ITIL 4 CDS common mistakes
ITIL 4 CDS · Understand the traps and how to avoid them.
When a scenario says a technician restored service by restarting a component or applying a temporary workaround, the candidate labels the activity 'problem management' because the fix looked technical; likewise they label a request for a new laptop an 'incident' because something was needed.
In most real jobs one person does both restore and root-cause work, so candidates sort by who did it or how technical it felt rather than by purpose. The exam separates the purpose of restoring service (incident management), investigating and eliminating cause (problem management), and fulfilling a normal predefined request (service request management).
What to do instead
Before reading the options, tag every action in the stem with one of three purposes: restore service now, find/verify cause, or fulfil a standard request. Only then look at the options and reject any whose purpose tag differs from the action the question is about.
The candidate uses 'change enablement', 'deployment management' and 'release management' interchangeably - for example choosing release management for a question about moving a tested component into a live environment, or choosing deployment management for a question about who authorises a high-risk change.
In tooling and everyday speech the three are used as one word ('we released a change'), so the distinctions are never practised. The exam tests three different purposes: enabling beneficial changes while managing risk (change enablement, including authorisation), moving components between environments (deployment), and making functionality available to users (release).
What to do instead
Memorise three one-line purposes and a verb for each: change enablement = assess/authorise/schedule; deployment = move/ship a component to an environment; release = make it available to users. On any stem, underline the verb and pick the practice that owns that verb.
The candidate reads 'a central team sets standards and tooling while business units own and run their own delivery' and selects 'decentralized', or reads 'a single team makes all IT decisions for the whole enterprise' and selects 'federated'.
The four structures are learned as labels rather than as answers to two separate questions (who decides? who delivers?), so near-miss options feel plausible. Candidates also assume 'federated' sounds more modern and therefore more likely to be correct.
What to do instead
Draw a two-column grid on scratch paper - 'who sets direction/standards' and 'who executes day to day' - and place the scenario in it before reading options. Central coordination with dispersed execution is federated; fully dispersed decision-making is decentralized; one authority for both is centralized; a deliberate blend of the three is hybrid.
The candidate treats 'shift left' as meaning move work to cheaper or offshore staff, or picks an option that keeps testing and verification as a late gate while calling it shift left.
'Left' has no intuitive meaning without the timeline picture, and the phrase is widely misused in industry as a cost-cutting slogan. The exam uses it to mean moving work earlier in the value stream and closer to the source of demand - developer testing, automated checks, first-contact resolution, user self-service.
What to do instead
Redraw shift left as a horizontal timeline with detect-fix-feedback points and ask 'which option moves the activity earlier or closer to the source?'. Reject any option whose main effect is later detection, extra handoffs, or purely lower unit cost.
Asked where to improve a described end-to-end flow, the candidate picks the step with the most visible work or the most skilled team, and builds their value stream map from activity boxes only, with no wait time, queue or handoff shown.
Process-mapping habits dominate: candidates draw what people do, not how long work waits. Touch time is what is memorable in a scenario, while waiting is invisible unless quantified, so the largest lever looks like the busiest step.
What to do instead
For every step in the stem, write two numbers - work time and wait time - then identify the largest queue. Pick the option that relieves that queue, not the one that speeds up the busiest worker. If an option cannot be tied to a measured wait in the stem, eliminate it.
Faced with a bottleneck and long lead times, the candidate selects 'add more staff', 'run steps in parallel', or 'increase the batch size to gain efficiency' rather than limiting work in progress and reducing batch size.
Utilisation thinking is deeply trained into candidates: idle capacity feels like waste, and adding resource is the intuitive fix. ITIL flow content instead treats excess WIP and large batches as the cause of queues and delayed feedback.
What to do instead
Use a fixed checklist for flow questions: limit WIP (stop starting, start finishing), reduce batch size, remove the constraint, make feedback loops shorter. Choose the option that matches one of those four; treat 'add people' or 'raise the batch for efficiency' as default wrong unless the stem explicitly shows a capacity ceiling.
The candidate defines an empowered or self-organising team as one with no manager involvement and no constraints, choosing options such as 'the team sets its own objectives and has no reporting line' or 'leadership steps back entirely'.
Autonomy is learned as an attitude word rather than a boundary condition, and candidates over-correct against command-and-control. Real high-autonomy teams have a very clear 'what and why' with discretion over 'how'.
What to do instead
Apply the split test: objectives, priorities and guardrails are set for the team; the method, sequencing and technical decisions are set by the team. Reject any option that removes the 'what/why' boundary or removes leadership entirely, and prefer options that describe guidance, coaching and clear constraints.
Given a scenario describing blame, fear of raising issues and poor cross-team trust, the candidate picks the tightening option - more audits, stricter KPIs, additional approval gates - instead of an intervention aimed at behaviour and safety.
Candidates default to control because it is measurable and looks like action, and because the cultural signal in the stem ('people hide mistakes') is read as a compliance gap rather than a culture symptom.
What to do instead
Classify the stem's symptom first: if it is 'people do not report / do not speak up / hide failures', the answer must change behaviour and safety (blameless review, visible leadership modelling, feedback loops); if it is 'the same error keeps reaching production', the answer may be technical control. Match the intervention type to the symptom type before reading options.
Asked how to develop a capability, the candidate selects a one-off training course, a certification programme, or a tool purchase as the complete answer, and ignores capacity, information, partners and the process side of the capability.
Capability is treated as a synonym for skill, and training is the most familiar lever. The exam expects capability to be built across the whole system - competent people with available capacity, defined ways of working, usable information/knowledge, supporting technology and the right partners - and sustained, not one-off.
What to do instead
Before choosing, run the stem through four questions: people (skills and capacity), process (defined and practised ways of working), technology/information (tools and usable knowledge), partners (supplier and interface capability). Reject single-lever options such as 'send the team on a course' as complete answers; prefer the option that addresses the gap the stem actually shows and that includes reinforcement over time.
On questions about improving a value stream, the candidate optimises one practice in isolation - 'improve incident management', 'improve testing', 'improve supplier management' - without linking it to the end-to-end flow or to what the customer experiences.
The exam is organised by practice, so candidates revise practice by practice and answer practice by practice, missing that CDS questions usually ask about the flow across practices, teams and suppliers rather than inside one of them.
What to do instead
For every 'improve' question, first state the value stream stage in the stem and the outcome the customer receives at the end of it, then ask which option shortens or de-risks that end-to-end path. Eliminate options whose benefit stays inside a single team or practice boundary with no effect on the flow or the customer outcome.