13. Network of Teams

Networks of teams balance autonomy with structured coordination through explicit interfaces and protocols.

Hierarchies are dead, but networks need structure

In 2010, a catastrophic explosion ripped through BP’s Deepwater Horizon oil rig, killing 11 workers and triggering what is widely regarded as the largest accidental marine oil spill in history.[1] Subsequent investigations revealed a confluence of technical failures—including problems with the blowout preventer and cement work.[2] Multiple reports also highlighted critical safety warnings that did not receive sufficient attention at senior levels; organizational silos and communication breakdowns contributed to a failure to act on these warnings.


  1. “Overview Page: Deepwater Horizon Oil Spill,” National Oceanic and Atmospheric Administration, accessed February 2025, https://response.restoration.noaa.gov/deepwater-horizon-oil-spill-case-study. ↩︎

  2. “Investigation Report Volume 1: Explosion and Fire at the Macondo Well,” U.S. Chemical Safety and Hazard Investigation Board, June 5, 2014, https://www.csb.gov/assets/1/7/vol_1_final.pdf. ↩︎

This tragedy illustrates the flaw in traditional organizational hierarchies. Designed for stable, predictable environments, these structures can fragment information as it moves up the chain, distorting reality with each layer it passes through.[1] By the time insights reach decision-makers, they may be sanitized, delayed, or lost entirely. Meanwhile, frontline workers—those closest to customers and operations—spend more energy satisfying bosses than solving real problems.


  1. Jay R. Galbraith, Designing Complex Organizations (Addison-Wesley, 1973). ↩︎

Yet organizations that simply declare “we’re flat now” or remove management layers without creating alternative structures often find themselves in equally dangerous territory. Decisions stall as no one knows who can make the call. Information scatters across disconnected teams. Hidden power dynamics grow more toxic than formal hierarchies.

Therefore...

Design the organization as a network of teams that connects autonomous but interdependent units around actual value creation. Unlike a traditional hierarchy organized by function or division, a network organizes teams end to end around customer journeys, products, or services—with explicit protocols for how they interact.

As a general example, imagine a large, traditional bank reorganizing its operations using this approach. Rather than structuring around traditional banking products or functions (loans, deposits, know your customer), they could organize into customer journey teams—“onboarding,” “everyday banking,” “lending”—each with all the skills needed to deliver their part of the customer experience. These teams could connect through explicit interfaces rather than managerial layers. In other similar situations, teams organizing in this way have reduced product development cycles from 18–24 months to less than five weeks.

A well-designed Network of Teams delivers immediate benefits in speed, innovation, and engagement. Teams can respond to customer needs without waiting for multiple approvals. Information flows directly to where it’s needed rather than through distorting hierarchical filters. People feel greater ownership and purpose when they see how their work directly contributes to outcomes.

The long-term benefit is resilience: networks reconfigure easily and readily as conditions change. Teams can form, merge, split, or dissolve based on emerging needs rather than through disruptive reorganizations.

You can apply this pattern in three phases: network design, network infrastructure, and network implementation. Consider testing this in a discrete area of the organization before expanding it to the broader whole.

1. Network Design

Network design goes beyond the sticks and boxes of an org chart. Effective teams in a network structure require thoughtful boundaries and interfaces that enable autonomy and coordination across the organization. Teams should be designed with clear ownership, direct connections to other teams, and transparent work processes.

Network Design Principles

Purposeful: Design the teams around complete chunks of value delivery, following the purpose, customer, platform (16) pattern. Each team should own a discrete domain (23) with clear customer-facing outcomes. At Spotify, each squad owns a specific aspect of the user experience (e.g., search functionality, playlist management, artist pages) with all the skills needed to deliver that feature from concept to production.[1]


  1. Henrik Kniberg and Anders Ivarsson, “Scaling Agile @ Spotify,” Crisp, October 2012, https://blog.crisp.se/wp-content/uploads/2012/11/SpotifyScaling.pdf. ↩︎

Connected: Enable teams to interact laterally without managerial intermediaries. Establish clear protocols for cross-team collaboration using consent & consensus (10) and structured decision-making (30). Morning Star creates explicit colleague letters of understanding (40) documenting how teams will interact, enabling coordination without traditional management.

Legible: Make work and dependencies transparent across team boundaries using work in public (50) and visualizing workflows through kanban (55). are.na has a public road map for its product, hosted on its website, for customers and backers to see and review.[1]


  1. “Roadmap,” Are.na, accessed February 26, 2025, https://www.are.na/about#roadmap. ↩︎

2. Network Infrastructure

Beyond individual teams, a network requires supporting infrastructure that enables coordinated action without hierarchical control. This infrastructure creates the conditions for autonomous teams to align their efforts toward common goals while sharing knowledge across boundaries.

Infrastructure Elements

Shared Purpose & Principles: Teams’ purposes will be informed and constrained by true purpose (1) and rule of law (2), holding the network together, providing alignment without control. W.L. Gore & Associates unites its network with four simple principles that guide all decisions: freedom, fairness, commitment, and waterline (don’t sink the ship).[1]


  1. “The Gore Story,” Gore, accessed February 2025, https://www.gore.com/about/the-gore-story. ↩︎

Good Boundaries: Create explicit agreements about where one team’s responsibility ends and another’s begins using domains, assets & standards (23) and boundary management (67). Amazon uses APIs between teams that clearly specify what each team provides to others and what they can expect in return.[1]


  1. “Stevey’s Google Platforms Rant,” GitHub Gist, published October 12, 2011, https://gist.github.com/chitchcock/1281611. ↩︎

Cross-Network Learning: Establish forums for knowledge sharing across the network through guilds (21), chapters (22), and regular retrospectives (66). Valve Corporation uses temporary interest groups to share learning and solve crosscutting problems without creating permanent hierarchical structures.[1]


  1. Valve, Handbook. ↩︎

3. Network Implementation

Moving from concept to reality requires a deliberate, step-by-step approach that works through both structural and cultural change. To transform your organization into a Network of Teams:

Network Implementation Phases

  1. Map value flows: Start by tracking how work actually moves through your organization to deliver customer value. Ignore official org charts and focus on real activity flows.
    1. Map the end-to-end journey of your products or services from conception to customer.
    2. Identify key handoffs, bottlenecks, and broken connections.
    3. Note where current reporting lines help or hinder this flow.

A large hospitality company mapped its guest experience journey from booking to checkout, discovering 32 separate systems and nine departments touched a single reservation. Despite having an official five-step process, frontline staff used 17 unofficial workarounds to meet guest needs promptly. Room maintenance requests averaged 26 hours for resolution due to handoffs between departments, though guests expected fixes within four hours.

  1. Design team boundaries: Create teams with end-to-end responsibility for discrete value streams, perhaps guided by the purpose, customer, platform (16) model.
    1. Purpose teams: Organized around specific customer needs or journeys
    2. Customer teams: Focused on distinct market segments or user types
    3. Platform teams: Providing services and capabilities that support multiple teams

For each team, clearly define:

  • The specific outcomes they own
  • Their scope of authority (what decisions they can make autonomously)
  • The resources and capabilities they need to deliver their outcomes

The company reorganized from traditional departments (housekeeping, front desk, food service) into journey-based teams responsible for complete guest experiences. They created arrival & welcome teams, stay-experience teams, and departure & return teams, each with staff from multiple disciplines empowered to resolve issues without escalation. Each property also established a small special requests team with expertise in handling unusual situations.

  1. Design interfaces: Make the connections between teams explicit and efficient.
    1. Create service agreements that clearly state what each team provides to others.
    2. Establish communication protocols (when teams sync, how they share updates).
    3. Define escalation paths for conflicts or dependencies.
    4. Build shared visibility systems so teams can see each other’s work.

They developed service agreements between teams that clearly specified handoffs and responsibilities. For example, the stay-experience and departure teams created explicit protocols for how guest checkout information would transition, including timing, data sharing, and exception handling. They implemented a digital guest journey board visible to all teams showing real-time status of all guests, enabling teams to coordinate without constant meetings.

  1. Build network capabilities: Invest in developing the skills needed to operate in a network.
    1. Train teams in conflict resolution and negotiation.
    2. Develop emergent leadership (58) across the organization.
    3. Teach effective meeting facilitation using facilitation (56).
    4. Build capability in giving and receiving advice (31).

Before fully implementing the network structure, the company ran decision simulations where teams practiced handling guest scenarios that crossed traditional boundaries. They created a network playbook with common terminology and protocols for cross-team collaboration. Every employee completed a workshop on collaborative problem-solving focusing on how to navigate situations without escalating to managers. They started with low-season pilot properties to refine their approach before rolling out company-wide.

  1. Create feedback mechanisms: Establish ways to continuously evolve the network.
    1. Regular health check (59) processes to assess network effectiveness
    2. Cross-team retrospectives (66) to identify improvement opportunities Periodic network mapping to spot emerging bottlenecks or gaps
    3. dissolvability (9) protocols to reconfigure teams as needs change

The hospitality company implemented experience retrospectives where representatives from all teams reviewed the complete guest journey data weekly, identifying improvement opportunities across team boundaries. They created shared metrics focused on guest outcomes rather than departmental KPIs. Then, they established a quarterly network evolution process where teams could propose boundary adjustments based on guest feedback and operational data.

Dealing with pushback

Resistance to this approach comes in a few forms:

“This will create chaos without clear reporting lines.”

Network structures actually reduce chaos by making coordination explicit rather than dependent on hierarchical authority. Traditional reporting lines often mask underlying confusion about who decides what. Establish clear decision protocols through structured decision-making (30) and build capability in conflict resolution (39).

“We’ll lose economies of scale without Centers of Excellence.”

Shared services and capabilities can still exist as platform teams (17) that support multiple value streams. These teams maintain specialized expertise while connecting directly to customer-facing teams instead of operating as isolated silos.

“Leaders won’t give up control.”

Control isn’t going away. It’s being redistributed to where it’s most effective, and we are gaining more ways to adjust as we learn. Use expanded available power (4) principles and begin with pilot teams to demonstrate success before scaling.

Reader resource

Use these slides to introduce Network of Teams and put it to work with your colleagues.

Download the chapter slides (PDF) · Version 1.2

Send me a receipt showing that you pre-ordered the book, and then sign in. You'll get a magic link and then all these chapters are yours to read.