Skip to content

Toolkit — Anti-Pattern Playbook

This playbook is a set of field countermeasures for common system failures. Each entry is structured for fast use:

  • What you’ll see (symptoms)
  • What’s really happening (mechanism)
  • Fast test (how to confirm)
  • Countermove (artifact / constraint / authority / cadence)
  • Default (what to do if people avoid the fix)

Use it in reviews, retro-style sessions, and whenever “process” starts feeling heavy but outcomes don’t improve.


Anti-Pattern 1 — “Alignment” as a Goal

What you’ll see

  • “We need alignment” becomes the problem statement.
  • Meetings increase; decisions don’t.
  • People leave with different interpretations.

What’s really happening

A failing decision type is being disguised to avoid conflict.

Fast test

Ask: “Alignment about which decision?” If nobody can answer in one sentence, it’s not alignment—it’s avoidance.

Countermove

  • Decision: force a decision type (priority/scope/ownership/…)
  • Artifact: create a decision log entry with “Decision changed” field
  • Constraint: timebox the decision + default outcome

Default

If the decision isn’t named in 10 minutes, stop the meeting and assign an owner to produce an Observable Failure Statement.


Anti-Pattern 2 — Ritual Without Artifact

What you’ll see

  • Recurring meetings exist “because we always do them.”
  • Outcomes are verbal; nothing persists.
  • The same topics reappear weekly.

What’s really happening

Cadence is substituting for a decision machine.

Fast test

Ask: “Show me the last 3 artifacts this meeting produced.” If there are none, it’s a ritual.

Countermove

  • Artifact: define a single required output (one page / one table)
  • Constraint: “No artifact, no meeting”
  • Cadence: shorten or remove until artifact exists

Default

If artifact isn’t produced, automatically cancel the next occurrence and run async.


Anti-Pattern 3 — Artifact Without Decision

What you’ll see

  • Dashboards/slides/docs are produced regularly.
  • People feel “informed,” but actions don’t change.
  • Artifacts are optimized for polish.

What’s really happening

Reporting is replacing decision-making.

Fast test

Ask: “Which decision changed because of this artifact?” If there’s no answer, it’s theater.

Countermove

  • Artifact change: add mandatory fields:

  • “Decision changed”

  • “Owner”
  • “Next action + date”
  • Constraint: artifact must contain a decision or it’s invalid

Default

If no decision is recorded, the system defaults to the previous decision (explicitly) and escalates the missing decision to the decider.


Anti-Pattern 4 — Constraintless “Guidelines”

What you’ll see

  • “It’s just guidance.”
  • Exceptions are constant.
  • Under stress, the system vanishes.

What’s really happening

There is no enforcement mechanism; the system is optional.

Fast test

Ask: “What happens when someone ignores this?” If the answer is “nothing,” it’s not a system.

Countermove

  • Constraint: add one enforceable rule
  • Default: define automatic behavior when ignored
  • Authority: name who enforces it

Default

If enforcement authority can’t be named, redesign the system to a smaller unit of analysis where authority exists.


Anti-Pattern 5 — Framework Stacking

What you’ll see

  • New systems are added; none are removed.
  • Multiple sources of truth exist for the same decision.
  • Teams spend time translating between systems.

What’s really happening

Systems collide or duplicate decision ownership.

Fast test

Ask: “Which artifact is the source of truth for priority/scope/etc.?” If there are two, you have a collision.

Countermove

  • Precedence rule: declare which system wins for that decision domain
  • Removal/subordination: retire or subordinate one system
  • Landscape matrix: map collisions explicitly

Default

No new system may launch until it names what it replaces/subordinates and where the single source artifact will live.


Anti-Pattern 6 — Consensus Veto Everywhere

What you’ll see

  • Decisions take too long.
  • “We need more buy-in” repeats.
  • The most risk-averse actor controls outcomes.

What’s really happening

Consensus creates distributed veto power and rewards avoidance.

Fast test

Ask: “Who decides if we disagree?” If unclear, you don’t have decision rights.

Countermove

  • Authority boundary: name the decider
  • Participation rule: consult vs inform
  • Constraint: timeboxed decision window + default outcome
  • Artifact: decision log with dissent recorded

Default

If the group can’t agree on a decider, escalate that as the decision to leadership immediately (ownership is the blocker).


Anti-Pattern 7 — Metric Capture

What you’ll see

  • Metrics improve; reality worsens.
  • People hide work to protect numbers.
  • Metrics become performance weapons.

What’s really happening

Measurement became an incentive system.

Fast test

Ask: “Who gets punished or rewarded by this metric?” If the answer is “individuals/teams,” gaming is predictable.

Countermove

  • Separate:

  • learning metrics (diagnosis)

  • evaluation metrics (performance)
  • Add “decision link”:

  • metrics must map to an action decision log entry

Default

If a metric is used for evaluation, it must have an explicit anti-gaming review and be paired with at least one qualitative truth check (incidents, audits, samples).


Anti-Pattern 8 — Legibility Over Truth

What you’ll see

  • Status looks green until failure is unavoidable.
  • Bad news travels slowly.
  • Artifacts are sanitized.

What’s really happening

Artifacts optimize safety/politics over reality.

Fast test

Ask: “What does this artifact make harder to say?” If uncertainty and risk aren’t representable, truth is being filtered.

Countermove

  • Artifact: require fields for:

  • uncertainty

  • risks with owners
  • assumptions
  • “what would change our mind”
  • Constraint: red/yellow flags require next action + date

Default

If uncertainty can’t be written, the artifact is invalid and must be replaced by a simpler decision-focused artifact.


Anti-Pattern 9 — “Everything Is Urgent”

What you’ll see

  • Priorities shift constantly.
  • Work starts but doesn’t finish.
  • Teams are always “busy,” outcomes lag.

What’s really happening

No enforced prioritization and uncontrolled WIP.

Fast test

Ask: “What are the top 3 priorities and what are we not doing?” If “not doing” is empty, priority is fake.

Countermove

  • Constraint: WIP limit or initiative cap
  • Default: new work displaces lowest-priority in-progress work
  • Artifact: ranked stack with capacity allocation

Default

If leadership introduces urgent work, they must name what it displaces (explicit tradeoff rule).


Anti-Pattern 10 — Ownership Fog (“Not My Problem”)

What you’ll see

  • Work gets blocked on dependencies.
  • Escalations replace collaboration.
  • Interfaces and responsibilities are disputed.

What’s really happening

Ownership and interfaces are the real object of control, but the system is trying to control “communication.”

Fast test

Ask: “Who owns this interface/domain and what authority do they have?” If unclear, you have boundary failure.

Countermove

  • Artifact: ownership map + interface contract
  • Constraint: every interface has an owner; changes require owner approval
  • Authority: escalation path defined

Default

If ownership is disputed, default ownership goes to the team operating it in production until reassigned by a named authority.


Anti-Pattern 11 — Hero-Dependent Systems

What you’ll see

  • One person makes the system “work.”
  • Everyone says “ask Alex.”
  • Progress stalls when the hero is absent.

What’s really happening

Interpretation and enforcement live in a person, not in artifacts and defaults.

Fast test

Ask: “Could a new hire run this system from artifacts alone?” If not, it’s hero-dependent.

Countermove

  • Move rules into:

  • decision logs

  • explicit constraints and defaults
  • ownership map
  • Reduce reliance on tacit knowledge

Default

If a step requires a person’s memory, treat it as a missing artifact and stop the workflow until it’s captured.


Anti-Pattern 12 — Fossilized Systems (Can’t Die)

What you’ll see

  • Nobody remembers why the system exists.
  • Removing it feels dangerous.
  • The system survives via identity.

What’s really happening

No review cadence and no retirement mechanism.

Fast test

Ask: “What failure was this designed to prevent?” If unclear or obsolete, it’s fossilized.

Countermove

  • Add:

  • sunset clause

  • review cadence
  • kill criteria (evidence-based)
  • Run the System Fitness Checklist

Default

If the system can’t justify itself within one review cycle, schedule retirement and protect a replacement only if a proven gap remains.


Fast “What Should We Change First?” Rule

When a system is failing, change in this order:

  1. Constraint + default (teeth)
  2. Artifact (inspectability)
  3. Authority boundary (enforcement)
  4. Cadence/trigger (operational fit)
  5. Unit of analysis (scale validity)

If you start with cadence or templates, you usually get theater.


Mini Index (Anti-Pattern → Countermove)

  • Alignment talk → force decision type + timebox + default
  • Too many meetings → “no artifact, no meeting”
  • Pretty reports → add “decision changed” + next action
  • Optional guidelines → add enforceable constraint + default
  • Too many frameworks → precedence + remove/subordinate
  • Consensus slow → decider + consult/inform + default
  • Metrics gamed → split learning vs evaluation + decision link
  • Green status lies → uncertainty/risk fields + red requires action
  • Everything urgent → WIP/initiative cap + displacement default
  • Blocked dependencies → ownership map + interface contract + escalation
  • Hero hub → externalize into artifacts + defaults
  • Fossilization → sunset clause + kill criteria