10

Apply it on Monday

The book is a toolbox, not a methodology: you pick the principles that pay off in your context. Start by scoring your team against the twelve problems from chapter 1 — your answers are stored only in this browser — and get a prioritized list of first moves.

How true is this of your team?

  1. 01Failure to correctly quantify economics

    Track proxy variables — efficiency, schedule, conformance — because nobody can say what a week of delay costs.

  2. 02Blindness to queues

    Design-in-process inventory is invisible — it's information, not parts — so nobody measures it.

  3. 03Worship of efficiency

    Keep every engineer 100% busy. Idle time looks like waste.

  4. 04Hostility to variability

    Variability is bad; eliminate it with Six Sigma-style process control.

  5. 05Worship of conformance

    Success means hitting the original plan, spec and budget.

  6. 06Institutionalization of large batch sizes

    Phase-gate processes move the whole project as one batch through requirements, design, test.

  7. 07Underutilization of cadence

    Meetings, reviews and releases happen when someone asks for them.

  8. 08Managing timelines instead of queues

    Monitor whether activities are on schedule — a lagging indicator.

  9. 09Absence of WIP constraints

    Start work as soon as it's approved; more started work looks like more progress.

  10. 10Inflexibility

    Specialists, fully loaded, each doing exactly their own role.

  11. 11Noneconomic flow control

    First in, first out — or whoever shouts loudest gets served first.

  12. 12Centralized control

    Decisions escalate to managers who see the big picture.

Your profile

123456789101112outer ring = “exactly us” — the bigger the shape, the more flow there is to win

Your playbook, most painful first

  1. 1.Failure to correctly quantify economicsSomewhat

    For your top 5 backlog items, estimate cost of delay per week. A rough number agreed in 30 minutes beats none.

    Measure: Every prioritization discussion references $/week of delay.

    E1E3E5

  2. 2.Blindness to queuesSomewhat

    Make queues visible: one board with every item in flight, including items waiting for review, test, decisions.

    Measure: Queue size per stage, tracked weekly, plotted as a cumulative flow diagram.

    Q1Q11Q15

  3. 3.Worship of efficiencySomewhat

    Stop targeting 100% utilization. Leave deliberate slack on bottleneck people and environments (aim ~70–80%).

    Measure: Wait time in front of the bottleneck, not its busyness.

    Q3Q9

  4. 4.Hostility to variabilitySomewhat

    Run more, cheaper experiments; cap the downside of each (feature flags, spikes, prototypes).

    Measure: Number of experiments per month and cost per experiment.

    V2V7

  5. 5.Worship of conformanceSomewhat

    Replace 'did we hit the plan?' with 'did we update the plan when we learned something?' at reviews.

    Measure: Decisions changed by new information, celebrated not punished.

    V1E9

  6. 6.Institutionalization of large batch sizesSomewhat

    Halve the batch: smaller PRs, smaller releases, smaller funding increments, smaller requirements chunks.

    Measure: Median PR size, deploy frequency, items per release.

    B1B3B12

An order of operations that usually works

Each step makes the next one cheaper. Economics first, because without a cost of delay you can't tell which of the others is worth doing.
  1. step 1
    Price delay

    Agree a cost of delay per week for the work that matters. Rough is fine; it only has to be right enough to rank decisions.

  2. step 2
    See the queues

    Put every in-flight item on one board, including waiting states. Plot a cumulative flow diagram weekly.

  3. step 3
    Cap WIP

    Set WIP limits per stage. Finish before you start. Watch cycle time fall while throughput barely moves.

  4. step 4
    Shrink batches

    Smaller PRs, releases, requirements and funding increments. Invest in lowering transaction cost (automation) so small is cheap.

  5. step 5
    Sequence by WSJF

    Order the queue by cost of delay ÷ duration. Recompute when either changes.

  6. step 6
    Add cadence

    Fixed rhythm for planning, reviews, demos and releases. Synchronize dependent teams on harmonic cadences.

  7. step 7
    Tighten loops

    Measure the time from change to signal (CI, review, customer). Attack the longest loop first.

  8. step 8
    Decentralize

    Publish decision rules and intent; let the people closest to the information decide within them.

Reinertsen's own advice: you don't need to adopt everything. Pick the principle with the biggest economic payoff in your situation, apply it, measure, repeat. The one non-negotiable is the first — if you don't quantify cost of delay, every other trade-off is decided by opinion.