Team & Org Communication Anti-Patterns

Team and organizational communication fails in ways that are structural, not individual. You can have skilled communicators in every seat and still end up with an organization where the right hand doesn't know what the left is doing. The anti-patterns below are emergent — they arise from how teams are structured, how incentives are aligned, and how information is (or isn't) designed to flow. Fixing them usually requires intervention at the team or org level, not at the individual level.

Structural Anti-Patterns

The Coordination Tax

Every cross-team decision requires 6 meetings and 4 Slack threads. Coordination overhead exceeds the value of the decisions being made.

Symptoms:
  - Engineers report "too many meetings" in surveys
  - Simple decisions take weeks
  - Nobody can name who owns what
  - "Let's set up a sync" is the reflexive answer to any
    ambiguity

Fix: Usually a Conway's Law problem. Teams are coupled too tightly because ownership is unclear. Reorganize so each decision has one owning team. Or reduce the number of dependencies between teams via better interfaces.

Silo Archipelago

Each team is internally high-functioning but invisible to other teams. Three teams solve the same problem independently. Breaking changes blindside downstream consumers.

Fix: Public channels by default. Shared roadmaps visible across teams. Guilds / communities of practice for horizontal knowledge flow. RFCs with explicit cross-team review requirements.

Matrix Mania

Everyone has two managers. Every decision requires four approvers. The matrix was meant to enable cross-functional work; it has become a veto forest.

Fix: Simplify. For each person, one primary manager, clear reporting line. For each decision, one DRI. Matrix structures can work but require obsessive clarity about decision rights; most orgs lack the discipline.

The Frozen Reorg

The last reorg was 5 years ago. The company has grown 5x. The structure is now obviously wrong — and nobody wants to be the one to propose changing it.

Fix: Periodic structural reviews (every 12-18 months minimum). Treat org design as a continuous practice, not a crisis response.

The Constant Reorg

The opposite: reorganizing every 6 months. Teams never develop stable identity. Knowledge is lost; relationships start over; nothing matures.

Fix: Reorg with purpose, not as a default move. Each reorg costs 3-6 months of productivity. Reserve for genuine misalignment.

The Overloaded Team

A team owns 15 services. Cognitive load far exceeds capacity. Work queues; engineers burn out; adjacent teams build workarounds.

Fix: Team Topologies cognitive-load framing. Cap what a team owns to what fits their cognitive capacity. Invest in platforms that reduce extraneous load. Split teams when consistently overloaded.

The Miniature Team

A 2-person "team" that owns a critical service. When one person leaves or goes on vacation, the service is at risk. The bus factor is 1.

Fix: Minimum viable team size of 3-4. If a service can't sustain a team of 3-4, it should be folded into a larger team's ownership.

Directional Flow Anti-Patterns

Broken Telephone

Strategy at the top. By the time it reaches ICs, it's unrecognizable. Each manager added interpretation; meaning was lost.

Fix: Leaders communicate directly to broad audiences in writing. Every IC can read the original source. Managers localize and reinforce but don't replace.

The Sanitized Upward Flow

Weekly statuses are all green. Quarterly reviews show all OKRs at 80%+. Reality is different: projects are at risk, engineers are burned out, trust is eroding. Leadership doesn't know.

Fix: Psychological safety (Edmondson). Skip-level 1:1s. Anonymous channels. Incident review without blame. Leadership must actively reward truth-telling, especially about bad news.

The Hidden Escalation

A team has a blocker. They don't raise it. They work around it for weeks until the blocker causes an incident. Post-mortem reveals they were afraid to escalate.

Fix: Explicit escalation norms. "If you're blocked for more than 2 days, escalate" as a cultural rule. Escalations are not failures; they're information.

The Executive Bubble

Executives talk only to executives. They haven't talked to a frontline engineer in 6 months. They don't know what the daily experience is like.

Fix: Skip-level 1:1s. MBWA (Management By Walking Around) — virtual equivalent for remote orgs. Leadership office hours. Engineering leaders should write code or join incidents periodically.

The All-Hands Theatre

All-hands meetings are performances, not communication. Slide decks polished; Q&A managed; no genuine information flows. Attendance is obligatory, but no one remembers what was said.

Fix: All-hands should have real substance. Honest updates (including bad news). Real Q&A (including anonymous submissions read aloud). If an all-hands can't generate any questions worth asking, cancel it.

The Meeting-Free Myth

An organization claims "meeting-free Wednesdays" or similar. In practice, meetings bleed onto adjacent days; "emergency" meetings on Wednesdays are constant. The norm is theatrical.

Fix: Actually defend the norm. Leadership models it. Escalation path for violators. If the norm is important, it must be enforced, not just stated.

Decision Anti-Patterns

The Consensus Trap

Every decision requires everyone's agreement. Decisions take forever. Quality suffers because nobody can make a bold call without offending someone.

Fix: Named decision-makers (DRI). Decision rights clear for each type of decision. Consensus is expensive; use it sparingly, for decisions with high implementation risk.

The HiPPO Decision

Highest-Paid Person's Opinion wins. Debate is theater; the outcome is pre-decided. Engineers stop engaging because their input doesn't matter.

Fix: Written decisions with documented alternatives (ADR / RFC). Structured input-gathering. Leaders who override expert input must explain why in writing. Track decision outcomes to see whether HiPPO decisions systematically underperform expert input.

The Committee That Decides Nothing

A group is formed to make a decision. 3 months later, still meeting. No decision. No clear DRI. Everyone defers to everyone else.

Fix: Every committee has a DRI with unilateral authority to make the call if consensus fails. Decision deadline set at formation. If deadline passes, DRI decides unilaterally.

The Escalation Cycle

A decision bounces up and down the hierarchy. Manager escalates to director; director asks manager to figure it out; manager doesn't have authority; back up. The issue dies without resolution.

Fix: Clear escalation authority. When something is escalated, the escalated-to person either decides or explicitly pushes back with reasoning. No silent ping-pong.

The Meeting That Replaces the Decision

A decision is scheduled for a meeting. The meeting happens. Attendees "align" on next steps (another meeting). No decision is made. The pattern repeats for 4 weeks.

Fix: Decisions are made in writing, with meetings as optional assists. Every meeting about a decision ends with the decision recorded, or with a named DRI committed to a decision deadline.

Cultural Anti-Patterns

The Fear-Driven Silence

Engineers see problems but don't raise them. Past experiences have shown that raising concerns damages their standing. Better to stay quiet and let it fail.

Fix: Psychological safety as active practice. Leaders explicitly reward people who raise concerns. Public acknowledgment of "near misses" that were caught early. Training on giving and receiving uncomfortable information.

The Blame Culture

When incidents happen, someone is blamed. Post-mortems become exercises in finger-pointing. Engineers hide mistakes; systems don't improve.

Fix: Blameless post-mortems. Focus on system and process fixes, not on individuals. Leaders model this by owning their own contributions to incidents.

Heroism Glorified

The engineer who worked 80 hours to save the launch is celebrated. The engineer who designed a system that didn't require heroism is invisible. Incentive distortion.

Fix: Recognize prevention and system design as first-class work. Promote engineers who make heroism unnecessary. Treat chronic heroism as a symptom of a broken system.

The Promotion Language Gap

Engineers who communicate loudly and confidently (often from specific demographic groups) advance faster than equally capable engineers who communicate differently. The org mistakes presentation for competence.

Fix: Promotion criteria based on impact, not presentation. Written artifacts as evidence (an RFC's quality is visible in a way that meeting presence isn't). Explicit attention to whose contributions are being noticed.

The Information Monopoly

A single person or small group holds critical context nobody else has. They're invaluable because only they know; they're also a single point of failure.

Fix: Written documentation requirements. Knowledge-sharing rituals. Rotation of key roles. The information monopolist should be encouraged (required) to write down what they know.

The Meeting-Centric Culture

Nothing is real until it's discussed in a meeting. Written communication is treated as second-class. Engineers who prefer writing are at a disadvantage.

Fix: Shift to written-first norms (see Async & Remote topic). Leadership models writing. Meetings become exceptional, not default.

Scale Anti-Patterns

The Startup-Forever Organization

A 300-person company operates with the same informal structures it had at 30. No written strategy. Slack for everything. "We all know each other" — except the newest 250 people don't.

Fix: Each scale break (30 → 100 → 300 → 1000) requires structural upgrades. Invest proactively in practices before they become crises.

The Premature Enterprise

A 40-person company imports enterprise practices wholesale. Weekly architecture review board. Five-layer approval for any change. Process overhead exceeds productive work.

Fix: Scale processes to company size. 40-person companies need lightweight processes. Import practices selectively, not wholesale.

The Growth Freeze

The company has doubled, but the founder still makes every decision. Bottleneck is structural. Decision velocity is crushed.

Fix: Founder must delegate actually — not ceremonially. Accept that delegated decisions will sometimes be different from what the founder would have made; that's the cost of scale.

The Middle-Management Void

Company grew fast but didn't invest in middle management. Each manager has 15 reports. Managers can't manage; everyone suffers.

Fix: Span-of-control norms (usually 5-8 direct reports). Invest in middle-management development. Sometimes hire experienced middle managers rather than promoting from within.

Tool & Artifact Anti-Patterns

Tool Sprawl

Information lives in 12 different tools. Slack + Email + Jira + Notion + Confluence + GitHub + Linear + Asana + ... Nobody can find anything.

Fix: Canonical tools by type. Decisions = one tool. Tasks = one tool. Docs = one tool. Chat = one tool. Ruthlessly consolidate. Migration is painful once, ongoing sprawl is painful forever.

Slack as Everything

Chat app used for: decisions, docs, project management, incident response, social conversation. Decisions evaporate when the thread scrolls past.

Fix: Use chat for chat. Move decisions into docs. Move tasks into a task tracker. Move incidents into incident channels with post-incident write-ups.

The Wiki Graveyard

A wiki or handbook is started with enthusiasm. Six months later, half of it is stale. Everyone has learned not to trust it.

Fix: Ownership per page. Review triggers. A stale wiki is worse than no wiki. Invest in maintenance, or accept that docs will degrade and plan accordingly.

Notification Overload

Every Slack channel notifies everyone. Every Jira ticket notifies the watchers. Every PR pings the reviewers. Nobody can focus.

Fix: Notification hygiene. Defaults off; opt-in per channel. Tags used intentionally. Teach engineers to configure their notifications.

Meta-Antipatterns

Framework Tourism

The org adopts framework after framework (Scrum, then SAFe, then Team Topologies, then Spotify Model). None of them stick because none of them were genuinely adopted.

Signal: A new framework announced every 12-18 months; none durably in practice.

Fix: Pick a framework that fits the organization. Commit to it. Iterate inside it rather than replacing it. Most frameworks work if genuinely adopted; none work if treated as veneer.

1. Can a new engineer find canonical docs for: strategy, decisions,
   how work flows, how decisions are made?
2. Does bad news travel up? Can leaders name risks their teams
   are managing?
3. When two teams need to coordinate, is there a clear mechanism?
4. Can you draw the decision-rights diagram for any active project?
5. What communication norms do we have written down, and are they
   actually practiced?
6. Is cognitive load appropriate for each team?
7. Is information flowing across (horizontal) as well as up/down
   (vertical)?

For a team:

1. Does everyone on my team know what the team is optimizing for
   this quarter?
2. Can I name the work happening in adjacent teams that might
   affect us?
3. When was the last time we updated our team handbook?
4. Are my team's interactions with other teams clean (one mode
   per pair) or blurred?
5. Who is the DRI for each of our active initiatives?