Principles Library / Second-Order Effects

Second-Order Effects

New Solutions Create New System Constraints

Act

Showing all 5 practices. Answer the Diagnose questions and the most relevant come first.

Name the new constraints before you commit

Before you approve a solution, list the new constraints it's likely to create and where they'll land.

Start by asking your team: What new constraints will this solution introduce?

Avoid · You’ll know it’s working when

Avoid

  • Assuming solutions are purely additive
  • Optimizing local problems without evaluating system impact

You’ll know it’s working when

Bottlenecks clear instead of moving, and new fixes rarely catch teams off guard.

Make constraint shifts explicit before implementation

Before you implement, write down which teams or systems will absorb the shifted constraints.

Start by asking your team: Which teams or systems will absorb those constraints?

Avoid · You’ll know it’s working when

Avoid

  • Treating second-order effects as unexpected
  • Treating second-order effects as unavoidable

You’ll know it’s working when

Bottlenecks stay put or disappear, and failures from recent changes are expected rather than surprising.

Assign owners for new components and side effects

When you add a component, name who owns it and who owns the side effects it creates.

Start by asking your team: What new knowledge, ownership, or coordination will be required?

Avoid · You’ll know it’s working when

Avoid

    You’ll know it’s working when

    Knowledge and control spread beyond the new components, and recent changes stop surprising teams.

    Remove something each time you add something

    When you add a tool, process or layer, remove or simplify an existing element.

    Avoid · You’ll know it’s working when

    Avoid

    • Responding to new constraints by adding more layers

    You’ll know it’s working when

    New layers are rare, and complexity holds steady while outcomes improve.

    Set human oversight before adding agents

    When you introduce agents, decide who monitors them, who steps in and who adapts them as constraints surface.

    Start by asking your team: As automation increases, what new human responsibilities emerge?

    Avoid · You’ll know it’s working when

    Avoid

    • Introducing AI or automation without redefining ownership and failure modes

    You’ll know it’s working when

    Agent changes arrive with their constraints named and owned, and downstream surprises from them are rare.

    Overview

    Principle intent

    Every solution introduced to address a problem creates new constraints elsewhere in the system.

    Read the full intent

    System health depends on anticipating and managing these second-order effects, not just solving the immediate issue.

    When it holds: what it does for you

    Fixes that stay fixed

    Bottlenecks get cleared instead of moved somewhere else.

    Fewer surprise failures

    Teams know which constraints a change brings before it goes live.

    A simpler system

    Something old goes out each time something new comes in.

    When it is ignored: from early signs to a lasting condition

    Early warning signs · Visible now, cheap to fix

    • New tools, processes or layers keep appearing to fix isolated problems.
    • After a new tool or process goes in, the bottleneck shows up somewhere new.
    • Teams are surprised by failures caused by recently introduced solutions.

    What happens over time · Accumulates, costly to reverse

    • Each new solution adds constraints faster than old ones get resolved.
    • Reliability drops as hidden dependencies multiply.
    • Improvement work loses credibility because every fix brings new problems.

    The condition it becomes

    Local Optimization Bias · Secondary

    Every team hits its targets. The product still misses.

    How this principle feeds it

    a new solution improves one area while its constraints land elsewhere, so end-to-end flow may not improve.

    Workload Saturation · Secondary

    Everything is in progress. Nothing is finishing.

    How this principle feeds it

    work queues at each new bottleneck, and compounding constraints leave more work in progress and less finishing.

    Dependency Density · Secondary

    The team owns the commitment. Nobody owns the gap.

    How this principle feeds it

    each new part adds interfaces owned by different people, so work needs more coordination to flow.

    Quality Fragility · Secondary

    Customers find defects before the team does.

    How this principle feeds it

    failures from dependencies nobody knew about surface late as incidents, though the root is added coupling.

    Attribution Failure · Primary

    We know what happened. Nobody can explain why.

    How this principle feeds it

    the failure is recorded against the team where it appeared, while the design decision that caused it goes untraced.

    Accountability Fragmentation · Secondary

    Accountability is assigned. Control is not.

    How this principle feeds it

    downstream teams answer for failures created by changes they did not control.

    Strategic Volatility · Secondary

    Every quarter starts over. Nothing compounds.

    How this principle feeds it

    leaders drop each backfiring initiative and start another, so improvement never compounds.

    Oversight Erosion · Secondary

    The review process still runs. It stopped meaning anything.

    How this principle feeds it

    oversight sized for the original deployment cannot keep up with constraints added at machine speed.

    Where is your system on this path?

    Diagnose

    Diagnose

    AI signs are on by default. Use the AI switch to exclude them.

    1. Which of these early warning signs are you seeing?

    • New tools, processes or layers keep appearing to fix isolated problems.
    • After a new tool or process goes in, the bottleneck shows up somewhere new.
    • Teams are surprised by failures caused by recently introduced solutions.
    • The system keeps getting more complex while overall outcomes stay flat.
    • Knowledge or control piles up around the new components.
    • Automation or AI fixes a local problem but adds coordination or oversight work. (AI)
    • Agent-introduced solutions keep causing downstream problems nobody planned for. (AI)
    • Constraints change faster than the organization can map and understand them. (AI)
    • New agent capabilities bring constraints that nobody names or manages. (AI)

    2. Over the last few months, have you noticed…

    For each consequence: Not seen / Sometimes / Persistently / Not sure.

    • Each new solution adds constraints faster than old ones get resolved.
    • Complexity grows faster than the organization's ability to handle it.
    • Reliability drops as hidden dependencies multiply.
    • Failures get blamed on downstream teams instead of the design that caused them.
    • Improvement work loses credibility because every fix brings new problems.
    • Poorly understood agent solutions spread constraints fast and hide the cause. (AI)

    3. The condition it becomes

    Based on your answers, recognize at most two conditions: No / Partly / Yes, that is us.

    Local Optimization Bias: Every team hits its targets. The product still misses.

    Workload Saturation: Everything is in progress. Nothing is finishing.

    Dependency Density: The team owns the commitment. Nobody owns the gap.

    Quality Fragility: Customers find defects before the team does.

    Attribution Failure: We know what happened. Nobody can explain why.

    Accountability Fragmentation: Accountability is assigned. Control is not.

    Strategic Volatility: Every quarter starts over. Nothing compounds.

    Oversight Erosion: The review process still runs. It stopped meaning anything.

    Take your Diagnosis to the next level by invoking AI Diagnostic Coaching tool.

    What else have you observed? Optional. Shared only if you choose to start coaching.

    Start a Diagnosis session: review and edit a draft containing your note, conditions you recognized in Step 3 (including whether they match partly), warning signs and consequence ratings before sending it. No scores or principle files are sent.

    See what practices will help you