ITIL 4 DPI cheat sheet

Direction answers 'what and why' (goals, boundaries, authority); planning answers 'how and when' (resources, sequence, responsibilities); improvement answers 'what change makes it better'.

Direction is set at every organizational level; the higher the level, the broader the scope and the longer the planning horizon.

Vision = desired future state; mission = purpose and who it serves; goals = broad aims; objectives = specific, measurable targets that support goals.

Output = a deliverable produced; outcome = the result or effect of that deliverable for stakeholders; value is perceived benefit, usefulness or importance.

All improvement is change, but not all change is improvement: an improvement is a deliberate, planned change with an expected positive result against objectives.

Direction cascades: governance sets boundaries, strategy sets direction, tactics align products and value streams, operations execute; each level constrains the level below.

DPI scope spans all four dimensions (organizations and people; information and technology; partners and suppliers; value streams and processes), not just process or technology.

Decision: if a decision affects only one value stream and stays inside policy, make it at that level; if it crosses boundaries or changes policy, escalate.

Three planning levels: strategic (long term, whole organization), tactical (medium term, products and value streams), operational (short term, day-to-day delivery).

Direction and planning apply to all products and services, all value streams, all practices, all four dimensions, and relevant partners and suppliers.

Scope covers both 'run' (ongoing service delivery) and 'change' (projects, improvements, organizational change).

Scope boundaries are set by the governing body through policy, delegation of authority and risk appetite; plans must remain inside those boundaries.

Typical planning outputs: strategy, policies, portfolios, budgets, roadmaps, project and improvement plans, and service and value stream plans.

The wider the scope, the greater the uncertainty, so plans at strategic level state outcomes and constraints, while operational plans state tasks and schedules.

Anything outside the agreed scope of a plan is a change to scope and needs the authority level that owns that plan.

Seven guiding principles: focus on value; start where you are; progress iteratively with feedback; collaborate and promote visibility; think and work holistically; keep it simple and practical; optimize and automate.

Principles are used as decision rules when no policy or procedure covers the situation, and to test whether a plan is sound.

Policy vs plan: policy is a rule or boundary that directs behaviour; a plan describes how, when and with what resources objectives will be met.

Business case content: benefits, costs, risks, timescales, plus comparison of options; a plan proceeds only when the case is justified.

Cascading objectives: organizational goals decompose into supporting objectives for value streams and teams, so contribution is traceable upward.

Decision-making choices: centralized (consistency, control) vs decentralized (speed, local fit); delegation of authority records who may decide what.

Prioritization must use agreed, weighted criteria aligned to objectives; undocumented or intuition-only prioritization is a design flaw.

Iterative planning: plan a step, take it, check feedback, re-plan; avoid producing large plans that are obsolete before execution begins.

Governance = the means by which an organization is directed and controlled; it is a component of the service value system and connects to all value chain activities.

Governing body activities are evaluate, direct and monitor (EDM): evaluate options and performance, direct via strategy and policy, monitor conformance and outcomes.

Governance is accountable for the achievement of objectives; management is responsible for planning and executing the work.

Governance instruments: policies, controls, delegated authority limits, escalation routes, reporting requirements and audit.

Compliance = adherence to laws, regulations, contracts, standards and internal policy; obligations must be identified, met and evidenced.

GRC is most effective when governance, risk and compliance share one set of objectives and one reporting flow rather than running as separate silos.

The governing body sets risk appetite and tolerance; plans and value streams must be designed to operate within them.

Governance adds value by enabling decisions and control, not only by restricting; over-control slows flow and is itself a risk.

Continual improvement model steps in order: what is the vision; where are we now; where do we want to be; how do we get there; take action; did we get there; how do we keep the momentum going.

Steps 1-4 and 7 are planning, baselining and embedding activities; step 5 is execution; step 6 is measurement against the defined target state.

The model is iterative and applies at strategic, tactical and operational levels; the same steps repeat as each improvement completes.

Continual improvement register (CIR) records, prioritizes, tracks and reports improvement opportunities and initiatives.

Improvement opportunities should be prioritized by value potential, effort, risk and alignment to objectives; keep the backlog visible.

Improvement methods are combined, not exclusive: Lean for waste and flow, Agile and DevOps for iterative delivery, Kanban for work visibility, project methods for large change.

Measure the before-state and the after-state using the same metrics, otherwise the improvement cannot be demonstrated.

Embedding is required: without reinforcement and updated procedures, training and measures, gains decay back to the old way of working.

Communication is two-way: the sender is responsible for confirming that the intended meaning was understood, which requires feedback and listening.

Communication plan attributes: who (audience), what (message), why, when, where/channel and how (format and tone).

Different stakeholders need different messages at different times; one generic message for all audiences is a design error.

Stakeholder analysis maps interest and influence, plus current attitude, to decide engagement effort and communication approach.

Organizational change management addresses the people side of change: sponsorship, stakeholder engagement, communication, training, resistance handling and reinforcement.

A change agent network (CAN) is a group of trained advocates distributed across the organization who support adoption and channel local feedback.

Resistance is information: it usually signals unmet needs, unclear benefit, loss of something valued or lack of capability, and should be diagnosed rather than suppressed.

Communication must precede, accompany and follow the change; stopping communication at go-live is a common cause of failed adoption.

Hierarchy: objectives → critical success factors (what must go well) → key performance indicators (how success is measured) → metrics and data.

Lagging indicators report results already produced; leading indicators predict future results and enable earlier intervention.

Goal-question-metric: state the goal, ask the questions that reveal whether it is met, then define metrics that answer those questions.

SMART objectives (specific, measurable, achievable, relevant, time-bound) make the associated measures meaningful.

Balance measures across dimensions and stakeholders to prevent sub-optimization, where one area improves at another's expense.

A good metric is relevant to an objective, defined precisely, cost-effective to collect, and owned by someone empowered to act on it.

Reports are decision-oriented: state the question answered, the trend, the variance from target and the recommended action, not just raw data.

Measurement itself consumes effort; before adding a metric, confirm the decision it will change.

Value stream = the series of steps and activities an organization performs to create and deliver products and services to consumers, starting with demand and ending with value delivery.

Value stream mapping is a visual method showing steps, information and material flows, handoffs, queues and delays to expose waste and constraint points.

Typical sequence: identify the value stream and its boundaries, map the current state, analyse waste and delays, design the future state, then implement and re-map.

Process time is the time work is actively transformed (value-adding); lead time is the total elapsed time from demand to delivery; waiting is the difference.

Flow efficiency = process time ÷ lead time × 100; a low percentage shows that waiting and handoffs dominate the end-to-end time.

Percent complete and accurate (%C&A) measures the quality of output passed between steps and reveals rework loops.

Common wastes to look for include waiting, over-processing, handoffs, rework, unnecessary approvals and duplicated information entry.

Mapping a value stream is a prerequisite for improving it: you cannot optimize flow across steps you have not made visible.

Risk = a possible event that could cause harm or loss, or make it more difficult to achieve objectives; risk includes potential positive outcomes (opportunity).

Risk assessment = identification, then analysis (probability and impact), then evaluation against risk appetite to decide whether treatment is needed.

Risk treatment options: avoid, reduce/mitigate, transfer, accept; for opportunities the parallel options include exploit, enhance, share and accept.

Risk appetite = the amount of risk the organization is willing to accept in pursuit of objectives; risk tolerance = the acceptable variation around that level.

Risk register entries record the risk description, cause, effect, owner, current rating, treatment actions, due dates and residual rating.

Inherent risk is the level before controls; residual risk is the level after controls are applied, and it is residual risk that is accepted.

Each plan level evaluates risk at its own horizon: strategic plans weigh uncertainty of direction, operational plans weigh delivery and continuity risk.

Every risk has one accountable owner and named actions with dates; unowned risks are unmanaged risks.

Risk management is a general management practice integrated with governance, so risk reporting follows the same escalation path as performance reporting.