Learner worksheet
Each item gives you a fragment of a case (a brief, a pair of documents, or a small set of figures), and asks what follows. Choose the most defensible response, then read the explanation. Nothing here is timed, and the arithmetic never needs more than a pencil.
Choose one answer for each item. On a judgement item, more than one option is often defensible. Pick the one you would rate most effective, then check the worked explanation for why the runner-up is weaker.
Reading the brief and the evidence
Warm up on the first ten minutes of a case: fixing the decision, and deciding what each document is evidence of.
1. case-analysis-applied-s01-q01
A case pack contains a market report, an internal sales sheet, and a customer survey. The brief asks whether the business should extend its opening hours. Where do you start?
- The longest document, since it will contain the most information
- The most recent document, since it reflects current conditions
- Re-read the question to fix the decision and the criteria it should be judged on, then read each document for what bears on those
- All three in the order they were provided
Answer:
2. case-analysis-applied-s01-q02
One document in the pack is a market-share analysis dated several years ago. The case is set in the present day.
- Use it as evidence about the period it covers, and state explicitly any assumption you make in carrying it forward
- Discard it; out-of-date evidence has no place in a current recommendation
- Use it as though it described the current position
- Average it with a more recent figure to smooth the difference
Answer:
3. case-analysis-applied-s01-q03
The pack includes a quotation from the chief executive: 'Our customers care about price above everything else.' A customer survey is also in the pack.
- Accept it; nobody knows the business better
- Reject it as opinion rather than evidence
- Carry it into the recommendation as a fixed constraint
- Treat it as a hypothesis and test it against what the survey in the pack actually says
Answer:
4. case-analysis-applied-s01-q04
Having read the pack, you conclude that two of the six documents have no bearing on the question asked.
- Find a use for something in each document so no evidence is wasted
- Note briefly that they do not bear on the decision, and set them aside
- Leave them out silently
- Include their contents for completeness
Answer:
5. case-analysis-applied-s01-q05
The same quantity appears in two documents with two clearly different values.
- Check the definition and the period behind each figure before concluding either is wrong; they may be measuring different things
- Use the larger figure, since understatement is the more common error
- Use the figure from the more official-looking document
- Take the average of the two
Answer:
Reconciling conflicting evidence
Two sources disagree. The useful move is almost never to pick one silently.
6. case-analysis-applied-s02-q01
A board memo states the plant runs at 92% utilisation. The operations appendix reports 78% for the same year.
- Use the memo figure; it is the more senior source
- Look for the denominator each is using, scheduled hours against calendar hours would explain a gap of this size, and quote the figure with its basis
- Report both figures without comment and let the reader choose
- Use the lower figure, to be conservative
Answer:
7. case-analysis-applied-s02-q02
A site manager's interview describes a problem that the financial data in the pack does not show.
- Rely on the financial data; numbers are less prone to bias than interviews
- Rely on the interview; the person on the ground sees what reports miss
- Set both aside as unreliable
- Treat the contradiction as a finding in its own right: state both, say which you rely on and why, and say what would resolve it
Answer:
8. case-analysis-applied-s02-q03
Two forecasts in the pack reach different conclusions. The difference traces almost entirely to one assumption about demand growth.
- Present both forecasts and let the reader weigh them
- Adopt the more cautious of the two
- Name the assumption, show how the conclusion moves as it varies, and say what evidence would settle it
- Build a third forecast between them
Answer:
9. case-analysis-applied-s02-q04
One document in the pack is plainly written to argue for a particular option. It is a proposal from the division that would benefit.
- Use its facts where they are checkable, discount its conclusions, and note the author's interest
- Discard it, since it is advocacy rather than analysis
- Treat it as a neutral source like the others
- Give it the same weight as the independent documents
Answer:
10. case-analysis-applied-s02-q05
A figure you need to complete the analysis is not in the pack anywhere.
- Estimate it plausibly and carry on
- State the gap, show the range of values over which your conclusion does not change, and say what would be needed to narrow it
- Decline to make a recommendation without it
- Use the midpoint of the plausible range without drawing attention to it
Answer:
Numbers inside a case
Short calculations of the kind a case turns on. Each one has a distractor that is arithmetically true and decision-irrelevant.
11. case-analysis-applied-s03-q01
A product line generates revenue of 800,000, has direct costs of 600,000, and carries an allocated head-office overhead of 250,000: a reported loss of 50,000. Should it be closed?
- Yes: it loses money as reported
- Yes, unless growth is expected to close the gap
- No: revenue exceeds direct costs, so a line like this should never be closed
- Not on this evidence: it contributes 200,000 towards overhead that would largely remain, so closing it worsens total profit unless that overhead is genuinely avoidable
Answer:
12. case-analysis-applied-s03-q02
Option A costs 100,000 and saves 40,000 a year. Option B costs 250,000 and saves 90,000 a year. Both run for five years and the budget allows either. Which is the stronger investment?
- A, because it pays back in 2.5 years against B's 2.8
- A, because it costs less and so ties up less capital
- B, because over the five years it returns 200,000 net against A's 100,000, and both recover their cost well inside the period
- Neither, without knowing the cost of capital
Answer:
13. case-analysis-applied-s03-q03
The pack includes a survey of 40 customers, conducted at one store on one Saturday afternoon. The chain has 120 stores.
- Treat it as indicative of that store on that day, and say so. It cannot support a chain-wide conclusion
- Treat it as representative; 40 responses is a workable sample size
- Discard it entirely as worthless
- Scale the findings up by the number of stores
Answer:
14. case-analysis-applied-s03-q04
A product sold at 100 with a unit cost of 75 had its price cut by 10%. Volume then rose by 12% and unit costs were unchanged. What happened?
- The cut worked: revenue rose
- The cut lost money: unit contribution fell from 25 to 15, so 12% more volume left total contribution well below where it started
- The cut worked if it also gained market share
- It cannot be judged without the competitors' response
Answer:
15. case-analysis-applied-s03-q05
This year's operating profit includes a one-off gain of 300,000 from the sale of a building.
- Include it; the cash is real and it was earned this year
- Include it, with a note that it was unusual
- Exclude it from the underlying comparison and disclose it separately, because it will not recur
- Restate the prior years to include an equivalent adjustment
Answer:
Framing the recommendation
Finish on the output. A case answer is judged on whether somebody could act on it, not on how much of the pack it covers.
16. case-analysis-applied-s04-q01
How should a written recommendation be structured?
- The recommendation first, then the two or three reasons that carry it, then the main risk and what would change your mind
- The analysis in the order you carried it out, ending with the conclusion
- Every option given equal weight, leaving the choice to the reader
- The risks first, so expectations are set before the proposal
Answer:
17. case-analysis-applied-s04-q02
After the analysis you favour option A over option B, but only narrowly, perhaps sixty-forty.
- Recommend neither, and present both fairly
- Recommend A, say plainly that the decision is close, and name the one piece of evidence that would reverse it
- Recommend the lower-risk option regardless of the analysis
- Ask for more time before recommending
Answer:
18. case-analysis-applied-s04-q03
Your recommendation rests on an assumption that is reasonable but could turn out to be wrong.
- Leave it unstated so the argument reads more strongly
- Record it in an appendix for completeness
- Present the recommendation without conditions and handle it if it arises
- State the assumption where it is used, show how sensitive the conclusion is to it, and give the fallback if it fails
Answer:
19. case-analysis-applied-s04-q04
A colleague's case answer recommends a course of action but says nothing about cost, sequence, or who would do it.
- That is fine; implementation is a separate exercise
- That is fine as long as the strategic direction is right
- It is incomplete: a recommendation that cannot be costed or scheduled cannot actually be decided on
- That is normal for a strategy-level case
Answer:
20. case-analysis-applied-s04-q05
Having worked through the pack, you conclude that the evidence does not support any of the three options the brief offered.
- Recommend the least bad of the three
- Recommend the one the client appears to favour
- Report that the evidence is inconclusive
- Say plainly that none of the stated options is supported, say what the evidence does support, and define the specific work that would settle it
Answer:
Worked explanations
case-analysis-applied-s01-q01
Knowing what decision you are making and on what criteria is what makes a document relevant or not, and it stops you absorbing pages that cannot change the answer. Starting with the most recent document is the sensible-sounding heuristic and does nothing for you here: recency says when a document was written, not whether it is about the question you were asked.
case-analysis-applied-s01-q02
Old evidence is still evidence about its own period, and it is often the only data available on that variable, what matters is that the bridge to today is visible and challengeable. Discarding it is the rigorous-looking move and throws away information; the discipline is labelling the assumption, not deleting the source.
case-analysis-applied-s01-q03
A senior person's belief is a claim you have been handed alongside data that can check it, and the check is the whole point of putting both in the pack. Accepting it on authority is the tempting deference, and it converts the most testable statement in the case into an assumption nobody examines.
case-analysis-applied-s01-q04
Saying what you excluded and why shows the reader that you read it and judged it, which is exactly what distinguishes selection from oversight. Using something from every document is the instinct 'use all the evidence' produces, and case packs routinely contain material chosen to be irrelevant, forcing a use for it dilutes the argument that does the work.
case-analysis-applied-s01-q05
Most apparent contradictions in a case pack are definitional (headcount versus full-time equivalent, calendar year versus financial year), and resolving that gives you a figure you can defend. Averaging is the tidy-looking compromise and produces a number that neither document supports and that means nothing at all.
case-analysis-applied-s02-q01
Utilisation has several standard denominators, and identifying which one each source used usually reconciles the two exactly. Taking the lower figure to be safe is the prudent-sounding option and is still a guess: a conservative choice between two undefined numbers is an undefined number that happens to be smaller.
case-analysis-applied-s02-q02
A gap between what people report and what the accounts show is often the most valuable thing in the pack, and naming it is more useful than quietly resolving it. Trusting the numbers is the reflex that feels most objective, and aggregated data can easily average away a local effect that is real and expensive at one site.
case-analysis-applied-s02-q03
Isolating the single driver turns two competing outputs into one question the decision-maker can actually answer or investigate. Presenting both is the balanced-looking option and leaves the reader exactly where you started, holding two numbers with no account of why they differ.
case-analysis-applied-s02-q04
An interested party usually holds the best operational data about their own area, so the answer is to separate their facts from their argument. Discarding it is the sceptical-looking move and throws away the pack's most detailed source over a bias you could simply have stated.
case-analysis-applied-s02-q05
Showing that the recommendation holds across a range converts a missing number from a blocker into a bounded uncertainty. Quietly using the midpoint is the option that keeps the analysis looking complete, and it is the most damaging: to a reader, an unlabelled assumption is indistinguishable from data.
case-analysis-applied-s03-q01
Allocated overhead does not leave with the product, so the relevant figure is the 200,000 contribution and the question is how much of the allocation is actually avoidable. The 'never close it' answer is the strongest wrong one because it has the right instinct and states it far too absolutely: capacity, avoidable costs, and what else the space could earn all bear on the decision.
case-analysis-applied-s03-q02
A returns 200,000 against 100,000 spent; B returns 450,000 against 250,000 spent: twice the net gain over the same horizon. Payback is the strongest distractor because it is a real criterion and A genuinely wins on it, but payback stops counting on the day the money is recovered and so ignores more than half the life of both options.
case-analysis-applied-s03-q03
The limitation is not the sample size but the sampling: one store on one weekend afternoon selects for a particular kind of customer. 'Forty is a workable sample' is the tempting statistical answer and addresses precision when the problem is bias. A larger sample drawn the same way would be just as unrepresentative, only more confidently so.
case-analysis-applied-s03-q04
Contribution per unit fell by 40% while volume rose by 12%, so on 100 units the total falls from 2,500 to about 1,680. 'Revenue rose' is the sharpest distractor because it is arithmetically true (revenue goes from 10,000 to about 10,080), and revenue is not what pays for anything.
case-analysis-applied-s03-q05
A recommendation about future trading has to rest on the part of this year's profit that will happen again. Including it with a note is the transparent-looking compromise and still leaves a year-on-year comparison in which the growth is mostly a building sale.
case-analysis-applied-s04-q01
Answer first means every subsequent paragraph has a job the reader can see, and it survives being read only halfway. Narrating the analysis in the order you carried it out is the natural-feeling structure, because it is how the work actually went, and it makes the reader carry all the evidence before learning what any of it was for.
case-analysis-applied-s04-q02
Recommending while being honest about the margin gives the reader both your judgement and its strength, and the flip condition tells them what to watch. Presenting both without a recommendation is the scrupulously balanced option, and it hands the decision back to the person who has read less of the evidence than you have.
case-analysis-applied-s04-q03
An assumption stated at the point it does its work can be challenged by someone who knows better, which is the only way it gets corrected before the decision. Putting it in an appendix is the technically-complete option and functions as concealment: nobody reads the appendix before deciding.
case-analysis-applied-s04-q04
Choosing between options means comparing what each costs and how long it takes, so an answer without either is a preference rather than a recommendation. 'Implementation is separate' is the reasonable-sounding scoping argument, and it withholds precisely the information the choice turns on, even an order-of-magnitude figure changes which option wins.
case-analysis-applied-s04-q05
Rejecting the menu is legitimate when you replace it with something: a different option, or a defined piece of work with a cost and a date. Stopping at 'inconclusive' is the honest half of the answer and leaves the decision-maker exactly as stuck as before, which is why it reads as an evasion rather than a finding.