High impact · high effort
Big bets
Worth doing, but plan them. These need real resources, so sequence them and protect the timeline.
This is a free educational guide. It is not affiliated with any university, consulting firm, or named framework owner. Frameworks are credited to their original authors where known. Examples and numbers are illustrative, not real business data.
This guide collects the thinking tools that show up in case competitions, guesstimates and case interviews. The point is not to memorise a flowchart. It is to recognise which tool a problem is asking for, then use it well. Everything here is a general method or a published framework explained in plain words, so you can apply it to hackathons, ideathons, business plan contests and interviews just as easily as to a case round.
The idea behind your answer matters more than the deck around it. Methods and structure exist to make a good idea legible, not to replace one.
A framework is a way to look at a problem, not a form to fill in. Use the part that fits the case and drop the rest. Forcing a framework is worse than using none.
Enough to get a first team started, and enough structure for anyone who already competes and wants to win more consistently.
Eleven chapters, from first principles to the frameworks and numbers you reach for under time pressure. Tap any tile to jump.
The thinking that happens before any slide. Understand the situation, aim at the right question, and find a path to the answer without analysing everything at once.
SCQA is a way to frame a problem so the whole room agrees on what you are actually solving. It comes from Barbara Minto's work on structured communication. Walk the four steps in order and you arrive at a question that is sharp enough to answer. Most weak submissions fail here, not in the analysis: they solve a problem nobody asked about.
The stable backdrop everyone accepts as true. State it plainly and keep it short. Nothing here should be in dispute.
The thing that changed and broke the status quo. This is the tension that makes the case worth solving.
The single question that follows from the complication. If it does not flow naturally from C, the framing is off.
Your recommendation, stated up front. It should resolve every part of the complication, not just one corner of it.
A good case question is specific, measurable and time bound. "Grow revenue" is a wish. "Lift annual revenue by 15% within two years without raising headcount" is a question you can build an answer against.
The answer is the part judges remember, so spend real time generating options before you commit to one. The reliable pattern is two phases that should never run at the same time. First diverge: widen the set of possible ideas without judging them. Then converge: filter, combine and stress test down to the strongest one. Most teams skip the first phase and defend the first idea anyone said out loud.
Ground rules for the diverge phase. These come from classic brainstorming practice (Alex Osborn) and they exist to protect early ideas from being shot down before they are understood.
No "that won't work" while generating. Criticism trains the room to stop offering ideas.
More ideas means more raw material to combine later. Volume first, quality second.
Add to what others said rather than replacing it. The best answer is often a merge of two weak ones.
An impractical idea is easy to tame later. A timid idea has nowhere to go.
A simple way to converge. When it is time to choose, the Disney method (popularised by Robert Dilts from Walt Disney's working style) splits the work into three mindsets that you adopt one at a time. Run them as separate passes, not as a single argument.
Imagine the idea fully realised, with no limits. What would the best version look like?
Turn the dream into a plan. What steps, resources and timeline would actually deliver it?
Attack the plan on purpose. Where does it break, what is missing, what would a judge poke at?
Edward de Bono's Six Thinking Hats give a team six lenses to look at a problem, one at a time. The value is that everyone wears the same hat at once, so a discussion stays on one track instead of mixing optimism, risk and emotion in the same breath. Call the hat out loud: "let us spend five minutes in the black hat" forces the whole team to look for risk together.
What do we actually know? What data is missing? Pure information, no opinion.
Gut reactions and intuition, stated without having to justify them. Useful to surface, then set aside.
Risks, weaknesses and reasons it might fail. The most useful hat, and the easiest to overuse.
Value, upside and why it could work. The deliberate counterweight to the black hat.
New options, alternatives and creative leaps. This is where generation lives.
Manages the discussion itself. Decides which hat is next and summarises the outcome.
Trying to analyse everything is how teams run out of time with nothing decided. Hypothesis-driven problem solving flips the order: start with a testable guess at the answer, then go looking only for the evidence that would prove or disprove it. Each loop ends in one of two ways. The evidence holds and you have a conclusion, or it fails and the failure tells you what the next hypothesis should be. Either way your effort stays pointed at the decision.
Question: where is margin leaking? Hypothesis: a single product line is dragging the blended margin down. Test: pull margin by line and check the spread. If one line is clearly worst, the guess holds and you size the fix. If margin is even across lines, the guess fails, and the next hypothesis moves to a shared cost like logistics or returns.
MECE stands for mutually exclusive, collectively exhaustive. It is a discipline for breaking a big question into branches: the branches must not overlap, and together they must cover the whole problem with no gap. An issue tree built this way lets you split the work, rule out whole branches quickly, and avoid double counting. MECE is general consulting practice rather than any one author's framework. The clean first cut is usually the hard part. Below, profit splits into revenue and cost, which is exhaustive by definition and has no overlap.
A quick MECE check: could any item belong in two branches at once? If yes, the cut is overlapping. Is there a driver the tree cannot reach? If yes, it is not exhaustive.
When ideas are hard to generate and no framework or data points the way, reverse the problem. Instead of asking how to succeed, ask how you would guarantee failure, then design against each of those causes. This is the inversion habit Charlie Munger is known for, and it works because criticism is often easier to produce than invention. Listing every way to wreck the goal tends to surface the real constraints faster than staring at the goal itself.
Once you have a long list of ideas, you have to choose what to actually do. The advice to not boil the ocean only helps if you have a tool to sort with. The impact versus effort matrix plots each idea on two axes and tells you what to do first. Start with the quick wins, plan the big bets, and drop the rest. The point is to make the cut visible so the team agrees on it.
Do these first. High return for little work. They build momentum and free up time for the harder calls.
Worth doing, but plan them. These need real resources, so sequence them and protect the timeline.
Cheap and minor. Do them only if there is slack, never at the expense of a quick win or a big bet.
Avoid. A lot of work for little return. The discipline of the matrix is being willing to cut this quadrant.
A close cousin for managing your own time is the Eisenhower matrix, which swaps the axes for urgent versus important: do the urgent and important now, schedule the important but not urgent, delegate the urgent but not important, and drop the rest.
You found the answer. Now arrange it so the room follows your logic without effort. Structure is what separates a correct answer from a convincing one.
The Pyramid Principle is Barbara Minto's idea that any argument reads best when the single main message sits at the top, a small set of supporting arguments sits below it, and evidence sits at the base. Each level answers the question the level above it raises. There are two directions to remember. You build the pyramid from the bottom up, gathering evidence into arguments into one conclusion. You present it from the top down, leading with the conclusion so the audience knows where you are going.
Cluster your findings into a few groups. Name the takeaway of each group. The takeaways become your arguments, and the one thing they all point to becomes your governing thought.
Reverse it on stage. Open with the governing thought, give the arguments that support it, and only then drill into the evidence for anyone who wants it.
The same pyramid powers SCQA from Chapter 01. Situation and complication set up the question, and the answer to that question is your governing thought.
There are two ways to order the same content under the governing thought, and the choice depends on your audience. Inductive order leads with the recommendation, then gives parallel reasons. It is fast to follow and works when the room already trusts you. Deductive order walks from what is happening, to why, to what to do. It is slower but earns agreement when the room might resist the answer if you stated it first.
Before anyone opens a slide tool, sketch the whole story as a row of one line headlines, one per intended slide. Read those headlines straight down. If they tell a coherent argument on their own, the deck has horizontal logic and you can start building. If they do not, no amount of formatting will save it. Storyboarding on paper or sticky notes also lets the team move whole blocks around and split the work cleanly, because every block is just a headline plus a placeholder.
Each headline should state the takeaway, not the topic. "Margin fell because input costs rose" beats "Margin analysis". A reader who only skims your titles should still get the argument.
If a slide needs two headlines, it is two slides. Every chart, number and line on the slide should support the single headline above it.
Dots and dashes is a quick habit for sketching a storyboard before you write full headlines. Mark each load bearing message with a dot. Mark each piece of supporting detail that hangs off a message with a dash. Reading only the dots top to bottom should give the spine of your argument. The dashes are there to support a dot, never to introduce a new one. If a dash starts to feel like its own message, promote it to a dot and give it its own slide.
• Costs are rising faster than price
– input prices up across two key lines
– price held flat to protect volume
• The cheapest fix is on the cost side
– two cost lines moved, the rest held
• Recommendation: renegotiate inputs, then revisit price
Read only the dots and you still have the whole argument. The dashes never carry a new point on their own.
An observation is not an insight. Judges hear "sales are down ten percent" from every team. What moves you ahead is what that fact forces. Apply the so what test to every finding: keep asking "so what?" until the answer is something the business must do. Each pass climbs one rung, from a fact, to what it means, to the action it forces.
A neutral fact pulled straight from the data. True, but on its own it tells the room nothing about what to do.
The same fact plus its consequence for the business. This is what belongs in an action title and what a judge actually rewards.
A correct answer presented badly still loses. The deck is where your structure becomes something a judge can absorb in real time.
A competition deck usually moves through the same five blocks. Knowing the blocks lets you assign slides before you have a single chart.
The recommendation and its impact, up front, on one slide. Built straight from your SCQA so the room knows the answer before the detail.
The sub problems and their causes. This sets the foundation that your recommendations later slot into.
The main body. Each recommendation gets enough space to be clear and specific. This is what you spend the most time on.
How the recommendation actually gets done. Sequence, owners and a rough timeline turn an idea into a plan.
No plan is risk free. Naming the risks and how you would handle them signals you have thought it through.
The useful rule of thumb is that one slide takes about two minutes to present well. Work backward from your time limit, not forward from your slide count. A rough split for a ten minute slot:
Treat these as a starting point and adjust to the brief. The recommendation should always get the largest share.
Every strong slide carries the same parts. Lead with the takeaway as a title, support it with one main visual, and frame it with light navigation and a source line.
Frameworks are shared lenses, not forms to fill in. Use the parts that move your argument and leave the rest. A half used framework that earns an insight beats a complete one that earns nothing.
This is the most fundamental strategy lens, and the fastest place to start almost any business case. Look at the situation from three angles: the Company, its Customers, and its Competitors. Strategy lives where all three meet, where the company can serve a real customer need in a way rivals cannot match. The 4C version adds Channel, or Collaborators, when how the product reaches the customer is itself part of the question.
Your strengths, costs, capabilities and constraints. What can this business credibly do?
Who buys, what they need, how they segment, and what they will pay for it.
Who else serves that need, how they are positioned, and where they are weak.
The fourth C. How the product reaches the customer, and the partners that make it possible.
SWOT sorts what matters into four boxes: strengths and weaknesses are internal and within the company's control, opportunities and threats are external and are not. It is a fast first scan of a situation. Its weakness is that it can become a list with no decision attached, so always end by asking which strength meets which opportunity, and which weakness exposes you to which threat.
What the company does well and controls. Assets, capabilities and positions rivals would struggle to copy.
Internal gaps and constraints. The places a competitor or a shock could exploit.
Outside trends the company could ride. New demand, channels, technology or regulation moving in your favour.
Outside forces that could hurt you. New entrants, substitutes, shifting demand or rules moving against you.
PESTLE scans the external environment across six forces. It pairs well with the opportunities and threats half of a SWOT, since it is a structured way to find them. Do not narrate all six. Pick the two or three that actually shape this case and go deep on those.
Government, policy direction, trade stance, stability. Anything driven by who holds power.
Growth, inflation, rates, employment, currency. The macro backdrop demand sits inside.
Demographics, values, lifestyle and behaviour shifts in the population you sell to.
New capabilities, automation and shifts that change cost structures or what is possible.
Regulation, compliance, licensing and liability. The rules you must operate within.
Climate, resources, sustainability expectations and physical constraints on the business.
Michael Porter's Five Forces explain why some industries are structurally more profitable than others. The forces act on the rivalry in the centre. The stronger they are, the more they squeeze the profit available to everyone competing. Use it to judge how attractive an industry is to enter or stay in, not to analyse a single company.
The marketing mix began as McCarthy's four Ps and was extended to seven by Booms and Bitner to cover services. It is a checklist for how a product reaches and serves a customer. Use it to find which lever a marketing or go to market case actually turns on.
What you sell and the need it meets.
What you charge and how the structure works.
Where and how customers buy it.
How you make the market aware and persuade.
The staff and service that shape the experience.
The steps from order to delivery and support.
The tangible cues that signal quality, online or in person.
The 7S model checks whether the parts of an organisation are aligned. Three are hard elements that are easier to change: strategy, structure and systems. Four are soft elements that are slower to shift: shared values, skills, style and staff. Shared values sit in the centre because everything else should line up around them. Use it for organisation, restructuring and change cases, where a good strategy fails because the soft elements were never aligned to it.
The growth share matrix from the Boston Consulting Group sorts a portfolio of products or business units on two axes: how fast their market is growing, and how large their share of it is. The four quadrants suggest where to invest, where to hold, and where to exit. It is a portfolio tool, useful when a case spans several lines and you need to decide where the money should go.
Leaders in a fast market. Fund them to defend the lead. They turn into cash cows as growth slows.
Fast market, weak position. Either invest hard to win share or step away. Do not drift.
Strong position in a slow market. Harvest the cash and use it to feed stars and question marks.
Weak position in a slow market. Usually a candidate to wind down or sell unless it serves a strategic role.
The hard part is not knowing the frameworks, it is reaching for the right one fast. Read the prompt, match it to a starting point, then adapt. These are defaults to open with, not a script. Most real cases combine two or three.
| When the prompt sounds like | Start with |
|---|---|
| "Profit dropped" / "margins are falling" | Profitability tree, then value chain to find where cost builds up |
| "Should we enter this market?" | Market entry logic, with PESTLE and Porter's Five Forces for attractiveness |
| "How should we price this?" | The three pricing lenses, anchored on willingness to pay |
| "How do we grow revenue?" | Growth levers and the Ansoff matrix to rank the paths |
| "Should we buy this company?" | The M&A and private equity logic, entry to value to exit |
| "How do we launch this product?" | Customer segmentation, a go to market plan, and unit economics |
| "Is this business healthy / aligned?" | The 3Cs for position, then 7S for internal alignment |
Name the framework to yourself, not to the judges. They want the insight it produces, not a tour of the tool.
A chart exists to make one comparison obvious. If a judge has to study it, it has failed. Edward Tufte's work on analytical graphics is the standard reference, and a few habits cover most cases.
Decide the single comparison you want the reader to make, then pick the chart that makes it obvious at a glance. The comparison comes first, the chart second. Most chart mistakes are a chart chosen before the point was clear.
Use a line. The eye reads the slope as the story.
Use a bar. Length is the easiest visual to judge accurately.
Use a stacked bar. Reach for a pie only with very few slices.
Use a scatter to show how two variables move together.
The chooser suggests the chart that makes that comparison easiest to read, and why.
Put the conclusion in the chart title and annotate the one number that matters. Do not make the reader hunt for the point.
Drop gridlines, heavy borders and 3D effects. Ink that carries no meaning competes with the ink that does.
Start bar axes at zero and keep scales consistent across panels. A truncated axis that exaggerates a gap costs you credibility.
A pie chart asks the eye to compare angles and areas, which people read poorly. Once there are more than two or three slices, or once two slices are close in size, nobody can tell which is bigger. A bar chart answers the same question instantly because the eye compares lengths along a shared baseline. Keep pies for the rare case of two or three very different parts.
Beyond the basics, three charts come up often in cases and signal that you have thought about the right comparison.
Shows how a starting number becomes an ending number through a chain of additions and subtractions. Perfect for a revenue or profit bridge.
Shows how a population shrinks as it moves through stages, from visitors to buyers. The drop between two bars is where you look.
A grid where each row is a group that started together and each column is a later period. Colour fading right shows retention decay.
Judges are usually senior people who have spent years reading dense spreadsheets. A wall of numbers reads as corporate banality. The same figures shown as a picture read as a point. Tap the toggle to see the difference on one set of marketing numbers.
For anything geographic, use a vector map you can recolour, not a screenshot. Shade each region by the value you are showing, keep to a single colour ramp, and let the reader switch what they are looking at. Tap a metric below to recolour this abstract region map. The shapes are illustrative, not a real place.
Three layouts carry a comparison without a chart. A before and after pairing shows change. A benchmark table sets you against peers. A sensitivity table shows how the answer moves when an assumption moves.
Put the two states next to each other so the change is the visual. Illustrative figures.
| Player | Margin | Growth |
|---|---|---|
| You | 14% | 9% |
| Peer A | 21% | 6% |
| Peer B | 12% | 15% |
Set yourself against named peers on the few metrics that matter. Illustrative figures.
A sensitivity table shows the answer across a grid of two assumptions, so a judge can see how fragile it is. The shaded cell is your base case.
| Price change \ Volume change | -10% | 0% | +10% |
|---|---|---|---|
| -5% | -9% | -4% | +1% |
| 0% | -5% | 0% | +6% |
| +5% | -1% | +5% | +11% |
Each cell is the change in profit. The spread across the grid tells the judge how much the recommendation depends on assumptions holding. Illustrative figures.
A risk matrix plots each risk by how likely it is against how much damage it would do. The point is to separate the risks worth a mitigation plan from the ones worth a sentence. The top right corner, high likelihood and high impact, is where your attention and your slide space should go.
Always note your source on the slide. A number a judge cannot trace is a number they can dismiss.
The work around the work. How a team divides effort, handles the room, and holds up when a judge starts pushing.
A case team wins on coordination more than on raw talent. Agree how you will work before you start solving, not halfway through the night when tension is high.
Map who is strongest at structure, analysis, design and delivery. Assign roles to those strengths rather than splitting the deck evenly.
Work apart, then come back together at set checkpoints. Long solo stretches end in two halves that do not fit.
Many hands build slides, but one person guards the through line so the final deck argues as a single voice.
The first round is usually a filter, not the final. The goal is to clear it cleanly without burning the energy you need later.
The first read for the gist, the second for the constraints and the exact question. Most disqualifying mistakes are answers to a question that was not asked.
Agree the spine of the answer before anyone polishes a slide. A clear structure with rough slides beats a beautiful deck with no argument.
Set hard internal deadlines for solving, building and rehearsing. Reserve real time at the end, because the last hour always overruns.
A finished answer to 80% of the problem beats a perfect answer to half of it. Judges grade the whole arc, not one slide.
Lead with the recommendation in the first thirty seconds. The room should know your position before you justify it.
Say where you are and where you are going. "Three reasons, here is the first" lets the room track you without effort.
Plan transitions between speakers. A smooth handoff signals a team that prepared, a clumsy one breaks the spell.
The slide supports you, it is not your script. Hold eye contact with the panel, not the screen.
Rehearse to land inside the limit with margin. Getting cut off mid recommendation is an avoidable loss.
Close by restating the recommendation and its impact. Last words are the ones the panel carries into scoring.
Q&A is where many results are actually decided. Judges probe to see whether you understand your own answer or only memorised it. Treat every question as a chance to show depth, not a threat to survive.
Hear the whole question and take a beat before answering. A short, composed pause reads as confidence, not hesitation.
Give the direct answer in one sentence, then the reasoning. Do not make the judge wait through a wind up for your position.
Be ready to defend any figure on your slides and the assumption behind it. "We assumed X, here is why" beats a blank look.
If a judge finds a real gap, acknowledge it and show how you would close it. A team that can take a hit and adjust scores well.
Estimating a number you cannot look up. Judges are not grading the final figure. They are grading whether your path to it is structured, stated out loud, and sane.
Every guesstimate follows the same arc. Clarify what you are being asked, choose a direction, break the number into parts you can reason about, estimate each part, then combine and sanity check. Say each step aloud. The reasoning is the answer.
Pin down exactly what is being estimated, the geography, the time period and the units. Resolve the ambiguity before you compute.
Top down from a large known total, or bottom up from a single unit. Choose whichever has the more reliable starting number.
Split the target into a chain of factors you can each estimate. Keep the segments clean and non overlapping.
Multiply through, then test the result against anything you know. If it is off by orders of magnitude, find the bad assumption.
The two directions reach the same number from opposite ends. Top down starts from a big population and narrows with filters. Bottom up starts from one unit and scales up. The example below is purely illustrative, with round numbers chosen to show the method, not to be accurate.
When a population is not uniform, split it into groups that behave differently, estimate each, and add them up. Urban and rural, or by age, are common cuts.
Start from a number you are fairly sure of, such as a country's population or a typical household size, and build out from that anchor.
Use round numbers so the arithmetic stays fast and visible. Precision here is false comfort and slows you down.
Be clear whether you want a total that exists at a point in time, a stock, or an amount per period, a flow. Mixing them is the classic error.
Getting the right power of ten matters far more than the leading digit. Aim to be in the right ballpark first.
Every estimate rests on assumptions. Name them as you go so a judge can challenge one number without collapsing the whole chain.
Drills that make estimation feel routine instead of improvised. Build the muscle on these before you need it under a clock.
Start from a big total and apply filters down to the answer.
population × % relevant × usage rate
Best when you trust a large published total, like a country's population, more than you trust a single unit.
Start from one unit and scale up to the total.
per unit value × number of units
Best when you can observe one case well, like sales per store, and the count of units is easy to reason about.
A strong answer runs one direction, then checks it with the other. If the two roughly agree, your estimate is probably sound.
Most guesstimates reduce to one of a few skeletons. Tap what you are estimating to see the decomposition to start from.
The builder shows the standard chain of factors to break that estimate into.
A few round anchors worth carrying in your head. Treat them as planning numbers to adjust, not precise facts.
| Anchor | Round value to start from |
|---|---|
| People per household | 2 to 4 |
| Human lifespan | ~75 to 80 years |
| Working hours per person per year | ~2,000 |
| Share of population that is working age | ~60% |
| Waking hours per day | ~16 |
| You often need | Value |
|---|---|
| Days in a year | 365 |
| Weeks in a year | 52 |
| Hours in a year | ~8,760 |
| Minutes in a day | 1,440 |
| Seconds in an hour | 3,600 |
| Thousand · Million · Billion | 10⁷ · 10⁶ · 10⁹ |
When a population is not uniform, reach for one of these cuts, estimate each group, and add them up.
Children, working age, retired. Behaviour differs sharply by life stage.
Urban versus rural, or by region. Density and access change usage.
Low, middle, high. Drives both who buys and how much.
Heavy, light, non users. The few heavy users often dominate volume.
Penetration today versus the addressable rest of the market.
| Weak | Strong |
|---|---|
| Jumps to a number | Lays out structure first |
| Silent computing | Narrates each step |
| Hidden assumptions | Names every assumption |
| No final check | Sanity checks and caveats |
Estimation is a timed skill. Set a target, talk through a guesstimate out loud, and land before the buzzer.
These are families of structure for the most common case prompts. None of them is a script to fill in. Each one is a starting tree you adapt to the specific question, then prune to the branches that matter.
When profit has fallen or needs to grow, break the number into its parts and find the branch that moved. Profit is revenue minus cost. Revenue is price times volume. Cost is fixed plus variable. Walk down each branch until you reach a number you can act on.
Michael Porter's value chain splits a firm into the activities that add value on the way to the customer. Primary activities move the product through the business. Support activities make the primary ones possible. Use it to locate where cost builds up, or where an advantage actually comes from.
When the question is whether and how to enter a new market, answer it in order. First decide if the market is worth entering. Then decide if the firm can win there. Only then choose the way in. Skipping to the mode of entry before the first two questions is the usual mistake.
Size, growth rate, margins and how those are expected to move. A large market that is shrinking is a different decision than a small one growing fast.
The strength of incumbents, barriers to entry, and whether the firm has a real edge in cost, product or distribution.
Whether the move suits the firm's capabilities, brand and strategy, and what it costs in focus elsewhere.
Build it alone, partner or joint venture, or acquire a player already inside. Each trades speed against cost and control.
There are three honest ways to set a price, and they often disagree. The right answer is usually a blend, anchored on the one that fits the product and the market. Name which lens you are using before you quote a number.
Add a target margin on top of what the unit costs to make and deliver. Simple and defensible, but it ignores what buyers will actually pay.
Anchor on what rivals charge, then position above or below. Easy to justify, but it cedes the pricing decision to others.
Price against the value the customer receives, measured against their next best alternative. Hardest to estimate and usually the most profitable.
Igor Ansoff's matrix sorts growth options on two axes: existing or new products, against existing or new markets. The four quadrants run from least risky to most. It is a fast way to lay out the paths a firm could take and rank them by how far they stretch from the core.
Sell more of what you have to the customers you already reach. Lowest risk. Pricing, share gains and usage.
Take the current product to new segments or geographies. Moderate risk in demand and distribution.
Build new products for the customers you know. Moderate risk in execution and fit.
New product and new market at once. Highest risk, since both unknowns compound.
An acquisition question has a strategy half and a deal half. First ask why the firm would buy at all, and whether buying beats building or partnering. Only if the logic holds do you move to value, price and what happens after the deal closes.
The reason to buy. Scale, a capability, a market, a team, or removing a competitor. If there is no clear reason, stop here.
What the target is worth on its own, plus the value created by combining the two. The premium paid should not exceed that combined value.
Testing the numbers and the risks before committing. Finances, legal exposure, customers and culture.
How the two businesses are joined afterward. Most deals that fail, fail here rather than in the price.
A private equity case usually asks whether a fund should buy a company, and how it would make money on the way out. The logic runs across the whole hold period, not just the purchase. Reason from the exit backward to the entry.
Why this target, at this price, now. A clear view of what is undervalued or fixable, and what the fund brings.
How returns are built during ownership. Growing the business, improving margins, paying down debt, or buying add ons.
How and to whom the fund sells, and the return that path implies. A sale to a strategic buyer, another fund, or a public listing.
The quantitative layer under every case above. Each idea here is one plain definition plus, where it helps, a calculator you can play with. All default figures are illustrative.
A revenue bridge explains how last year's revenue became this year's, one driver at a time. It turns a single number into a story a judge can follow.
Margins are just revenue with costs peeled off in layers. Take revenue, subtract the cost of making the product to get gross margin, then subtract running costs to get operating margin. Contribution margin is a fourth view: price minus the variable cost of one unit, which is what each sale contributes toward fixed costs.
Break-even is the number of units where total contribution covers fixed costs. Below it you lose money, above it you make it.
Payback is how long an investment takes to return its own cost in cash. Quick to compute and easy for a judge to grasp, though it ignores anything after the payback point.
Customer acquisition cost is what you spend to win one customer. Lifetime value is the profit that customer brings over their whole relationship. The ratio between them tells you whether growth is worth funding. A common rule of thumb is that LTV should be at least three times CAC.
Three nested views of market size. The total addressable market is everyone who could ever buy. The serviceable available market is the slice your model can actually reach. The serviceable obtainable market is the share you can realistically win in the near term.
A dollar later is worth less than a dollar now, so future cash flows are discounted back to today. Net present value sums those discounted flows and subtracts the upfront cost. A positive NPV means the project creates value. The internal rate of return is the discount rate at which NPV would be exactly zero, so compare it against the rate the money would otherwise earn.
No model is certain, so show how the answer moves when an assumption moves. Flex one driver at a time across a sensible range and watch the output. The driver that swings the answer most is the one to defend hardest. The sensitivity table in the data chapter is the clean way to present this on a slide.
Every other chapter is a piece. This is how they connect, from the moment you read the prompt to the moment you defend the answer. When you feel lost in a case, find your step on this ladder.
Restate the question in your own words and confirm what success looks like. Most failures start here.
Pin the goal, the time horizon and the limits. What counts as a win, and what is off the table.
Break the question into MECE branches so you can split the work and rule out whole areas fast.
Make a testable guess at the answer so your analysis has a target instead of wandering.
Gather only the evidence that would prove or disprove the hypothesis. Skip the rest.
Trace the symptom down to the driver behind it. Fix the cause, not the surface.
Diverge before you converge. Put several real options on the table, not one.
Score them against the objective and constraints. Make the trade-offs explicit.
Commit to one answer you can defend, sized with its impact and its risks.
Arrange the answer as a pyramid and storyboard the headlines before touching a slide.
One message per slide, action titles, one visual that carries the proof.
Anticipate the hard questions and build the backup slides you hope they ask for.
The loop is not always clean. A falsified hypothesis at step five sends you back to step four. That is the method working, not failing.
Competitions weight criteria differently, but most rubrics look close to this. Knowing roughly where the points sit tells you where to spend your last hour.
| Criterion | Weight | Share |
|---|---|---|
| Analytical depth | 20% | |
| Feasibility | 20% | |
| Problem understanding | 15% | |
| Originality | 15% | |
| Business impact | 15% | |
| Slide clarity | 10% | |
| Q&A handling | 5% |
Depth and feasibility usually carry the most weight. A clever idea that cannot be executed scores worse than a plain one that can.
Live cases reward speed and composure with numbers. A few shortcuts keep you from fumbling arithmetic while a judge watches.
To find how long something takes to double, divide 72 by the growth rate in percent. At 8% growth, doubling takes about nine years. It works in reverse too: to double in a target number of years, divide 72 by that number to get the rate you need.
Awkward percentages become easy when you swap them for a fraction. To take 25% of a number, divide by four. Memorise this short list and most percentage work turns into division.
| Percent | Fraction | To apply it |
|---|---|---|
| 50% | 1 / 2 | Halve it |
| 33% | ~1 / 3 | Divide by three |
| 25% | 1 / 4 | Divide by four |
| 20% | 1 / 5 | Divide by five |
| 12.5% | 1 / 8 | Divide by eight |
| 10% | 1 / 10 | Move the decimal one place |
Separate the front digits from the zeros. For 8 million times 3 thousand, multiply 8 by 3 to get 24, then count the zeros: six plus three is nine. The answer is 24 followed by nine zeros, which is 24 billion. Do the small multiplication, then reattach the magnitude.
x% of y always equals y% of x. So 16% of 25 is the same as 25% of 16, which is just a quarter of 16, or 4. When one side is an easy fraction, flip the problem so the easy side does the work.
Find 10% by moving the decimal, then build the number you want from it. 30% is three tens. 5% is half a ten. 15% is a ten plus a five. Most live percentage maths is just adding and halving tens.
The jargon used across this guide, in one place and in plain language. Type to filter.
This guide was written from scratch as original work. It follows the familiar shape of a university consulting field guide, but no text, cases, figures or numbers were copied or adapted from any existing casebook.
Every explanation, diagram and example here was written and drawn fresh. The examples and numbers are invented for teaching and are not meant to be accurate. The methods themselves are long established in business and consulting practice, and are taught here in plain language.
Keep this attribution. Treat each named framework as the work of its author. This guide makes no claim over the methods it explains, only over the way it explains them.
The pyramid principle and the SCQA structure for ordering an argument.
The six thinking hats method for separating modes of thinking.
The five forces of industry competition and the value chain.
The classic rules of brainstorming behind the idea generation section.
The product and market growth matrix used in the growth section.
The growth share matrix for sorting a portfolio of business units.
The principles of clear analytical graphics behind the data section.
Standard planning frameworks in common use across business education.