KSD key terms

A means of delivering value to customers by facilitating outcomes they want to achieve, without the customer owning specific costs and risks.

The outcome or result a customer is trying to achieve; the reason the service exists and the anchor for judging whether work adds value.

A measurable change in the customer's situation that the service enables, as opposed to an activity or deliverable produced internally.

Requests or work that directly serve the customer purpose and, if fulfilled, would create value for the customer or the organization.

Demand caused by a failure to do something right the first time, such as repeat contacts, escalations, rework, or complaints.

Examining incoming work by type, source, and frequency to understand what customers ask for and how much demand is avoidable failure demand.

A documented description of a service's purpose, customers, offerings, and commitments, used to set expectations and scope what the team supports.

The sequence of steps and handoffs work passes through from a request being accepted to the customer receiving value.

An explicit or implicit rule that governs how work is handled, such as approval gates, batch sizes, or prioritization rules, and shapes the system's behavior.

A policy that is written down, visible, and understood by everyone affected, so it can be questioned and changed deliberately rather than followed by habit.

Predefined categories of work with agreed policies for how each is prioritized, pulled, and treated, such as expedited, fixed date, standard, and intangible.

A class of service for urgent work that must jump the queue, typically limited to one or two active items to protect the rest of the system.

A class of service for work with an externally imposed deadline, managed by starting late enough to avoid excess waiting while still meeting the date.

The default class of service for routine work, pulled in priority order and subject to normal work-in-progress limits.

A class of service for work whose value is not immediately visible, such as improvement, technical debt reduction, or capability building, still protected with capacity.

The deliberate split of available team capacity across classes of service or work types so that important but non-urgent work is not permanently starved.

A recurring mechanism that returns information about results to the people doing the work so they can adjust behavior and decisions.

An observable indicator in the workflow, such as growing queues, aging items, or blocked work, that suggests where flow is degrading and attention is needed.

A small, time-boxed change with a stated hypothesis and expected measure, run to learn whether a modification improves the system before scaling it.

A testable statement predicting that a specific change will produce a specific measurable effect, written before the experiment so results can confirm or refute it.

The regular rhythm at which the team meets, reviews flow, and adjusts plans, providing predictable opportunities for feedback and coordination.

Approaching an organization as a set of interconnected parts whose behavior emerges from their interactions, rather than optimizing each part in isolation.

Improving one part of the system in a way that may worsen overall performance, often by creating queues or starving downstream steps.

The limiting factor in a system that determines overall throughput; improving anything other than the constraint does not increase the system's output.

A management approach that identifies the system's limiting factor, exploits it fully, subordinates other work to it, elevates it, and then repeats the cycle.

A step or resource whose capacity is lower than the demand placed on it, causing work to accumulate in front of it and limiting flow downstream.

A recurring pattern of behavior in complex systems, such as shifting the burden or fixes that fail, that helps explain unintended consequences of well-intentioned actions.

Systems Thinking Approach to Introducing Kanban: a structured method to model a system's purpose, demand, workflow, and policies before designing a kanban system.

A category describing the kind of work flowing through the system, such as feature, defect, or support request, used to analyze demand and design appropriate policies.

A visual representation of the steps, handoffs, and waiting times work passes through, used to identify delays and non-value-adding activity.

The proportion of total lead time spent actively working on an item rather than waiting, expressed as a percentage and used to expose hidden delay.

A visual display of workflow columns and work items that makes current work, its state, and its progress visible to everyone involved.

A cap on how many items may be in a given workflow state at once, used to reduce multitasking, expose bottlenecks, and encourage finishing before starting.

A way of working where downstream capacity signals readiness and work is drawn into a step only when there is room, rather than pushed ahead of demand.

A line of work items waiting for the next step; queues hide delay, grow silently, and are a primary source of poor flow and customer dissatisfaction.

A relationship stating that average lead time equals average work in progress divided by average throughput, useful for predicting the effect of limiting WIP.

The elapsed time from when a customer's request is accepted or committed to when the work is delivered, including all waiting.

The elapsed time from when work actively starts on an item to when it is finished, excluding time spent waiting before work began.

The number of work items completed per unit of time, used with work in progress to understand and forecast delivery capability.

The elapsed time since a work item entered its current state, used to spot items that are stalled or at risk before they breach expectations.

A stated probabilistic forecast of how long work of a given type will take, such as a percentage of items completed within a number of days.

A condition preventing a work item from progressing, which should be made visible immediately and treated as a signal to act rather than a private problem.

A regular examination of measures such as lead time, throughput, and work in progress to detect trends and decide the next improvement step.