KSD flashcards
KSD · Build recall with a focused review session.
33 cards. Read the term and the definition here. The deck at the top of this page is optional.
- Service purpose
- A one-sentence statement naming the customer need the service meets, the outcome it produces, and what it deliberately does not do, so every later policy decision has an anchor.
- Value vs constraint
- Value is defined by the customer's changed condition; internal stakeholders and approvals are constraints on how value is delivered, not the definition of value itself.
- Demand and capacity
- Demand is the rate and pattern of work arriving at the system; capacity is the rate the system can complete work. Purpose is achievable only when both are understood together.
- Purpose drift
- The condition where work items accumulate that no stated purpose explains, consuming capacity and queue space without producing a customer outcome.
- Classes of service
- Named categories of work, such as expedite, fixed-date, standard, and intangible, each with explicit policies for priority, WIP allocation, and expected time in system.
- Expedite policy
- A reserved capacity slot and a hard cap on expedited items, each requiring explicit sponsor approval, so urgent work cannot permanently starve standard work.
- Fixed-date class
- Work pulled by date risk rather than arrival order; the start point is calculated from historical lead time and the due date, at the last responsible moment.
- Intangible class
- Low or deferred cost-of-delay work that is pulled only when capacity is otherwise idle, so it never displaces work that carries real customer value.
- Written policy
- A rule stating who decides, on what information, within what limit. Unwritten conventions are not policies because they cannot be audited or changed deliberately.
- Definition of ready
- An entry policy at the start of the workflow specifying the conditions a work item must satisfy before it may be pulled into the system.
- Little's Law
- Average cycle time equals average WIP divided by average throughput. It predicts the lead-time effect of a WIP change before the change is made.
- Work item age
- Elapsed time since an unfinished item started. Items older than the 85th percentile of historical cycle time are at risk now, not at their deadline.
- Lead time vs cycle time
- Lead time runs from customer request to delivery; cycle time runs from work start to finish. Report which clock is used or the numbers are not comparable.
- Cumulative flow diagram
- A stacked chart of items per stage over time. Widening bands show where work accumulates; flat arrival and departure lines indicate balanced flow.
- Next experiment format
- Hypothesis, one policy or limit change, metric and threshold, time box, and a decision rule to adopt or revert. One variable per experiment keeps results interpretable.
- Feedback loop cadence
- Scheduled reviews, such as a daily look at blocked and aging items and a periodic review of policies and classes of service, each tied to data rather than anecdote.
- Value stream mapping
- Charting the full path from customer request to outcome, including queues, handoffs, approvals, and rework loops, not only the team's own steps.
- Local optimization
- Improving one station's cycle time while increasing total lead time. The test is whether end-to-end flow improves, not whether a single step looks faster.
- Queue and utilization
- Queue size grows sharply as capacity utilization approaches 100 percent, producing long and unpredictable lead times even when every worker appears busy.
- Bottleneck rule
- System throughput is set by the constraint. Improving any non-constraint step increases work in progress rather than delivered output.
- Rework loop cost
- Items returning to an earlier stage consume capacity twice. Measuring the percentage of returning items sizes a hidden drain on throughput.
- Handoff cost
- Each transfer between people or teams adds waiting time and information loss. Counting handoffs is a direct diagnostic for delay in a value stream.
- STATIK definition
- Systems Thinking Approach To Introducing Kanban: a six-step method for designing a kanban system fitted to a specific service's purpose, demand, and capability.
- STATIK step one
- Understand what makes the service fit for purpose by identifying customer expectations and the specific sources of dissatisfaction the service must address.
- STATIK demand analysis
- Analyzing demand sources and patterns, including work types, volumes, variability, and predictability, to shape classes of service and policies.
- STATIK capability analysis
- Assessing the system's capacity, skills, and current performance against demand, so WIP limits and policies are grounded in what the system can actually do.
- STATIK workflow modeling
- Discovering the actual steps and handoffs work passes through, then visualizing them on a board that reflects reality rather than an intended process.
- STATIK kanban design
- Setting WIP limits, cadences, and feedback meetings, then running the system and iterating. STATIK is a design loop, not a one-time event.
- Pull vs push
- Pull starts new work only when capacity frees under a WIP limit; push assigns work regardless of capacity, creating queues and longer cycle times.
- WIP limit commitment
- When a column reaches its limit, no new item starts and the team swarms on finishing. The limit is a binding commitment, not a suggestion.
- Blocked work visibility
- Blocked items are marked with reason and owner and remain counted in WIP, because blocked work still consumes capacity and must be measured.
- Intention in pull
- The pull decision uses class of service and item age, not just worker availability, so priority and risk drive selection within a class.
- Board as policy surface
- A board that displays classes of service, definitions of ready and done, and per-column WIP limits makes the governing policies visible and auditable to everyone.