sees CRM data.

Engineering sees its development environment.

Product sees the roadmap.

Finance sees the financial plan.

Manufacturing sees production.

People sees hiring.

Each picture is accurate within its function.

But leadership and cross-functional teams need another level of understanding.

What does the company plan require?

What outcomes matter now?

Which important objectives are On-Course?

Which are Off-Course?

What do the critical KPIs show?

Where has a dependency changed?

What decision does another team need to know about?

Peak creates that shared execution layer through the One-Year Plan, OKRs, KPIs, Weekly Camp, and Triage.

The goal is not to move every functional data point into Peak.

It is to make the organizationally consequential information visible enough that different teams can reason from a common picture.

Shared Data Is Not the Same as Organizational Intelligence

This difference becomes increasingly important as companies adopt more software and AI.

It is now easier than ever to share information.

That does not mean organizations automatically make better decisions.

The FAA's own modernization work illustrates the distinction. Its current Flow Management Data and Services program is intended to combine real-time flight, weather, schedule, and position information, improve predictive understanding, consolidate multiple displays, and streamline live-data exchange between the FAA and airspace users so participants can make more effective collaborative decisions.

Information has value because it improves understanding and action.

The same should be true inside a company.

A dashboard says:

Engineering is four weeks behind.

Organizational Intelligence asks:

Which company objective does that affect?

Which other teams depend on the milestone?

Is there still enough time to recover?

Does another commitment need to change?

Who owns the decision?

Does this remain a local Engineering issue or has it become an organizational issue?

The number did not change.

The organization's understanding of its consequence did.

That is what turns visibility into intelligence.

How Do You Coordinate Teams Using Different Systems?

This is a recurring buyer problem in scaling companies.

Engineering uses one system.

Sales uses another.

Finance another.

Manufacturing another.

Product another.

Leaders sometimes conclude that organizational visibility requires moving everyone onto a single platform.

Usually, that is neither realistic nor desirable.

FAA Collaborative Decision Making offers a more useful conceptual model.

Different participants maintain their own operational roles and systems while exchanging information needed to create shared situational awareness and support better system-level decisions. The FAA's modernization efforts similarly focus on live data exchange and integrated situational displays without eliminating the distinct roles of airspace users and traffic managers.

Peak works at a similar organizational layer.

The CRM can remain the CRM.

Engineering tools remain Engineering tools.

Financial systems remain financial systems.

The manufacturing environment remains specialized.

Peak creates an operating layer connecting the outcomes those systems are supporting.

What does the organization need to accomplish?

What is happening?

What matters?

Who owns it?

What has changed?

Where does another team need visibility?

What requires collective judgment?

This is why an operating system is different from a system of record.

It organizes organizational execution, not every unit of functional work.

The Consequences of a Decision Matter Beyond the Decision-Maker

One of the most interesting phrases in the FAA's description of CDM is its emphasis on awareness of the consequences decisions may have on the system and its stakeholders.

That idea deserves more attention in companies.

We tend to think about decision rights primarily as:

Who gets to decide?

That matters.

But complex organizations need a second question:

Who will be affected by the decision?

The Head of Product may own a Product decision.

But if that decision changes Engineering capacity, customer commitments, revenue timing, and hiring requirements, those consequences matter.

The CFO may own a capital-allocation decision.

But if that decision changes a program milestone, the operational consequence matters.

Engineering may own a technical architecture decision.

But if it changes Manufacturing or customer delivery, the system consequence matters.

Clear Roles and Responsibilities remain essential.

Peak explicitly uses them to establish ownership and increase Empowerment.

But good organizational decision-making requires ownership plus consequence awareness.

The person making the decision needs enough shared context to understand when a local decision is no longer purely local.

This Is Why Dependencies Need Visibility

A dependency is essentially evidence that one team's decision can affect another team's ability to execute.

Companies often discover those dependencies too late.

Product changes a requirement.

Engineering absorbs it.

The date changes.

Marketing does not know.

Sales communicates the old date.

Customer Success begins preparing customers.

The local decision moved faster than the information about its consequences.

Peak's Team-of-Teams planning is intended to surface more of these relationships before execution.

Input comes upward from teams.

Leadership creates organizational direction.

The plan moves back down.

Teams examine what it means.

Dependencies move across.

Material issues return upward.

Then the organization executes from a stronger shared picture.

During execution, Weekly Camp and Triage keep the information moving when circumstances change.

The objective is not to make every decision cross-functional.

It is to recognize which decisions have cross-functional consequences.

That is a much more scalable principle.

Collaborative Decision-Making Is Not Consensus

The name can create a misunderstanding.

Collaborative Decision Making does not mean everyone gets equal authority over every FAA traffic-management decision.

FAA CDM is a government-industry information-sharing and decision-support framework. Stakeholders contribute information and preferences, while the FAA retains its responsibilities within the National Airspace System. The collaboration improves the information and awareness underlying decisions; it does not erase organizational roles.

Companies need the same distinction.

Collaboration is not consensus.

Cross-functional planning does not mean every executive has veto power over another function.

Shared visibility does not mean everyone becomes a decision-maker.

A company needs:

broad enough information flow

and

clear enough decision authority.

Peak addresses that combination through Roles and Responsibilities, objective ownership, team participation, and Triage.

The team may discuss the issue.

Different experts contribute.

Alternatives become visible.

Then someone has to decide.

That is coordinated decision-making, not decision-by-committee.

Local Autonomy Becomes More Valuable When Shared Context Is Better

There is a counterintuitive relationship here.

Leaders sometimes assume that more organizational coordination requires more centralization.

It can be the opposite.

When teams have poor shared context, leaders have to pull more decisions upward because they cannot trust that local decisions will account for company consequences.

When shared context improves, teams can make more decisions locally.

The Sales leader understands the One-Year Plan.

The Engineering leader understands customer consequences.

Finance understands current organizational priorities.

Product understands capital constraints.

Teams can see important OKRs and KPIs.

Dependencies are discussed.

Ownership is clear.

Now people can exercise greater judgment because their decisions are informed by more than their local function.

This is one of the strongest connections between Organizational Intelligence and Empowerment.

Better shared understanding can enable greater distributed autonomy.

How Do You Stop Every Cross-Functional Decision From Escalating to the CEO?

This buyer problem becomes easier to diagnose through the CDM lens.

If the CEO is constantly resolving cross-functional issues, ask:

Do the teams share the same picture of the plan?

Are relevant constraints visible?

Do people understand the consequences of their decisions on other functions?

Are Roles and Responsibilities clear?

Do company-level outcomes have owners?

Does the organization have a known decision forum for issues that genuinely require cross-functional judgment?

Or is the CEO the only person who regularly sees enough of the whole system to connect the pieces?

That last condition is common.

The founder becomes the company's human integration layer.

Peak is designed to distribute more of that context.

The Mission, Vision, One-Year Plan, OKRs, KPIs, Roles and Responsibilities, Weekly Camp, and Triage give the leadership team and other teams enough shared organizational information that every cross-functional decision does not have to be reconstructed inside the founder's head.

The goal is not to remove the CEO.

It is to remove unnecessary CEO dependency.

Information Sharing Changes the Quality of Negotiation

FAA CDM also highlights another important aspect of multi-stakeholder systems.

Participants have preferences.

An airline may prefer one response to congestion.

Another participant may prefer something else.

A shared view of the system makes the tradeoffs more visible.

The FAA says that through information exchange and the sharing of values and preferences, stakeholders learn from one another and build a common pool of knowledge that can improve overall air-traffic-management decisions.

Executive teams negotiate continuously.

Sales wants investment.

Engineering wants capacity.

Finance wants discipline.

Product wants focus.

People wants hiring lead time.

When those conversations occur without a shared picture, they become positional:

My function needs this.

With shared organizational context, the conversation can become:

Given the One-Year Plan, current capacity, important OKRs, KPIs, and dependencies, what creates the best company outcome?

The function still has a point of view.

But the discussion is anchored in the system.

That is a much healthier form of executive conflict.

Local Metrics Can Accidentally Drive the Wrong Organizational Behavior

Another form of local optimization appears through metrics.

Sales is measured on bookings.

Manufacturing on efficiency.

Engineering on throughput.

Customer Success on retention.

Finance on runway.

Each metric may make sense.

But the company can still experience dysfunctional system behavior if teams optimize the measure without understanding the larger objective.

Sales closes revenue the organization cannot deliver profitably.

Manufacturing maximizes efficiency by resisting product variation the market needs.

Finance preserves runway by delaying a capability required to hit the plan.

Engineering optimizes throughput by avoiding important but difficult cross-functional work.

This is one reason Peak separates KPIs from OKRs.

KPIs provide important measures of performance.

OKRs can identify the capabilities or cross-functional outcomes the organization needs to build or accomplish.

The One-Year Plan provides the larger context.

Now leadership can ask:

Is this metric helping the organization achieve the plan, or are we optimizing the metric at the expense of the plan?

Metrics are valuable.

Mission matters more.

The Whole System Can Be the Right Unit of Optimization

FAA CDM explicitly seeks decisions and actions that create value for the National Airspace System rather than allowing every participant to operate solely from its local interests.

Companies need a similar leadership mindset.

A functional executive sits on two teams.

They lead their function.

And they are part of the leadership team responsible for the enterprise.

Those identities can come into conflict.

The Head of Engineering may know what is best for Engineering.

The leadership team has to decide what is best for the company.

The CFO may know the most conservative financial choice.

The company may rationally accept greater financial risk to achieve a strategic objective.

Sales may want one customer commitment.

The organization may decide the operational cost makes it unattractive.

This is part of becoming an enterprise leader.

The executive stops thinking only:

How do I optimize my function?

and begins asking:

What decision best advances the organizational Mission and plan, given what we know about the whole system?

That is the Symbiosis mindset at the leadership level.

Shared Understanding Must Change as Conditions Change

Air traffic conditions are not static.

Weather changes.

Demand changes.

Airport capacity changes.

Flights move.

FAA CDM therefore depends not simply on creating one common picture but on continuously exchanging updated information so participants can adapt. The FAA's current traffic-flow modernization efforts explicitly use live data and predictive modeling to help stakeholders respond more proactively to changing conditions.

Companies need the same dynamic quality.

Alignment achieved during Annual Planning gradually decays as new information enters different parts of the company.

Sales hears something from customers.

Engineering learns something technically.

Finance gets new capital information.

The board changes its perspective.

A competitor moves.

A critical hire falls through.

The operating picture changes.

Peak's recurring cadence brings teams back together to update shared understanding.

Weekly Camp handles near-term execution.

Quarterly or Semiannual Sessions create a higher-level review of the One-Year Plan and next outcomes.

Annual Planning refreshes the longer horizon.

The shared picture is not an artifact.

It is something the organization repeatedly recreates.

External Partners Make the Team-of-Teams Problem Larger

This is where FAA CDM becomes particularly useful for frontier-tech companies.

Most business operating-system discussions assume the relevant teams are inside the company.

But many frontier companies execute through an ecosystem.

Aerospace organizations depend on suppliers, contract manufacturers, government customers, and research partners.

Defense companies may coordinate primes, subcontractors, federal customers, and outside technical partners.

Robotics companies may depend on component suppliers and deployment partners.

Advanced manufacturing may involve equipment providers, logistics partners, customers, and regulators.

No CEO controls the entire ecosystem.

This is closer to the FAA CDM problem.

Different organizations retain their own authority.

They have their own priorities.

They may have different information.

Yet their decisions affect the shared outcome.

The organization therefore needs mechanisms for asking:

What information do we need from the partner?

What information do they need from us?

What assumptions are we making about their capacity?

What commitments are they making?

Where is an external dependency becoming Off-Course?

Who owns the relationship?

When does an external issue become material to the company plan?

These questions belong in organizational execution even when the team sits outside the org chart.

Frontier Tech Needs Ecosystem Intelligence, Not Only Organizational Intelligence

This suggests an extension of the Organizational Intelligence concept.

A company can have excellent internal visibility while still being surprised by the system around it.

A critical supplier slips.

A government milestone changes.

A partner changes scope.

A contract manufacturer alters capacity.

A regulatory requirement changes.

The internal organization was aligned.

The ecosystem was not sufficiently visible.

FAA Collaborative Decision Making provides a useful reminder that complex execution sometimes requires a shared picture among actors who do not share one hierarchy.

For frontier-tech companies, the operating system therefore needs to account for consequential external dependencies.

Peak does not need to bring a supplier into Weekly Camp.

But the supplier dependency should become visible when it threatens an important OKR or One-Year Plan objective.

The internal owner needs to be clear.

The consequence needs to be understood.

The appropriate team needs to decide what to do.

External complexity has to enter the organizational operating picture somewhere.

Collaborative Decision Making Depends on Trustworthy Information

There is one more important lesson.

Shared information only helps if participants can rely on it.

FAA CDM membership and data sharing are governed through agreements and data-quality requirements because collaborative decisions depend on the quality of the information participants contribute.

That has a direct organizational parallel.

An executive team cannot build shared understanding when status updates are consistently optimistic.

A KPI nobody trusts is not useful.

An OKR marked On-Course because the owner fears reporting bad news weakens the entire system.

A forecast intentionally padded to avoid accountability reduces Organizational Intelligence.

The shared operating picture depends on people communicating reality.

This is why an Off-Course status should be treated as information.

If leaders punish bad news, people delay bad news.

Then the system becomes less intelligent.

Reliable information is an organizational behavior as much as a data-quality problem.

Better Decisions Come From Better Shared Context

The deepest lesson from FAA Collaborative Decision Making is not that every business decision should become collaborative.

It is not that companies should copy air-traffic-management processes.

And it is not that everyone should see every piece of information.

The stronger insight is:

When independent teams make decisions inside an interdependent system, the quality of those decisions improves when participants share enough context to understand conditions, constraints, preferences, and system-wide consequences.

That is a powerful Team-of-Teams principle.

Peak independently addresses the business version through:

shared Mission and direction,

company and functional outcomes,

visible KPIs,

clear Roles and Responsibilities,

up-, down-, and across-team planning,

Weekly Camp,

Triage,

and recurring learning.

The organization creates enough shared understanding for specialized teams to retain their expertise and authority while making decisions that work for the whole.

The Best Team-of-Teams Organizations Optimize the Whole Without Controlling Every Part

Complexity creates a leadership temptation.

Centralize.

Put the CEO in the middle.

Require more approvals.

Create one master process.

Make every function conform.

That can increase control.

It can also destroy speed and expertise.

The opposite extreme is equally dangerous.

Allow every function to optimize independently.

Now the company becomes a federation of silos.

FAA Collaborative Decision Making demonstrates a different possibility in a very different domain.

Different actors can retain different responsibilities and preferences while operating from a more shared understanding of the system and the consequences of their choices.

Peak points toward the same organizational balance.

Shared context without identical perspectives.

Clear authority without unnecessary centralization.

Functional autonomy without functional isolation.

Local expertise with awareness of system consequences.

Information sharing tied to decisions.

Team performance connected to organizational performance.

For CEOs building aerospace, defense, robotics, advanced manufacturing, autonomy, physical AI, and other frontier-tech organizations, this becomes increasingly important as the company and its ecosystem become more interdependent.

The question is no longer only:

Can each team make good decisions?

It becomes:

Can each team make good decisions while understanding enough of the larger system that those decisions combine into a good organizational outcome?

That is one of the central challenges of Team-of-Teams execution.

And it is one of the capabilities an organizational operating system should help create.

Related Insights

What Is Peak OS?

What Is Organizational Execution?

What Is Organizational Intelligence?

What Is a Business Operating System?

What Is Operating Rhythm?