CGMA common mistakes
CGMA · Understand the traps and how to avoid them.
Producing a full set of ratio or variance calculations and then stopping, without translating the numbers into a stated financial consequence and a recommended action.
Candidates treat the integrated case study as a technical arithmetic test and assume the numbers 'speak for themselves'; marks are awarded for interpretation and application, not for the calculation alone.
What to do instead
After every calculation block, force one written sentence in the form 'This means [impact on profit/cash/risk], therefore I recommend [action]', and check that each numeric working has at least one such sentence beside it.
Discounting nominal cash flows at a real discount rate, or applying a pre-tax hurdle rate to post-tax cash flows, when appraising an investment option.
Under time pressure candidates grab the first rate shown in the scenario without checking whether tax, inflation and financing assumptions are on the same basis as the cash flows.
What to do instead
Before any NPV or appraisal working, write a one-line basis check: 'Flows are nominal/post-tax, so the rate must be nominal/post-tax WACC' - and adjust the rate or the flows explicitly if the bases differ.
Making recommendations that carry no quantified financial impact - for example 'improve cost control' or 'invest in marketing' with no figure attached.
Candidates default to general management commentary because it feels safe, and they fear that a quantified estimate could be 'wrong'; the exam rewards justified estimates over vague assertions.
What to do instead
Impose a rule of at least two quantified impacts per major recommendation (a currency amount, a percentage, or a payback period), each with the assumption stated in one clause so the marker can follow the logic.
Answering an operational requirement with generic textbook theory about process improvement, rather than analysing the specific constraint, bottleneck or quality problem described in the scenario.
Memorised frameworks are easier to retrieve under exam pressure than a fresh analysis of the scenario, so candidates reproduce model content and never apply it to the named process.
What to do instead
Anchor every operational point to a named entity, volume or step from the scenario, and delete any sentence that could have been written without reading the case.
Recommending an operational change (new shift pattern, outsourcing, quality initiative) with no implementation plan and no treatment of the capacity or service-level trade-off it creates.
Candidates stop at the 'what should we do' question and treat 'how would it actually work' as a separate, optional requirement; operational marks hinge on feasibility.
What to do instead
Attach a standing four-line implementation paragraph to every operational recommendation: what changes, who owns it, over what timeframe, and the single measure that will show it is working.
Recommending a strategic option without explicitly evaluating the alternatives or testing it against the organisation's stated objectives and competitive position.
Candidates commit to the first plausible option and then justify it, rather than using the case facts to compare options - evaluation verbs such as 'evaluate' and 'assess' require a comparison to be visible.
What to do instead
Use a short decision matrix: two or three criteria drawn from the organisation's own objectives, each option scored against them, and one sentence naming the option rejected and why.
Presenting a strategic model such as SWOT, PESTEL or Porter's five forces as a descriptive list of factors with no implication or action attached to any item.
The model is used as a way to generate content and fill answer space, when the marker is looking for the 'so what' - the strategic consequence of each factor.
What to do instead
Convert every bullet into a two-part sentence: 'Factor - therefore implication - therefore action', and prune any factor that cannot be taken through to an action.
Recommending a cost reduction or restructuring without addressing the people, ethical, regulatory or reputational consequences of how it is carried out.
Candidates focus narrowly on the financial case and treat stakeholder and governance issues as separate topics, missing that the management domain tests the ability to weigh competing interests.
What to do instead
Run a fixed three-line check before finalising any significant recommendation: who is affected, what could go wrong ethically or reputationally, and what control or mitigation is needed.
Spending disproportionate time on the first requirement and leaving the final, often most integrative, requirement answered in two or three rushed lines.
The first requirement usually contains the most familiar calculations, so candidates over-invest there and lose track of the marks-per-minute budget for the remaining sections.
What to do instead
Before starting, convert the mark allocation into minutes per requirement, write the end time for each against the clock, and move on when the time is up even if the working feels unfinished.
Writing the answer as an unstructured list of technical points, with no situation summary and no recommendation stated up front, so the marker has to search for the decision.
Candidates write in the order they thought of things rather than the order a senior decision-maker needs, and the professional-format marks are lost even when the technical content is sound.
What to do instead
Apply a fixed skeleton to every answer: one-line situation, key analysis, recommendation stated first and then justified, implementation and risk - and put the recommendation in bold at the start of that section.