Advanced Secondary 2 entrepreneurship vocabulary becomes useful when a learner can turn a business idea into a sequence of questions that can actually be tested. This manual teaches 100 entrepreneurship terms for customer discovery, unit economics and market research, including problem interviews, behavioural evidence, willingness to pay, market sizing, contribution margin, customer acquisition cost, lifetime value, retention, runway, bottlenecks, sensitivity analysis and decision rules. The purpose is not to make a Grade 8 business project sound like a venture-capital presentation. It is to help a student distinguish an exciting assumption from an observation, a popular product from a viable unit, and a persuasive pitch from a decision supported by evidence.

Students searching for advanced Grade 8 entrepreneurship vocabulary, business terms with definitions and examples, startup vocabulary, market research vocabulary or unit economics explained for students often meet broad glossaries that name marketing, finance and enterprise concepts separately. Real entrepreneurial reasoning connects them. A customer interview changes a problem hypothesis. A price changes contribution margin. A conversion rate changes acquisition economics. A supplier delay changes capacity and cash flow. A forecast changes when one assumption changes. This collection therefore teaches the relationships among terms, not only their dictionary meanings.

Alicia, Tricia and Kai Kai work through a fictional school-scale enterprise laboratory. Every customer, dataset, price, cost and decision in the worked cases is invented for teaching unless a source is explicitly linked. The manual does not provide investment, legal, tax or personal financial advice, and it does not promise that a classroom business model would succeed commercially. Use the foundation Secondary 2 entrepreneurship glossary when the learner needs broad business vocabulary first. Use this advanced extension when the learner is ready to test assumptions, interpret evidence and explain why one next decision is better supported than another.

The 60-second route through this manual

For customer discovery and research design, begin with Terms 1–20. For value, markets and competitive position, move to Terms 21–40. Terms 41–60 build business models and unit economics. Terms 61–80 connect growth metrics with operations and cash. Terms 81–100 cover forecasting, uncertainty, metrics, ethics and decision records. The later laboratories combine these ideas in original evidence packets so that students have to calculate, compare, qualify and recommend rather than copy definitions.

Secondary 2 and Grade 8 are overlapping learner labels in some education systems, not a universal business curriculum. The technical level here is extension. The student does not need to memorise all one hundred terms in order. Select the words required by the current project, learn their boundaries, use them in a decision and return after a delay. The eduKate Vocabulary Learning System owns the wider meaning-to-retrieval-to-transfer method; this article owns the advanced entrepreneurship application.

A boundary before the first business case

The examples are educational models. Keep projects low-risk, age-appropriate and supervised. Do not ask students to spend money they cannot afford to lose, collect unnecessary personal information, sign contracts, provide regulated financial services or test products whose safety requires professional validation. A school project can learn a great deal from fictional data, prototypes, interviews about harmless everyday problems and transparent arithmetic. The learning target is disciplined reasoning under uncertainty, not the performance of being a real startup founder.

The U.S. Small Business Administration planning resources provide a public example of how market research, competitive analysis, startup costs and break-even analysis connect in business planning. The NSF I-Corps programme provides an example of customer discovery as an experiential entrepreneurial process built around talking with potential customers and stakeholders. These sources provide context; they have not reviewed or endorsed this manual or its fictional cases.

Opening case: the idea everyone loved until the first useful question

Kai Kai proposes a reusable study-planner kit. His friends say it looks impressive. Alicia wants to design the packaging. Tricia asks what problem the kit solves better than a notebook, calendar app or free printable. Nobody can answer clearly. The group has enthusiasm, several features and no evidence that a target user experiences the problem strongly enough to change behaviour or pay for a solution. Their first business task is therefore not branding. It is discovery.

They write three columns. Column one contains what they currently believe: students forget tasks, want a physical system and would pay for a compact planner. Column two contains what they have actually observed. Column three contains what test would most reduce uncertainty. The vocabulary below gives names to those relationships. A term is mastered when it changes what the group asks, measures, calculates or decides.

Terms 1–20: customer discovery and research discipline

1. Customer discovery

Customer discovery is a structured process for learning about potential customers, users, partners and the situations in which a proposed solution might create value. It focuses on evidence before commitment. Teams investigate problems, behaviours, alternatives and decision criteria instead of beginning with a finished product and asking people to praise it. In the fictional planner project, discovery begins by asking how students currently organise work, what fails, what they do when it fails and which alternatives they already use.

Natural phrases include conduct customer discovery, discovery interview and evidence from customer discovery. The NSF I-Corps model treats customer discovery as a central entrepreneurial activity in evaluating market potential and a possible business model. That does not mean a Secondary 2 student should imitate a professional programme’s interview count or commercial ambitions. The transferable idea is simpler: talk to relevant people to learn, not merely to sell. Discovery should be designed to change a belief when evidence disagrees with it.

2. Problem interview

A problem interview explores how a person experiences a problem before the interviewer promotes a particular solution. Useful questions ask about recent events, current workarounds, consequences and priorities. “Tell me about the last time you forgot an assignment” can produce more concrete evidence than “Would you buy our amazing planner?” The first invites an account of behaviour. The second combines a sales pitch, a proposed solution and a hypothetical purchase question.

Write run a problem interview and probe the current workaround. Tricia asks for a recent example and follows the sequence: what happened, what the student tried, how long the repair took and what mattered most. She avoids turning one emotional story into evidence that every student has the same problem. The interview is a source of qualitative evidence about this participant’s experience. Repeated patterns can inform a hypothesis, but the sampling method and diversity of participants still matter.

3. Solution interview

A solution interview explores how a potential user responds to a proposed approach after the problem has been investigated enough to justify showing one. It can examine whether the concept is understandable, useful, missing a requirement or incompatible with existing behaviour. The goal is not to collect compliments. A participant saying “looks nice” may tell the team little about whether the solution changes a real outcome.

Use test the proposed solution and distinguish explanation from persuasion. Alicia shows a low-cost paper prototype and asks the student to complete a realistic planning task. She observes where the user hesitates and asks what they expected to happen. This provides richer evidence than describing every feature first. A solution interview should not erase the discovery phase: if the team has misunderstood the problem, polishing the prototype may simply optimise the wrong thing.

4. Open-ended question

An open-ended question allows a respondent to construct an answer rather than choose from a small fixed set. “How do you currently decide what to study first?” can reveal routines the researcher did not anticipate. Open-ended does not mean vague. A good question can be specific about time, event or behaviour while leaving the content of the answer open.

Natural phrases include ask an open-ended question and probe the response. Kai Kai asks, “What do you like about planners?” and receives general praise. Tricia changes it to, “Walk me through yesterday evening from the moment you opened your school bag.” The second question anchors memory in an event. Students should learn that interview quality depends on the information need. Open questions are valuable for discovery; closed questions can be useful later when a well-defined category needs structured comparison.

5. Leading question

A leading question suggests a preferred answer through its wording or assumptions. “Don’t you think this colour makes the planner easier to use?” signals the desired response and mixes usability with colour preference. A more neutral alternative asks what, if anything, made the planner easier or harder to use. Leading questions can create evidence that reflects the interviewer’s expectations as much as the respondent’s experience.

Use the wording leads the respondent and identify the direction of pressure. Not every question with context is leading. The problem appears when the wording narrows the acceptable answer or embeds an unverified premise. Alicia reviews the team’s script and removes praise words before testing. She also records the exact question used when a surprising answer appears. That allows the group to determine whether the result came from genuine evidence or from how the interview was framed.

6. Observation

Observation records behaviour, events or conditions directly within a defined setting. In entrepreneurship, observation can reveal workarounds that people forget to mention. A student might say they “usually plan ahead” while their desk shows loose task sheets, reminders on a phone and a timetable with handwritten changes. The observation does not reveal private motives automatically; it records what was seen under the stated conditions.

Use observe the current process and record behaviour without interpretation as separate stages. Tricia first writes, “The student checked three different places before choosing the first task.” She does not immediately write, “The student is disorganised.” The first is an observation. The second is a broad personal judgement. Advanced research separates behaviour from interpretation so that the team can ask whether the pattern is repeated, costly or actually connected to the proposed problem.

7. Behavioural evidence

Behavioural evidence is information about what people actually did, chose, used, purchased, abandoned or repeated, rather than only what they said they might do. Clicking a reminder, returning to a prototype or choosing one option when alternatives are available can provide behavioural evidence. Its meaning still depends on context. A click may indicate curiosity rather than purchase intent.

Write the behavioural evidence shows only for the observed action. Alicia sees that participants repeatedly add their own colour labels to the prototype. That behaviour may suggest a need for flexible categorisation. It does not prove that users would pay more for custom colours. A mature entrepreneur treats behaviour as stronger for some questions, while recognising that an observation still needs interpretation and may be shaped by the test environment.

8. Stated preference

A stated preference is what a person says they prefer, intend or value in response to a question. Stated preferences are useful for understanding attitudes, priorities and perceived trade-offs. They can differ from later behaviour because circumstances, memory, social pressure or real costs influence action. “I would definitely use this every day” is therefore evidence about the response, not a guarantee of daily use.

Use reported preference and preserve the question wording. Kai Kai records eight positive statements and calls them eight future customers. Tricia narrows the result: eight interviewees said they preferred the proposed layout to the comparison version in that interview. The repair retains the information while removing an unobserved purchase conclusion. Stated preference becomes especially useful when compared with what users currently do and what constraints might prevent the preferred action.

9. Revealed preference

A revealed preference is inferred from choices people make when alternatives and consequences are present. If a participant repeatedly chooses a simpler planner over a feature-rich version when both are available, the behaviour reveals something about the decision under those conditions. It does not establish a timeless preference across every price, task or context.

Write choice behaviour suggests rather than pretending the preference is an internal fact directly observed. Alicia distinguishes a free classroom choice from a paid market choice. The absence of a real price means the exercise cannot establish willingness to pay. Revealed preference is powerful because action can expose trade-offs that hypothetical answers hide, but its interpretation still depends on what options the participant actually had and what they knew when choosing.

10. Research question

A research question states what the team is trying to learn from an investigation. “Do students want our planner?” is broad and invites confirmation. “Which current planning failures cause students to miss tasks, and how often do they occur among the participants we can study?” is more specific. A good research question identifies the uncertainty that matters to a decision.

Use the research question determines the evidence needed. Tricia writes one question per investigation rather than collecting every possible opinion at once. This discipline improves both method and analysis. A survey designed to estimate frequency differs from an interview designed to understand causes. When students complain that data are confusing, the problem often began before collection: the team never decided exactly which question the data were meant to answer.

11. Hypothesis

A hypothesis is a testable proposed explanation or prediction. In entrepreneurship, a hypothesis can concern a customer problem, preferred solution, price, channel or business model. “Secondary 2 students who currently use three or more separate planning tools will value one integrated weekly view” is more testable than “students want organisation.” The wording identifies a group, current condition and proposed relationship.

Natural phrases include test the hypothesis, evidence supports the hypothesis and revise the hypothesis. A hypothesis is not a promise to prove the founder correct. Its usefulness comes partly from the possibility of being wrong. Alicia writes her expectation before seeing the result so that she can tell whether the evidence changed the idea or the idea was rewritten afterward to fit whatever happened.

12. Falsifiability

Falsifiability is the property of a claim being framed so that conceivable evidence could show it to be wrong. “People will like it somehow” is difficult to falsify because any positive comment can be used to rescue it. “At least six of ten target users will complete the planning task without explanation using this layout” gives the team a criterion that can fail.

Use make the hypothesis falsifiable. Kai Kai initially treats every negative reaction as a bad participant. Tricia asks what result would actually cause the team to change direction. If no possible result can do that, the activity is promotion rather than a meaningful test. Falsifiability does not mean one failed test kills every related idea. It means the particular claim has a defined relationship to evidence and can be revised when that relationship breaks.

13. Assumption

An assumption is something treated as true for the purpose of planning or reasoning without yet being established. Every business model contains assumptions: who has the problem, what they use now, what they value, how much a solution costs to deliver and how people might discover it. Assumptions are unavoidable. The danger is forgetting which statements are assumptions and speaking about them as facts.

Write state the assumption and test the assumption. Alicia creates an assumption register with three columns: belief, evidence and consequence if wrong. “Students will pay twelve dollars” sits in the belief column until an appropriate test supports it. This simple separation turns confidence into a manageable research plan. The most important assumptions are not always the most uncertain; they are often the ones that would damage the model most if wrong.

14. Riskiest assumption

The riskiest assumption is the unverified belief whose failure would most seriously undermine the proposed model or next decision. A team may be uncertain about package colour, but if customers do not experience the target problem, colour is not the first uncertainty to reduce. Risk combines uncertainty with consequence.

Use identify the riskiest assumption before building an elaborate prototype. Tricia ranks assumptions by two questions: “How uncertain is this?” and “What breaks if it is wrong?” The integrated-planner team decides that the existence and severity of the problem matter before logo design. That changes their next action from graphic design to customer interviews. The term is valuable because it converts an overwhelming list of unknowns into a priority order for learning.

15. Experiment

An experiment is a planned test designed to reduce a specific uncertainty by observing what happens under defined conditions. In an entrepreneurial setting, an experiment may use a prototype, comparison, landing-page simulation, interview script or fictional dataset. It should state the hypothesis, measure, success criterion and decision that the result could affect.

Use design a low-cost experiment and avoid calling every activity experimental merely because it is new. Kai Kai posts a colourful image and receives reactions; without a defined hypothesis or appropriate comparison, the activity may produce interesting feedback but weak evidence for pricing. Alicia instead tests whether participants can complete a core task using two layouts. The simpler experiment answers a narrower question and therefore produces a more interpretable result.

16. Testable metric

A testable metric is a defined quantity or outcome used to evaluate a hypothesis. It should connect directly to the uncertainty being tested. If the claim concerns whether a layout reduces missed steps, task-completion errors may be relevant. Likes on a promotional image are not a direct metric of planning accuracy.

Natural phrases include define the metric before the test and the metric does not match the hypothesis. Tricia requires each metric to have an operational definition: what counts, when it is measured and which denominator is used. This prevents a team from declaring success by switching to whichever number looks strongest after the results arrive. Metrics are not neutral simply because they are numeric. Their usefulness depends on the relationship between the measure and the claim.

17. Sample

A sample is the group of people, cases or observations actually studied. Five interviewees can reveal useful problem patterns while remaining a very limited basis for population-level estimates. Sample size matters, but selection matters too. Ten carefully chosen target users can answer a different question from one hundred random visitors to an unrelated page.

Use the observed sample and the target population separately. Alicia writes “six of eight interviewed students” instead of “75% of students” because her sample is not the entire population. The count stays visible and the conclusion remains bounded. Advanced entrepreneurship requires resisting the temptation to convert every small test into a market statistic. A sample can guide learning without pretending to represent everyone.

18. Sampling bias

Sampling bias occurs when the selection process systematically over-represents or under-represents people relevant to the research question. Testing a planner only with students already enrolled in a study-skills club may produce unusually organised participants. The problem is not merely that the sample is small; it is that membership may be related to the behaviour being studied.

Write the recruitment method may bias the sample and explain the pathway. Tricia asks who had no chance to be included. Increasing the number of participants from the same club may produce a larger biased sample. A useful repair could recruit across several relevant routines or narrow the conclusion explicitly to the group reached. The term should lead to a design change or scope change, not become a vague reason to dismiss every survey.

19. Response bias

Response bias occurs when the way people answer is systematically influenced by question wording, social expectations, memory limitations or the setting. A student may praise a prototype more strongly when its creator is standing beside them. A hypothetical purchase question may produce optimistic answers because no real cost is present.

Use possible response bias and name the mechanism. Alicia uses neutral wording and, where appropriate, lets participants complete a task before discussing opinions. That does not remove every influence. It makes the process more interpretable. Response bias is different from sampling bias: one concerns who enters the sample, the other concerns how answers may be distorted after participation. A strong evaluation can identify both without treating the entire project as worthless.

20. Insight

A customer insight is an evidence-grounded understanding that changes how the team interprets a problem, customer or decision. “Students like blue” is an observation if respondents chose blue. “Students need one place where changing deadlines remain visible” may become an insight when interviews, observation and task behaviour repeatedly support that underlying need.

Use derive an insight from evidence and keep the source trail visible. Tricia distinguishes an insight from a slogan. An insight should explain something important enough to change what the team tests or builds next. It also remains provisional. New evidence can refine or overturn it. The learner shows advanced control when they can trace the path from observation to pattern to interpretation to next experiment without turning the final interpretation into an unquestionable fact.

Terms 21–40: value, markets and competitive position

21. Problem–solution fit

Problem–solution fit is the degree to which a proposed solution meaningfully addresses a problem that a defined customer group actually experiences. It is earlier and narrower than product–market fit. The team is asking whether the problem matters and whether this approach appears capable of helping, not whether a repeatable market engine has already been established.

Use evidence of problem–solution fit carefully. Alicia’s paper prototype reduces missed steps in a classroom task for several target users. That supports the usefulness of the proposed mechanism under the test conditions. It does not establish demand at a commercial price or long-term use. A strong student can explain which evidence supports the problem, which supports the solution and which commercial questions remain unanswered.

22. Value proposition

A value proposition explains why a defined customer should consider an offering: which important job, problem or desired outcome it addresses and how the proposed value differs from alternatives. “A beautiful planner for students” describes audience and appearance. “A weekly planning system that keeps changing deadlines visible in one place” makes the intended outcome clearer.

Write test the value proposition, not simply write a value proposition. Tricia treats the sentence as a hypothesis about what customers care about. Interviews and task behaviour may reveal that portability matters more than deadline visibility, forcing a revision. A value proposition should guide evidence collection and product choices. It becomes weak when it turns into a generic promise such as high quality, convenient and affordable without explaining whose problem is solved or which trade-off is improved.

23. Job to be done

A job to be done is the progress a person is trying to make in a particular situation. The phrase shifts attention from a product category to the customer’s desired outcome. A student does not necessarily want a planner for its own sake. They may want to remember commitments, decide what to do next and reduce the stress of discovering forgotten tasks late.

Use the customer is trying to accomplish before using the technical label. Kai Kai initially lists features: tabs, stickers and pockets. Alicia asks what job each feature serves. If a feature does not improve the important job, it may add cost without adding value. The framework should not be treated as a magic method that automatically reveals motivation. It is a way of asking a sharper question about progress, alternatives and circumstances.

24. Pain point

A pain point is a specific frustration, cost, delay, risk or difficulty experienced by a customer in a process. “Studying is hard” is broad. “Deadlines are scattered across several systems, so students repeatedly spend time reconstructing what is due” is more actionable. Severity matters: a mild annoyance may not justify switching behaviour or paying for a solution.

Write evidence of the pain point and avoid inventing emotional intensity. Tricia records frequency, consequence and current workaround. A problem becomes more commercially relevant when people already spend time, money or effort to manage it. That does not mean every painful experience should be monetised. For a school project, the concept helps students prioritise problems that are concrete enough to investigate and solve responsibly.

25. Gain

A gain is a positive outcome a customer values, such as saving time, reducing errors, gaining confidence or making a process easier to coordinate. Gains are not always the opposite of pain points. A user may tolerate the current process but still value a new capability, such as sharing a weekly plan with a parent or study partner.

Use desired gain and distinguish it from a feature. A colour-coded dashboard is a feature; faster identification of the next task may be the gain. Alicia asks participants which outcome matters and how they currently achieve it. This prevents the team from treating every attractive capability as value. A useful gain should connect to a decision the customer actually cares about.

26. Willingness to pay

Willingness to pay is the maximum amount a customer would be prepared to exchange for an offering under specified conditions. It is not the same as saying an item looks useful or attractive. Hypothetical answers can differ from real purchase behaviour because a real payment requires giving up money that could be used elsewhere.

Write evidence about willingness to pay rather than converting compliments into prices. Kai Kai asks, “Would you pay twelve dollars?” after a long explanation of how much effort the team invested. Tricia recognises social pressure and changes the test. In a school setting, use low-risk simulations or teacher-approved choices rather than pressuring peers to spend money. The key lesson is that price is a trade-off, not a popularity score.

27. Switching cost

A switching cost is the time, effort, money, learning or risk a customer faces when moving from an existing solution to a new one. The cost can be psychological or practical as well as financial. A free planning app can still have a high switching cost if a student has already built months of habits, labels and reminders in another system.

Use reduce the switching cost and identify its source. Alicia realises that asking users to re-enter every existing task makes the new system less attractive even if the interface is clearer. The team considers an import or transition method in its fictional design. Advanced entrepreneurship recognises that value is judged relative to the status quo. A new product can be better in isolation and still fail to justify the cost of changing.

28. Early adopter

An early adopter is a customer willing to try a relatively new solution before broad adoption, often because the problem matters enough to tolerate imperfection or because the new approach offers special value. Early adopters are not simply people who say yes quickly. They should have a reason connected to the problem and the proposed value.

Use identify a plausible early-adopter group. Tricia notices that students with frequently changing co-curricular schedules struggle more with static weekly planners. That group may be more motivated to test a flexible layout. Evidence from early adopters can accelerate learning, but their needs may differ from later mainstream users. A successful early test should not be generalised automatically to everyone.

29. Beachhead market

A beachhead market is a deliberately narrow initial market segment chosen so a team can focus its learning and resources before expanding. The term comes from a strategic metaphor and should be explained rather than used theatrically. A school project might focus first on students managing several weekly activity schedules instead of claiming to serve all teenagers everywhere.

Write focus the initial market and state why the segment is useful. Alicia looks for common needs, reachable users and a realistic test environment. A narrow market is not automatically good if it is too small or unrepresentative of the future model. The concept is valuable because focus makes assumptions easier to test. It reduces the temptation to solve different problems for different people in one confusing prototype.

30. Segmentation

Segmentation divides a broader market into groups that differ in relevant needs, behaviours, constraints or buying conditions. Age alone may be a poor segment when scheduling complexity is the variable that changes the value proposition. Useful segments are connected to a decision: product design, message, channel, price or service model.

Use segment by behaviour, need or another justified dimension. Kai Kai creates personas based only on favourite colours. Tricia asks whether colour preference changes the planning problem. If not, the segmentation may be decorative. A strong segment explains why customers within the group are similar in a way that matters to the business decision, while groups differ in a way that changes the model.

31. Positioning

Positioning is the place an offering aims to occupy in a customer’s mind relative to alternatives on dimensions that matter. It answers questions such as: for whom is this product, in what category, for which use and why should it be chosen? Positioning is not merely a slogan. It connects customer, competitor and value proposition.

Use position the product as and name the comparison. Alicia positions the fictional planner not as “the best planner” but as “a low-friction weekly system for students whose schedules change often.” The statement can then be tested against customer needs and alternatives. Strong positioning accepts trade-offs. If the team tries to be simultaneously cheapest, most premium, simplest, most customisable and most comprehensive, the message loses meaning.

32. Differentiation

Differentiation is a meaningful difference that gives a customer a reason to choose one offering rather than alternatives. A difference matters only if the target customer values it enough to affect the decision. A unique packaging shape can be distinctive without being useful. A faster transition from existing tools may be less visible but more valuable.

Write meaningful differentiation and link the difference to evidence. Tricia asks, “Different in what way, for whom, and why does that matter?” A feature is not automatically differentiation because competitors can copy it or customers may ignore it. The strongest school project answer explains the customer outcome created by the difference and the evidence suggesting that outcome matters.

33. Substitute

A substitute is an alternative way a customer can accomplish the same or a similar job. The substitute may belong to a different product category. A paper planner competes not only with other planners but with calendar apps, sticky notes, messaging oneself, memory and combinations of those methods. Ignoring substitutes can make a market appear emptier than it really is.

Use current substitute and ask why customers continue using it. Kai Kai searches for products with the same shape and finds few competitors. Alicia observes that students already solve the problem through several free tools. Those substitutes reveal the true cost of switching and the minimum improvement a new solution must offer. Competition begins with the customer’s job, not with matching category labels.

34. Competitor

A competitor is another organisation or offering seeking to satisfy overlapping customer needs or capture related demand. Competitors can be direct, indirect or potential. Their existence is not proof that a new opportunity is impossible. It can show that customers already allocate attention and money to the problem.

The SBA’s planning guidance treats competitive analysis as a way to understand competing offerings, market position, strengths, weaknesses and barriers. In a student project, compare what alternatives do well before listing their flaws. Tricia creates a feature–outcome–price table but refuses to declare a winner from a single score. Competitor analysis should reveal trade-offs, underserved needs and assumptions to test, not become a ritual insult to existing solutions.

35. Total addressable market (TAM)

Total addressable market, often abbreviated TAM, estimates the total demand available if an offering could serve the entire relevant market under a stated definition. TAM is not a forecast of what a new venture will actually sell. Its usefulness depends on how the market boundary, time period and price basis are defined.

Use estimate TAM and show the arithmetic. Kai Kai multiplies the number of all students in the world by an imagined price and calls the result revenue. Tricia repairs the model: that calculation is, at most, a broad market-size thought experiment under unrealistic universal adoption. It says nothing about reach, competition, ability to pay or product fit. Large market numbers should make assumptions more visible, not less.

36. Serviceable available market (SAM)

Serviceable available market, or SAM, is the portion of the broader market that fits the venture’s actual offering, geography, capabilities or customer definition. It narrows the theoretical opportunity to the area the business model could plausibly serve in its current form.

Write define the serviceable market and explain the constraints. Alicia’s fictional planner is only available in one language and sold through a school event in the model. Those limitations make the serviceable market far smaller than a global student population. SAM is not automatically the same as the beachhead market: a venture may be capable of serving several segments while choosing one initial segment for focus.

37. Serviceable obtainable market (SOM)

Serviceable obtainable market, or SOM, is an estimate of the portion of the serviceable market the venture could realistically capture within a defined period and set of assumptions. It should reflect channel reach, competition, capacity and conversion rather than a founder’s enthusiasm.

Use build a bottom-up SOM estimate. Tricia starts from reachable events, expected visitors, observed conversion assumptions and production capacity. She does not take one per cent of a giant global market merely because the result looks small enough to sound conservative. A smaller estimate can still be unrealistic if its pathway is unexplained. Good market sizing shows how the number could be reached.

38. Market share

Market share is the proportion of a defined market’s sales, units, users or other relevant activity associated with one organisation or offering. The numerator and denominator must use compatible definitions and periods. A share of sales revenue can differ from a share of units when prices differ.

Write market share by revenue or by unit volume when the distinction matters. Kai Kai divides his fictional sales by the total number of students and calls the result market share. Tricia asks whether non-buyers belong in the denominator for the chosen measure. A correct share requires a clearly defined market and comparable data. The concept is simple mathematically and demanding conceptually because the denominator can change the story.

39. Demand

Demand concerns the quantity of an offering customers are willing and able to obtain under particular price and market conditions. Interest alone is not demand in the economic sense. A hundred people saying a free sample looks useful does not establish how many would buy at the proposed price.

Use evidence of demand at this price and condition. Alicia treats pre-orders in a fictional low-risk simulation differently from social reactions. The closer the observed action is to the real decision and constraint, the more directly it may inform demand. Still, one event or one customer group does not define a full demand curve. Advanced writing states the conditions under which the behaviour occurred.

40. Price sensitivity

Price sensitivity describes how strongly customer choice or demand changes when price changes, holding other relevant conditions sufficiently comparable. It is related to, but not identical with, the formal economic concept of price elasticity. A school project can compare choices across stated prices without claiming it has estimated a full market elasticity.

Write responses were more sensitive to price in this test when the evidence supports it. Tricia avoids asking one group at six dollars and a completely different group at twelve, then attributing every difference to price. The comparison needs comparable participants or an appropriate design. Price sensitivity matters because a small change in price can affect conversion, margin and perceived positioning simultaneously. It therefore belongs to both market research and unit economics.

Terms 41–60: business models and unit economics

41. Business model

A business model explains how an organisation creates value for a defined customer, delivers that value through activities and partners, and captures enough value to sustain the work. It links customer, offer, channel, operations, revenue and costs. A product idea is therefore not yet a business model. “We sell planners” says what is offered but not who buys, how customers discover it, what it costs to deliver or why the exchange remains viable.

Use test the business model and separate its components into assumptions. Alicia maps the fictional planner project: target user, problem, value proposition, sales channel, production process, price and cost. Tricia marks the weakest evidence in each area. The model is not a decorative canvas to complete once. It is a connected hypothesis system. Changing one component can alter others: a lower price may require a cheaper channel; a different customer segment may need another feature or service level.

42. Revenue model

A revenue model describes how money enters the organisation from customers or other payers. Examples include one-time sales, subscriptions, usage fees, licensing and transaction fees. It is one part of the broader business model. Two organisations can deliver similar products but use different revenue models, creating different relationships with customers and costs.

Write one-time sale or recurring revenue rather than assuming subscription is automatically superior. Kai Kai wants a monthly planner subscription because subscriptions sound modern. Alicia asks what recurring value customers receive and what recurring service the team must provide. If the product is mostly a durable physical item, a forced subscription may create friction. A revenue model should fit customer value and delivery economics rather than imitate a fashionable category.

43. Pricing model

A pricing model defines how a price is structured: per unit, per user, per bundle, per month, by usage or through another basis. It differs from the specific price level. Ten dollars may be the price; “per kit” is part of the pricing model. The model changes what customers compare and how revenue scales with use.

Use pricing basis and state what is included. Tricia compares a single-kit price with a bundle containing refills. A lower headline price can be misleading if essential components are excluded. Conversely, a higher bundle price can offer better value when the customer would otherwise buy several items separately. The advanced task is to connect the model to customer use and cost behaviour rather than treat price as one isolated number.

44. Fixed cost

A fixed cost is a cost that does not change directly with the number of units produced or sold within the relevant range and time period. Examples in a simplified school enterprise might include a one-off display fee or a monthly software subscription. Fixed does not mean permanent. A cost can change when the scale, contract or time period changes.

Use fixed for this period and capacity range. Alicia separates the fictional thirty-dollar stall fee from paper used in each planner. The fee remains thirty dollars whether ten or twenty units are sold at that event. That distinction matters for break-even analysis because fixed costs must be covered by contribution from units sold. Avoid calling every large cost fixed or every small cost variable; classification depends on behaviour, not size.

45. Variable cost

A variable cost changes with the volume of units, orders or activity. Paper, packaging and payment fees may behave approximately this way in a simple model. The relevant unit must be defined. A delivery fee can be variable per order while the number of items inside the order also varies.

Write variable cost per unit and identify which costs the figure includes. Kai Kai calculates four dollars of materials and calls that total variable cost. Tricia notices a per-sale payment fee. Omitting it inflates the unit contribution. The goal is not accounting perfection in a school exercise; it is consistency between the model and the real costs that move with sales. If a cost changes only after certain volume thresholds, the simple model may need refinement.

46. Marginal cost

Marginal cost is the additional cost of producing one more unit or serving one more customer, under the conditions being analysed. It can differ from average variable cost when discounts, capacity limits or step changes appear. In a simple classroom model with constant materials and fees, marginal cost may be approximated by the variable cost per additional unit.

Use cost of the next unit to explain the idea. Alicia asks whether making the fifty-first kit triggers the need to buy another whole pack of special material. If so, the immediate incremental cost can jump. The concept helps students understand why cost behaviour is not always a smooth straight line. It also supports decisions about whether extra sales remain worthwhile at the margin.

47. Sunk cost

A sunk cost is a cost already incurred that cannot be recovered by the current decision. Rationally, it should not determine whether the team continues when future costs and benefits now point elsewhere. This does not mean past spending is meaningless; it can provide learning. It means the decision should not be justified simply because money or effort has already been spent.

Write do not chase the sunk cost carefully. Kai Kai argues that the team must keep a complicated feature because they spent six hours designing it. Tricia asks whether the feature improves the future product enough to justify future production and support costs. The six hours have already passed. The learning from them remains useful; the obligation to keep the feature does not. This distinction protects teams from escalating commitment to weak ideas.

48. Opportunity cost

Opportunity cost is the value of the best alternative forgone when a choice is made. Resources used for one experiment cannot be used simultaneously for another. In a student project, the main scarce resource may be time rather than money. Spending three sessions refining packaging means not spending those sessions interviewing users or testing the core task.

Use the opportunity cost of this choice and name the realistic alternative. Alicia compares two next steps with the same available afternoon. One improves the logo; the other tests the riskiest assumption. The opportunity-cost framework does not make the second automatically correct. It forces the team to recognise what is being given up and whether that sacrifice matches the project’s current uncertainty.

49. Contribution margin

Contribution margin is the amount of revenue from a unit or sale left after the relevant variable costs are subtracted, available to cover fixed costs and then contribute to profit. If a fictional planner sells for twelve dollars and incurs five dollars of variable cost, the unit contribution is seven dollars. This is not the same as total profit because fixed costs remain.

Use contribution margin per unit when the calculation is price minus variable cost per unit. SBA break-even guidance uses the relationship between fixed costs, selling price and variable cost to determine break-even units. Alicia checks that every cost in the variable figure belongs to the same unit basis. A contribution margin can be positive while the project still loses money overall if fixed costs are high or sales volume is low.

50. Gross margin

Gross margin describes gross profit relative to revenue, using the cost classification applied in the accounting model. In a simplified educational product example, if revenue is twelve dollars and cost of goods sold is five, gross profit is seven and gross margin is about 58.3%. Exact professional definitions depend on accounting treatment, so a school model should state what costs are included rather than imply universal equivalence with contribution margin.

Write gross margin under this simplified model. Tricia distinguishes an amount from a percentage: seven dollars is a gross-profit amount; 58.3% is a margin percentage. Kai Kai initially says “we make 58% profit,” which can imply net profitability after all expenses. The repair names the level of the calculation. Advanced vocabulary should make the financial statement narrower and more informative, not merely more technical.

51. Markup

Markup is the amount added to a cost to arrive at a selling price, often expressed as a percentage of cost. If a product costs five dollars and sells for eight, the three-dollar markup is sixty per cent of the five-dollar cost. Margin uses selling price as the denominator, so the corresponding gross margin is 37.5% in this simplified example.

Use markup on cost and margin on sales to preserve the denominator. Alicia writes both formulas beside each other because the same price and cost can produce different percentages. This is a common business-language trap: a 50% markup is not a 50% margin. The repair begins with the denominator, not the label. A student who can reconstruct the ratio has stronger knowledge than one who remembers the words but confuses their bases.

52. Break-even point

The break-even point is the level at which total revenue equals total cost under the model, producing neither profit nor loss at that point. In a single-product simplified model, break-even units can be estimated as fixed costs divided by contribution margin per unit. If fixed costs are seventy dollars and contribution per unit is seven dollars, ten units are needed to cover those fixed costs.

Use break-even under these assumptions. SBA guidance presents the same basic formula using fixed costs divided by price minus variable cost. Tricia refuses to call ten units guaranteed sales or proof that the project is viable. Break-even is an analytical threshold based on assumptions about price, costs and units. If the price changes, the variable cost changes or unsold inventory remains, the threshold changes too.

53. Unit economics

Unit economics describes the revenues and costs associated with a meaningful unit of business activity, such as one product, order or customer. The unit must be chosen deliberately. A subscription business may analyse one customer relationship; a physical product project may begin with one sold unit. The point is to ask whether growth in that unit creates or destroys economic value before fixed overhead and broader strategic effects.

Use positive unit contribution rather than “good unit economics” when a precise calculation is available. Kai Kai celebrates every sale even though the promotional discount is below variable cost. Alicia shows that more such sales increase the loss in the simplified model. That does not mean loss-leading strategies are never used professionally; it means the economic purpose and limits must be explicit. Students should understand the unit before interpreting growth.

54. Customer acquisition cost (CAC)

Customer acquisition cost, usually abbreviated CAC, is the cost attributed to acquiring new customers under a defined method and period. A simple educational model might divide specified acquisition spending by the number of new customers attributed to that campaign. The definition of cost and attribution matters: including only advertisement spend produces a different figure from including staff time, discounts and tools.

Write CAC for this campaign under this cost definition. Tricia avoids dividing by all customers when the numerator concerns spending used to acquire only new ones. She also distinguishes a customer from a website visit or lead. CAC becomes meaningful only when the numerator and denominator describe the same acquisition system. A low figure can be attractive, but not if the acquired customers generate insufficient contribution or do not remain customers.

55. Customer lifetime value (LTV)

Customer lifetime value, often abbreviated LTV or CLV, estimates the economic value generated by a customer relationship over the relevant duration, using a stated model. Professional calculations vary in sophistication. A simple school model can use expected contribution per purchase multiplied by expected number of purchases, while clearly labelling those values as assumptions.

Use estimated LTV under the model rather than presenting a forecast as a fact. Kai Kai multiplies one enthusiastic customer’s first-week purchases by five years. Tricia asks what evidence supports the purchase frequency and retention period. LTV is powerful precisely because it combines several uncertain assumptions. Changing any one of them can change the result substantially, so sensitivity analysis later in the manual becomes essential.

56. CAC-to-LTV relationship

The CAC-to-LTV relationship compares what it costs to acquire a customer with the economic value expected from that relationship. A simple ratio can be informative, but it is not a universal success score. The timing of cash flows, gross contribution, retention quality and uncertainty in the LTV model all matter.

Write LTV appears to exceed CAC under these assumptions rather than using a rigid ratio rule as a guarantee. Alicia tests a case where CAC is four dollars and estimated LTV contribution is twelve. The ratio is three to one, but if the twelve-dollar value arrives slowly while cash is needed now, the business may still face a financing problem. A ratio compresses several relationships into one number; advanced analysis reopens them when a decision depends on timing or uncertainty.

57. Payback period

The payback period is the time required for cumulative contribution or cash inflows associated with a customer or investment to recover the initial outlay under the model. In customer acquisition, it asks how long it takes for the contribution from an acquired customer to recover CAC. A shorter period can reduce cash pressure, but the calculation depends on retention and timing assumptions.

Use estimated payback period. Tricia models a four-dollar CAC and two dollars of monthly contribution. Ignoring churn and timing complexities, the simple payback is two months. If many customers leave after the first month, that model fails. Payback therefore connects acquisition economics to retention. The student should state the assumptions before celebrating the number.

58. Conversion rate

A conversion rate is the proportion of a defined group completing a defined target action. Ten purchases from one hundred qualified visitors gives a 10% visitor-to-purchase conversion rate under that definition. Ten purchases from fifty people who began checkout gives 20% checkout-to-purchase conversion. Both can be correct because the denominator differs.

Use conversion from A to B and name the stage. Kai Kai says “conversion is twenty per cent” without identifying the funnel step. Alicia makes the denominator explicit. This prevents different channels or periods from being compared unfairly. Conversion rates are powerful because they compress behaviour into a ratio, but the ratio is meaningful only when the start and end states are clear.

59. Retention

Retention describes the proportion of customers or users who remain active, continue purchasing or satisfy another defined continuation condition over a stated period. Retention is not the same as satisfaction. A person can like a product and stop using it because the need disappears; another can keep using a product because switching is difficult despite low satisfaction.

Write 30-day retention under this activity definition or another precise period. Tricia asks what counts as retained before calculating. Logging in once, purchasing again and keeping a subscription active are different measures. Retention affects LTV and payback, so an optimistic definition can make downstream unit economics appear stronger than the underlying behaviour warrants.

60. Churn

Churn measures customers, users or recurring revenue that leave or stop during a defined period, using a stated denominator. Customer churn and revenue churn can differ when customers pay different amounts. A project with no recurring relationship may not need a churn metric at all.

Use monthly customer churn or another clear basis. Alicia notes that retention and churn are related but not always exact complements across every professional definition, especially when customers can return, upgrade or downgrade. In a simple school model with one stable cohort and no reactivation, a 90% retained share implies 10% churn over the same interval. The advanced habit is to define the cohort and event before performing the subtraction.

Terms 61–80: growth metrics, cash and operations

61. Cohort

A cohort is a group of customers or users who share a defined starting event or period and are followed together over time. Customers acquired in September form a different cohort from customers acquired in October if the analysis aims to compare their later behaviour. Cohort analysis helps separate changes in the mix of customers from changes in how one group behaves as time passes.

Use September acquisition cohort or another precise label. Kai Kai compares all current users with last month’s new users and attributes the difference to product improvements. Tricia points out that the groups have different histories. Tracking cohorts makes retention and repeat purchase easier to interpret. A cohort is not automatically a representative sample; it is an organisational device for preserving a shared starting condition.

62. Funnel

A funnel is a model of stages through which potential customers move, such as awareness, visit, trial, purchase and repeat purchase. The funnel does not claim that every customer follows one simple path. It is a measurement structure that helps teams locate where large drop-offs occur and which conversion rate belongs to which transition.

Write funnel stage and stage-to-stage conversion. Alicia’s fictional event records two hundred visitors, eighty people who examine a sample, forty who try the planning task and twelve who place an order. The purchase rate is 6% of visitors but 30% of trial participants. Those rates answer different questions. A funnel helps reveal that the biggest problem may be getting visitors to try the product rather than persuading trial users to order.

63. Acquisition channel

An acquisition channel is a pathway through which a business reaches and gains new customers. Examples include referrals, search, events, partnerships or paid advertising. A channel is not simply where people can see a message; it becomes an acquisition channel when its role in bringing customers can be investigated.

Use customers acquired through the event channel. Tricia labels channels before comparing CAC because different channels have different costs, audiences and conversion patterns. A school event may produce low cash acquisition cost but require substantial time. A paid advertisement may reach more people but attract less qualified interest. Channel analysis should preserve the full cost and customer quality, not celebrate one cheap click metric in isolation.

64. Channel mix

A channel mix is the combination of channels used to reach, acquire or serve customers. Diversifying channels can reduce dependence on one source, but more channels also add coordination and measurement complexity. The best mix depends on target users, economics, capabilities and the stage of the venture.

Write change in channel mix when comparing overall metrics across periods. Kai Kai sees CAC rise and assumes every channel became less efficient. Alicia notices that the team shifted from mostly referrals to mostly paid promotion. The overall figure changed partly because the composition changed. Segmenting by channel reveals whether performance within each channel changed or the portfolio simply changed weight.

65. Referral

A referral occurs when an existing customer, user or partner directs another potential customer towards an offering. Referrals can reduce acquisition costs and carry trust, but they can also produce biased samples when people refer others similar to themselves. Referral volume therefore says something about spread, not necessarily representativeness.

Use referral source and distinguish organic referrals from rewarded programmes if the incentive changes behaviour. Alicia tracks whether a referred customer actually purchases rather than counting every shared link as acquisition. Kai Kai wants to pay for referrals before knowing contribution margin. Tricia checks unit economics first. A referral programme that rewards more than the acquired customer contributes can grow activity while worsening economics.

66. Average order value

Average order value, often abbreviated AOV, is the average revenue per order over a defined period: total order revenue divided by number of orders. It is not the same as revenue per customer when one customer can place several orders. The average can also hide a wide distribution of small and large baskets.

Write AOV increased from this period to that period and inspect why. A bundle can increase AOV while reducing order count or contribution margin. Kai Kai celebrates a higher AOV after offering a heavily discounted three-pack. Alicia calculates the contribution per order and discovers that the discount changed economics substantially. A larger basket is useful only when the additional revenue and costs support the objective.

67. Repeat purchase rate

Repeat purchase rate is the proportion of customers who make another qualifying purchase within a defined period or observation window. The denominator and eligibility rules matter. Customers who bought yesterday have had less time to repeat than customers who bought six months ago, so mixed-age cohorts can distort the comparison.

Use repeat purchase within ninety days or another stated window. Tricia compares mature cohorts rather than counting every recent buyer as a non-repeater. She also asks whether the product is naturally purchased often. A durable planner may not need monthly repurchase to succeed. Metrics should fit the product’s natural use cycle rather than force every business into a subscription-style pattern.

68. Cash flow

Cash flow describes money moving into and out of a business over time. Profit and cash flow are related but not identical. A sale can be recorded while payment arrives later; inventory may require cash before the related products are sold. A business can therefore appear profitable in an accounting model and still face a cash shortage.

Use cash inflow, cash outflow and cash timing. Alicia maps when the fictional team pays a supplier and when customers pay. Tricia notices that a large pre-order creates cash before production, while an invoice paid later creates the opposite timing. For a school project, the arithmetic can remain simple, but the timeline should be explicit. Cash cannot pay a bill before it arrives.

69. Burn rate

Burn rate is the rate at which a venture spends its available cash, commonly discussed for organisations operating before or while not generating enough cash to cover expenses. Definitions vary between gross spending and net cash decline, so the method must be stated. A negative cash balance trend is not made safer by using the word burn instead of spending.

Write monthly net cash burn under this simplified model. Kai Kai subtracts all expenses from zero and calls the result burn without recognising incoming cash. Tricia calculates starting cash, inflows and outflows by month. If the fictional project begins with six hundred dollars and loses one hundred net each month, the simple burn is one hundred per month. The number becomes useful when connected to runway and a plan for reducing uncertainty before cash runs out.

70. Runway

Runway estimates how long available cash can support the venture at the assumed rate of net cash use before additional financing or improved cash generation is required. A simple calculation divides cash by monthly burn. The result is highly sensitive to changing revenue, costs and one-off payments.

Use estimated runway at the current burn rate. Six hundred dollars divided by one hundred dollars of monthly burn suggests six months under the simplified steady-rate assumption. Alicia refuses to call that an expiry date. If revenue grows or costs rise, the runway changes. The metric is a planning clock: how much time does the team have to learn, change the model or stop responsibly?

71. Working capital

Working capital generally refers to current assets minus current liabilities, a measure of short-term financial resources available to support operations. In an advanced school example, the concept can be introduced through the practical gap between paying for inventory and receiving customer cash. Professional accounting definitions require proper classification of balance-sheet items.

Use working-capital requirement when describing cash tied up in the operating cycle. Tricia models a supplier requiring payment today while customer invoices are paid thirty days later. The unit can be profitable and still create a short-term cash need. This helps students see why margins alone do not settle operating viability. Timing is part of the business system.

72. Inventory turnover

Inventory turnover describes how frequently inventory is sold and replaced over a period, commonly calculated in professional accounting using cost of goods sold relative to average inventory. A high turnover can indicate fast movement, while an extremely high rate may coexist with stockouts. A low rate can indicate slow-moving inventory or intentional safety stock, depending on context.

Use inventory turns under the stated formula. A school exercise can compare days of stock on hand rather than imitate professional accounting precision without the needed records. Alicia notices that ordering one hundred kits reduces unit purchase price but leaves cash tied up if demand is uncertain. Inventory metrics connect purchasing economics with cash, demand and service level.

73. Stockout

A stockout occurs when inventory is unavailable when demand arrives. The consequence may include lost sales, delayed fulfilment, customer frustration or substitution. Avoiding every stockout by holding huge inventory can create different costs and risks. The decision is a trade-off, not a rule that more stock is always safer.

Use stockout rate only when its denominator is defined, such as unfilled demand events divided by total demand events. Kai Kai views a sold-out event as proof of success. Tricia asks how much unmet demand was lost and whether the shortage came from unexpectedly strong demand or under-ordering. A stockout can signal opportunity and operational weakness at the same time.

74. Lead time

Lead time is the elapsed time between initiating a process and receiving its result, such as ordering materials and receiving them, or receiving an order and completing delivery. Different lead times can exist within one supply chain. Average lead time can hide variability that matters for planning.

Use supplier lead time or order fulfilment lead time. Alicia initially reorders when stock becomes low without considering the supplier’s ten-day delay. Tricia calculates when inventory must be ordered to arrive before expected depletion. The concept turns time into an operating variable. A cheap supplier with unreliable long lead times may create more safety-stock need than a slightly more expensive nearby supplier.

75. Throughput

Throughput is the rate at which a process produces completed output over a defined period. In the fictional planner project, throughput might be complete kits packed per hour, not individual paper sheets cut. The output should correspond to the customer-ready or process-relevant unit.

Use system throughput, not the fastest station’s output, when evaluating the entire flow. Kai Kai reports that printing can handle sixty sets an hour and assumes the team can ship sixty kits an hour. Alicia finds that binding can complete only twenty. The system’s completed output is constrained elsewhere. Measuring one fast step without the rest of the process produces an attractive but unusable capacity claim.

76. Bottleneck

A bottleneck is the process step whose limited capacity or performance constrains the overall system under the current flow. Improving a non-bottleneck may create more waiting rather than more completed output. The bottleneck can move after an improvement, so diagnosis is iterative.

Use the current bottleneck rather than treating it as a permanent label. Tricia maps printing, binding, packing and payment processing. Binding has the lowest effective output and a queue forms before it. Adding another printer does not increase completed kits. Improving binding may. The term becomes powerful when it changes the investment decision from “speed up everything” to “find the constraint that actually limits the objective.”

77. Capacity

Capacity is the maximum sustainable output a resource or process can provide under defined conditions. Theoretical maximum, practical capacity and demonstrated output can differ. A person might bind one kit in two minutes for a short demonstration but not sustain thirty kits an hour for several hours without breaks, errors or setup time.

Use practical capacity and state the conditions. Alicia distinguishes one-station capacity from total-system capacity. If demand is below capacity, the constraint may be acquisition rather than operations. If demand exceeds capacity, long lead times and quality problems can appear. Capacity planning therefore connects market forecasts with operational evidence.

78. Utilisation

Utilisation is the proportion of available capacity actually used over a defined period. A machine used for eight hours out of ten available hours has 80% time utilisation under that simplified basis. Capacity utilisation can be measured differently depending on output, time and quality, so the definition must be stated.

Use high utilisation can reduce flexibility. Kai Kai wants every station operating at 100% continuously because idle time looks wasteful. Tricia points out that variation, rework and rush orders need some flexibility. In many systems, operating every resource at maximum apparent utilisation can create queues and delays. The educational lesson is not one universal target; it is that efficiency of each part and responsiveness of the whole system can conflict.

79. Quality control

Quality control consists of checks used to determine whether outputs meet defined requirements. It differs from quality assurance, which more broadly concerns the system designed to prevent defects and maintain quality. A final inspection can catch some wrong quantities or damaged products but may not reveal why defects occurred.

Use check against the specification. Alicia defines critical product requirements before packing. Tricia samples completed kits and records defects by type. Kai Kai initially rejects every imperfect item without tracking the pattern. The record turns quality control into learning: if most errors arise at one step, the process can be improved. A defect count becomes useful when connected to root cause, cost and customer impact.

80. Service level

A service level is a defined performance target for serving demand or customers, such as the proportion of orders fulfilled on time or the probability of avoiding a stockout under a specific inventory model. The phrase is broad and must be operationally defined. A 95% service level can mean different things in different systems.

Use on-time fulfilment service level when that is what is measured. Tricia sets a fictional target that at least nineteen of twenty orders be ready by the promised time. She then tracks whether the promise itself is realistic. A high service target can require extra inventory or capacity, increasing cost. Entrepreneurship requires deciding which service promise creates enough customer value to justify the resources needed to keep it.

Terms 81–100: forecasting, uncertainty, metrics and responsible decisions

81. Forecast

A forecast is an estimate of a future outcome based on stated information and assumptions. Forecasts can concern sales, cash, demand, staffing or inventory. They are not promises. A forecast becomes more useful when readers can see which assumptions drive it and how actual results compare afterward.

Use base forecast and specify the period. Kai Kai predicts one hundred sales because fifty people liked a post. Tricia builds a route from reachable audience to trial, conversion and capacity. Her forecast may still be wrong, but its logic is inspectable. Advanced forecasting means replacing one impressive total with a chain of assumptions that can each be tested and updated.

82. Scenario analysis

Scenario analysis compares several coherent sets of assumptions to explore how outcomes change under different plausible conditions. A base case, stronger-demand case and weaker-demand case can reveal which resources or risks become important. Scenarios are not probability statements unless probabilities are separately justified.

Write under the downside scenario rather than calling one scenario “what will happen.” Alicia models lower conversion together with higher material cost because those conditions can interact. A random collection of extreme numbers is not automatically a meaningful scenario. The assumptions should form a story about a possible operating environment and help the team prepare decisions before uncertainty becomes a crisis.

83. Sensitivity analysis

Sensitivity analysis changes one assumption at a time, or in a controlled set, to see how strongly an outcome responds. If break-even units change dramatically when variable cost moves by fifty cents, cost control may deserve more attention. If the result barely changes when package colour spending changes, that assumption may be less important economically.

Use the forecast is sensitive to conversion rate. Tricia keeps other assumptions fixed while adjusting one variable, making the relationship easier to interpret. This differs from scenario analysis, which often changes several assumptions together. Sensitivity analysis helps identify leverage and model fragility. A precise model does not become robust merely because it contains many decimal places.

84. Base case

A base case is the central set of assumptions used as the main reference for comparison. It should reflect the team’s best current evidence rather than the outcome they hope investors, judges or classmates will prefer. A base case can be revised when evidence improves.

Use base-case assumptions. Alicia chooses observed conversion from the most comparable test rather than the highest conversion ever recorded. Kai Kai objects that this makes the forecast look smaller. Tricia points out that a forecast’s purpose is planning, not motivation. A realistic base case creates a more useful comparison with upside and downside scenarios and makes later forecast error easier to diagnose.

85. Downside case

A downside case is a plausible set of less favourable assumptions used to examine resilience, cash needs and contingency actions. It is not the most catastrophic event imaginable. Useful downside cases focus on uncertainties capable of materially affecting the plan, such as lower conversion, higher cost or supplier delay.

Write downside case and state why its assumptions are plausible. Tricia tests whether the fictional project can still pay committed costs if sales are thirty per cent below the base forecast. The exercise does not predict failure. It asks what the team would do if a reasonable adverse condition occurs. That preparation can reveal a decision threshold before emotion and sunk costs interfere.

86. Expected value

Expected value is the probability-weighted average outcome across possible results when probabilities and values are meaningfully specified. A fifty per cent chance of gaining twenty units and a fifty per cent chance of losing ten produces an expected value of five units in the simplified calculation. Expected value is not the outcome that must occur.

Use expected value under the assumed probabilities. Kai Kai treats the five-unit result as a guaranteed profit. Alicia explains that the actual outcome could be twenty or minus ten in the two-outcome model. Risk tolerance, cash constraints and repeated opportunities can make decisions with the same expected value feel very different. Do not invent precise probabilities merely to make a calculation possible; acknowledge when they are uncertain.

87. Risk

Risk is exposure to uncertainty that can affect objectives. In entrepreneurship, risk may concern demand, cost, execution, reputation, safety, legal requirements or cash. Risk is not identical to uncertainty: uncertainty describes what is not known, while risk focuses on how uncertain events can affect valued outcomes.

Use identify the risk, likelihood, consequence and control. Tricia refuses to rank risks only by emotional vividness. A small supplier delay may be more probable than a dramatic event and therefore deserve attention. A school project should never create real safety or privacy risk merely to make the exercise realistic. Fictional cases can teach the reasoning without exposing students to harm.

88. Contingency

A contingency is a prepared response to a possible event or condition. A contingency plan identifies a trigger and action rather than saying vaguely that the team will adapt. “If the supplier confirms a delay beyond Friday, switch the classroom simulation to the backup material dataset” is a bounded contingency.

Use contingency trigger and fallback action. Alicia distinguishes contingency from panic purchasing. Preparation occurs before the event and preserves options. Good contingencies also recognise their own cost: keeping a backup supplier, extra inventory or spare capacity may reduce risk while increasing expense. The business decision balances resilience with resource use.

89. Decision rule

A decision rule states how evidence will lead to an action. For example: “If at least seven of ten target users complete the core task without help and no critical usability failure appears, proceed to the pricing test; otherwise revise the layout.” The rule prevents the team from changing the standard after seeing a disappointing result.

Use set the decision rule before the experiment. Kai Kai wants to continue regardless of the result because the prototype took days to design. Tricia points back to the agreed threshold. A decision rule need not be purely numeric; qualitative failure conditions can matter. The key is transparency about what evidence would cause continuation, revision or stopping.

90. Kill criterion

A kill criterion is a pre-defined condition strong enough to stop an experiment, feature or project path because continuing would no longer be justified. The phrase sounds severe, but its purpose is disciplined resource allocation. It protects teams from endless continuation driven by sunk cost or attachment.

Use stop criterion when a gentler phrase fits the classroom. Tricia sets one: if no interviewed participant reports the target problem after a diverse first research round, the group will pause product design and revisit the problem definition. A kill criterion should not reward superficial testing designed to end inconvenient ideas. It should identify evidence that genuinely undermines a critical assumption.

91. Key performance indicator (KPI)

A key performance indicator, or KPI, is a metric selected because it reflects performance on an important objective. Not every available number is key. A project can track page views, interviews, conversion, on-time fulfilment and defect rate, but only some may be central to the current objective.

Use KPI tied to the objective. Alicia chooses on-time fulfilment as a KPI for an operating phase where late orders are the main problem. Kai Kai prefers social impressions because the number is larger. Tricia asks which metric changes the decision. A KPI should focus attention. Tracking too many “key” indicators defeats the purpose and can encourage cherry-picking.

92. Leading indicator

A leading indicator is a metric that tends to change before a later outcome and may provide an earlier signal for action. Qualified trial completion might precede purchase conversion in a fictional funnel. Leading does not mean causal or guaranteed. Its usefulness depends on an established relationship with the outcome that matters.

Use possible leading indicator until evidence supports the relationship. Tricia examines whether users who complete the planning task are more likely to order later. One small cohort is suggestive, not conclusive. A good leading indicator gives time to intervene. A poor one creates false confidence because it is easy to measure but weakly connected to the final objective.

93. Lagging indicator

A lagging indicator reflects an outcome after the activities contributing to it have already occurred. Total monthly sales, refund rate or end-of-term profit can be lagging indicators. They are important for accountability but may arrive too late to diagnose what should change in the middle of a process.

Use leading and lagging indicators together where useful. Alicia tracks task completion during the funnel and later repeat purchases. The first can signal friction earlier; the second confirms whether customer behaviour persisted. A lagging metric is not inferior. It often measures the objective more directly. The planning question is whether the team also needs an earlier signal that can guide action before the final result arrives.

94. Vanity metric

A vanity metric is a number that looks impressive but provides little guidance for the decision being made because its connection to value, behaviour or causation is weak. Large follower counts, total impressions or cumulative downloads can become vanity metrics when they are reported without relevant denominators or outcomes.

Do not label every awareness metric useless. The same metric can be useful for one question and vanity for another. Kai Kai reports ten thousand impressions to prove customers love the product. Tricia asks how many qualified users tried it, purchased it and returned. Impressions may still matter for reach; they simply do not establish purchase satisfaction. The term should critique the metric–decision relationship, not mock large numbers automatically.

95. Actionable metric

An actionable metric provides information that can reasonably guide a decision because its meaning and relationship to an objective are sufficiently clear. Stage-specific conversion, defect rate by process step or retention by cohort can be actionable when the team knows what intervention each result might trigger.

Use actionable for this decision. A metric can be accurate without being actionable. Alicia knows total historical page views but cannot use them to decide whether the new onboarding step is confusing. She instead measures completion before and after a controlled change. The advanced move is to design the measurement around a decision rather than collect data first and search for a story afterward.

96. Experiment cost

Experiment cost is the money, time, attention and opportunity consumed by a test. An experiment should reduce enough important uncertainty to justify those resources. The cheapest test is not always best if it answers the wrong question; the most realistic test is not always necessary if a smaller test can eliminate a weak assumption first.

Use learning per unit of experiment cost as a planning idea, not a universal formula. Tricia chooses a paper prototype before producing twenty finished kits because layout usability is still uncertain. If the paper version fails, the team learns early. This approach reduces waste while preserving the ability to invest later when the remaining question genuinely requires a more complete product.

97. Learning velocity

Learning velocity describes how quickly a team converts important uncertainty into reliable learning and improved decisions. It is not simply the number of experiments completed. Ten superficial surveys can produce less learning than one well-designed test of the riskiest assumption.

Use the phrase qualitatively unless a project has a defensible measurement system. Alicia records assumptions closed, revised or replaced each week. Tricia also records whether the learning changed a decision. Kai Kai initially optimises for activity count. The group learns that speed matters only when the evidence is trustworthy enough to change what happens next. Fast confusion is not progress.

98. Ethical marketing

Ethical marketing communicates value without deceptive claims, hidden material information, exploitative pressure or inappropriate targeting. It requires the evidence behind a claim to fit the wording. A classroom enterprise should not say “proven to improve grades” because several students liked the planner or completed a task faster.

Use substantiate the marketing claim. Tricia rewrites “Never miss homework again” as “Designed to keep changing deadlines visible in one weekly view.” The second statement explains the intended function without promising an outcome the team cannot guarantee. Ethical marketing is not weak marketing. It is a constraint that forces creativity to remain inside the truth of what the evidence and product can reasonably support.

99. Privacy by design

Privacy by design means considering privacy needs and data minimisation while designing a process rather than adding protection only after unnecessary data have been collected. A student project studying planning habits may not need names, home addresses, exact grades or private message screenshots.

Use collect only what the research question needs. Alicia assigns anonymous participant codes for a fictional exercise and stores only the bounded observations used in analysis. Tricia asks who can access the key and when it should be removed under the school’s procedure. Privacy is part of research quality because participants answer differently when they fear unnecessary exposure, and harm can occur even when the analysis itself is mathematically correct.

100. Decision log

A decision log records an important decision, the evidence and assumptions available at the time, the alternative options considered and the condition that would trigger reconsideration. It prevents later memory from rewriting why the team chose a path. A decision that later proves wrong can still have been reasonable given the evidence then available.

Use record the decision and its evidence. Kai Kai writes only the outcome: “Use Version B.” Tricia expands the log: Version B met the task-completion rule, showed stronger conversion in the comparable test and remained within the unit-cost ceiling; the decision will be revisited if the next cohort’s completion rate falls below the threshold. The log closes the hundred-term collection by linking vocabulary to accountability. Business learning improves when choices remain traceable to what the team actually knew.

Advanced entrepreneurship investigation studio

The following ten packets are invented teaching cases. Their customers, prices, costs, conversion rates, suppliers and outcomes do not describe a real business. Each packet is designed to make several terms interact so that vocabulary changes the reasoning. Attempt the task before reading the worked discussion. The answer should state what the evidence supports, what it does not support, and which next action would reduce the most important uncertainty.

Use a four-line discipline in every case. Observation: what does the packet actually say happened? Calculation: what quantity follows from the supplied numbers? Interpretation: what does that quantity mean under the stated assumptions? Decision: what should the team do next, and what condition would change that decision? Keeping those lines separate prevents a neat calculation from becoming a stronger commercial claim than the dataset can support.

Investigation 1: twelve compliments and no discovery

Packet. Kai Kai shows a polished planner mock-up to twelve classmates during lunch. He begins every conversation by saying, “We spent weeks making this easier than ordinary planners. Don’t you think it looks useful?” Ten say yes, one says maybe and one says no. He then asks, “Would you use it?” Nine say yes. The team concludes that 75% of students have the target problem and that the product has problem–solution fit. No participant is asked how they currently plan, when they last forgot a task, what alternatives they use or whether the proposed feature would change an actual routine.

Task. Identify at least four research problems. Explain what the percentages do and do not measure. Rewrite the next interview as a discovery sequence that could change the team’s beliefs. Use leading question, stated preference, behavioural evidence, sample and problem–solution fit accurately.

Worked analysis. Ten of twelve is 83.3%, while nine of twelve is 75%. Those figures describe answers from this convenience sample under this interview script. They do not establish that 83.3% of all students find the product useful or that 75% would become users. The opening question is leading because it praises the product and suggests the expected answer. The second question collects stated intention without a real use situation, switching cost or price. The sample is also limited: lunchtime classmates available to speak with the creators may not represent the intended customer population.

The team also skipped the problem. Problem–solution fit requires evidence that a meaningful problem exists for a defined group and that the proposed approach appears capable of helping. A positive reaction to a mock-up supplies neither part strongly. A better sequence begins with behaviour: “Walk me through the last time you had several deadlines change in one week.” Then: “Where did you record the changes?” “What did you check first when deciding what to do?” “What went wrong, if anything?” “What did that cost you in time, stress or missed work?” Only after understanding the current process should the team show a prototype and observe whether it improves a realistic task.

Behavioural evidence. Give participants a fictional set of changing deadlines and ask them to plan the week using their normal method and then the prototype, with order balanced across participants. Record completion errors, time and requests for help. Such a task still does not prove long-term adoption, but it tests a more concrete proposition than “would you use it?” If the prototype reduces errors only after a long explanation, the team has learned something important about usability.

Decision. Do not proceed to a pricing claim based on these compliments. Run a discovery round focused on the problem and current substitutes. Predefine a stop criterion: if the team cannot find repeated evidence of the target planning problem among an appropriately varied group, pause the planner concept and revisit the problem definition. Alicia keeps the original compliments in the record; Tricia changes what they are allowed to mean. The evidence is not deleted. Its scope is repaired.

Investigation 2: useful at six dollars, admired at twelve

Packet. A fictional low-risk classroom choice experiment presents comparable participants with one of two price conditions for the same planner concept. At six dollars, twenty-four of forty participants choose the planner over a generic alternative and forty-six do not participate in the choice because they are outside the defined target group. At twelve dollars, twelve of forty comparable target participants choose the planner. A separate social poll shown only to followers asks whether the planner “looks worth twelve dollars”; seventy of one hundred respondents tap yes. No money is collected; the exercise uses tokens with an equal stated opportunity cost.

Task. Calculate the target-group choice rates at six and twelve dollars. Explain why the social poll is a different evidence type. Discuss price sensitivity without claiming a complete demand curve or real willingness to pay. Identify what a stronger next test would need to preserve ethically and methodologically.

Worked calculation. At six dollars, twenty-four of forty target participants choose the planner: 60%. At twelve dollars, twelve of forty choose it: 30%. The observed choice rate is lower under the higher price in these comparable classroom conditions. This pattern is consistent with price sensitivity. It does not establish that doubling the price always halves demand. Only two price points are observed, and the exercise uses tokens rather than actual personal spending.

The social poll produces a 70% positive stated response among followers who chose to answer. It differs in sample, context and consequence. “Looks worth twelve dollars” combines appearance with value judgement and involves no trade-off against an alternative. Treating the seventy positive taps as seventy customers would convert stated preference into behavioural demand. The poll can inform language or perception, but it is weak evidence for a real purchase rate.

Willingness to pay boundary. Real willingness to pay concerns a maximum exchange under actual conditions. A school project does not need to pressure minors into spending money to learn the principle. A teacher-approved token allocation can create a meaningful trade-off while remaining a simulation. The report should label it as simulated choice evidence. If the team later runs a real event under school rules, it can compare observed purchases with the earlier simulation and update its model.

Decision. Do not select six or twelve dollars from conversion alone. Combine the observed choice pattern with unit economics. A lower price may convert more customers while contributing too little to cover fixed costs. A higher price may produce stronger contribution but too few purchases. The next investigation should calculate contribution and break-even at both prices under explicit cost assumptions. Tricia writes the insight as a relationship: price changes customer choice and business economics simultaneously.

Investigation 3: a billion-dollar TAM that sells thirty kits

Packet. Kai Kai finds an online estimate that there are roughly one hundred million students in a broad international age group. He multiplies that number by a proposed ten-dollar price and writes “TAM = $1 billion; capturing only 0.01% gives $100,000 revenue.” The fictional project can currently serve only one school event with capacity for sixty kits. Based on past event attendance, the team expects three hundred visitors, of whom about one hundred and fifty fit the chosen target segment. In a comparable earlier test, 20% of target participants placed a simulated order at the proposed price.

Task. Separate TAM, SAM, SOM and a sales forecast. Build a bottom-up obtainable estimate for the single event. Explain why the global multiplication is not wrong arithmetic but weak decision evidence. Identify at least three assumptions that should appear in a market-size note.

Worked reasoning. One hundred million multiplied by ten dollars is indeed one billion dollars. But that number represents a highly abstract total-addressable-market construction only if the broad population, product relevance and price basis are defensible. It does not account for geography, access, competition, buying authority or the project’s actual capability. Taking 0.01% of a giant number does not create a pathway to customers.

For the event, the reachable serviceable audience is closer to the one hundred and fifty target visitors expected. Applying the 20% simulated order rate gives thirty expected orders under the assumption that the new event is comparable. Thirty kits at ten dollars produces a three-hundred-dollar revenue forecast, subject to capacity. Capacity is sixty, so the current forecast does not hit the operational ceiling. That thirty-order result is better described as an event forecast or bottom-up SOM estimate for the immediate test, not the project’s lifetime market.

Assumptions. Event attendance must resemble the reference event; the share of target users must be similar; the simulated 20% conversion must transfer to the new context; the ten-dollar price must remain; and supply must be available. Each assumption can fail independently. The estimate becomes educationally valuable because the learner can trace how thirty was produced. A global one-billion-dollar figure is harder to falsify at the scale of a classroom project.

Decision. Plan inventory around a range, not a single global story. The team might prepare a base forecast of thirty units, a lower scenario of eighteen and a higher scenario of forty-five while checking the cost of unsold stock and the cost of a stockout. Alicia sees that market sizing and inventory planning are connected by actual reachable demand. Tricia records the global TAM only as strategic context, clearly separated from the near-term obtainable forecast.

Investigation 4: lower price, higher sales, worse economics

Packet. The fictional planner has fixed event costs of ninety dollars. At the standard formulation, variable cost is four dollars per kit. Two price scenarios are considered. Scenario A sells at twelve dollars and expects thirty units. Scenario B sells at seven dollars and expects fifty units because the team believes lower price will increase conversion. Ignore taxes and other costs for this teaching model. The group announces that Scenario B is better because it sells twenty more units.

Task. Calculate contribution margin per unit, total contribution, break-even units and simplified profit after the fixed event cost for both scenarios. Explain why sales volume alone is a vanity-like metric for this decision. Identify the minimum evidence needed before trusting the fifty-unit forecast.

Scenario A. Contribution per unit is twelve minus four = eight dollars. Thirty units contribute 30 × 8 = $240. After the ninety-dollar fixed cost, simplified profit is $150. Break-even units are 90 ÷ 8 = 11.25, so at least twelve whole units are needed to cover the fixed cost in this simple model.

Scenario B. Contribution per unit is seven minus four = three dollars. Fifty units contribute 50 × 3 = $150. After the fixed cost, simplified profit is $60. Break-even units are 90 ÷ 3 = 30 units. Although B sells more units under the forecast, it produces less total contribution and lower simplified profit. The extra volume also consumes more capacity and inventory.

The comparison does not prove twelve dollars is the universal optimal price. Scenario A’s thirty-unit and Scenario B’s fifty-unit forecasts are assumptions. If real conversion at twelve dollars is much lower than expected or if variable cost changes with volume, the result changes. The lesson is that unit economics and demand must be evaluated together. “More units sold” is an activity metric; the decision concerns value created after relevant costs and capacity.

Next test. Run a teacher-approved comparable pricing experiment or use a supplied dataset that estimates conversion at several prices. Predefine the decision rule using both expected contribution and a minimum customer-value condition. Tricia also checks whether reducing price alters positioning: a lower price can affect perceived quality and the kinds of buyers attracted, not merely conversion. The final price decision needs evidence about customers and economics, not one spreadsheet column.

Investigation 5: CAC looks excellent until retention enters the model

Packet. Two fictional acquisition channels promote a refill subscription attached to the planner concept. Channel R is a referral programme that costs sixty dollars in rewards and produces thirty new customers. Channel A is a paid advertisement simulation costing one hundred dollars and producing twenty new customers. Each active customer is assumed to produce three dollars of contribution per month. After the first month, twenty-four of the thirty referral customers remain active, while eighteen of the twenty advertising customers remain active. After the second month, eighteen referral customers and sixteen advertising customers remain. The group compares only first-month CAC and declares R superior.

Task. Calculate initial CAC for both channels, first- and second-month cohort retention, and a simplified two-month contribution per acquired customer. Discuss why CAC alone does not settle channel quality. Do not project lifetime value beyond the supplied evidence without labelling assumptions.

CAC. Referral CAC = $60 ÷ 30 = $2 per acquired customer. Advertising CAC = $100 ÷ 20 = $5. On acquisition cost alone, referral is cheaper. Retention. Referral retention after month one = 24 ÷ 30 = 80%; after month two = 18 ÷ 30 = 60%. Advertising retention after month one = 18 ÷ 20 = 90%; after month two = 16 ÷ 20 = 80%. The advertising cohort is smaller and more expensive to acquire but retains better in this short fictional window.

Two-month contribution. If contribution is three dollars for each active customer in each month, referral produces 24 × 3 + 18 × 3 = $126 over months one and two. Dividing by the thirty acquired customers gives $4.20 of two-month contribution per acquired customer. Advertising produces 18 × 3 + 16 × 3 = $102, or $5.10 per acquired customer. Subtracting CAC as a simplified comparison yields $2.20 and $0.10 respectively over the observed two-month period. Referral still looks stronger on this narrow cumulative calculation, but the retention trajectories differ.

Projecting a full LTV from two months would require assumptions about future retention and contribution. Advertising could overtake if its customers remain much longer; referral could remain superior if later churn converges. The right conclusion is not “R wins forever.” It is “R has lower initial CAC and stronger two-month net contribution in the supplied data, while A shows higher retention and deserves continued cohort observation.”

Decision. Continue the channels under bounded tests if resources allow, while improving referral retention and monitoring whether advertising payback occurs. Tricia writes separate channel cohorts into the decision log. Kai Kai learns that a cheap customer is not automatically a valuable customer and an expensive customer is not automatically a bad acquisition. The relationship among CAC, retention, contribution and time determines the economics.

Investigation 6: ten thousand impressions hide the real funnel

Packet. A fictional campaign records 10,000 impressions, 2,000 reached accounts, 400 landing-page visits, 160 people who begin the planner task, 80 who complete it and 16 simulated orders. The team compares this campaign with an earlier smaller campaign that had 2,000 impressions, 800 reached accounts, 240 visits, 120 task starts, 84 completions and 21 simulated orders. Kai Kai announces that the new campaign is five times more successful because it has five times the impressions.

Task. Calculate reach-to-visit, visit-to-start, start-to-completion and visit-to-order conversion for both campaigns. Identify the stage where the newer campaign is weaker. Explain when impressions are useful and when they become a vanity metric.

New campaign. Reach-to-visit = 400 ÷ 2,000 = 20%. Visit-to-start = 160 ÷ 400 = 40%. Start-to-completion = 80 ÷ 160 = 50%. Visit-to-order = 16 ÷ 400 = 4%. Earlier campaign. Reach-to-visit = 240 ÷ 800 = 30%. Visit-to-start = 120 ÷ 240 = 50%. Start-to-completion = 84 ÷ 120 = 70%. Visit-to-order = 21 ÷ 240 = 8.75%.

The new campaign produces many more impressions and more visits, but its later conversion is weaker at every stated stage. It creates sixteen orders compared with twenty-one in the smaller campaign. The issue is not that impressions are meaningless. They describe distribution volume. If the objective is awareness, they may be relevant. They become vanity-like for a purchase decision when reported as success without the downstream actions the campaign is supposed to create.

The largest relative weakness in the new funnel appears in start-to-completion: 50% versus 70%, though reach-to-visit and visit-to-start also fall. The team should investigate whether the landing page attracts less qualified visitors, whether the task is harder or whether the campaign promise creates a mismatch with the actual experience. A single total cannot distinguish these explanations.

Decision. Keep distribution metrics as leading context, but make stage-specific conversion actionable. Tricia assigns one owner to investigate completion friction before buying more reach. Alicia compares the messages used in both campaigns. Kai Kai learns that amplification can magnify a weak funnel as easily as a strong one. More people entering the top does not guarantee more useful outcomes at the bottom.

Investigation 7: retention improves even while the overall rate falls

Packet. A project has two customer cohorts. January starts with twenty customers and retains sixteen after one month and twelve after two months. February starts with eighty customers and retains sixty after one month. At the end of February, the dashboard reports seventy-two active retained customers out of one hundred acquired customers in total and declares “retention fell to 72%, therefore the product became worse.” The team has mixed customers of different ages into one denominator.

Task. Calculate January month-one and month-two retention and February month-one retention. Explain why the 72% overall figure does not measure the same stage for every customer. Discuss how cohort analysis changes the interpretation.

January. Month-one retention = 16 ÷ 20 = 80%. Month-two retention = 12 ÷ 20 = 60%. February. Month-one retention = 60 ÷ 80 = 75%. Comparing month-one with month-one shows February slightly lower than January: 75% versus 80%. The combined seventy-two active customers consist of twelve January customers at month two plus sixty February customers at month one. Dividing 72 by 100 mixes different lifecycle ages.

The mixed overall figure may still describe the proportion of all acquired customers active at one calendar date, if that is the business question. It does not answer whether month-one retention improved or whether product changes caused a difference. Cohort analysis preserves the starting month and allows equal-age comparisons. The result in this packet is not “retention improved.” It is that February month-one retention is five percentage points lower than January month-one retention, while later February retention remains unknown.

Decision. Wait for comparable later data while investigating possible reasons for the month-one difference. Check acquisition-channel mix: perhaps February included a new channel with different customer intent. Check product changes and onboarding. Do not diagnose a cause from the retention table alone. The cohort metric identifies where to ask a question; it does not answer the causal question automatically.

Alicia redraws the dashboard by cohort and age. Tricia adds both calendar-active and cohort-retention views because they answer different operating questions. Kai Kai stops demanding one “true retention number.” Advanced metrics work when each number has a defined job.

Investigation 8: profitable on paper, out of cash on Tuesday

Packet. A fictional school-enterprise supplier requires payment of $400 for materials on 1 October. The team begins the month with $250 cash. It has already secured orders worth $720 with variable production cost included in the $400 material payment. Customers pay on 20 October. Other fixed cash expenses of $100 are due on 10 October. Ignore accounting details outside this cash-timing model. Kai Kai says the project is profitable because $720 revenue exceeds $500 of stated costs, so there is no financial problem.

Task. Build a simple dated cash sequence. Explain the difference between simplified profit and cash availability. Calculate the minimum additional temporary cash needed to meet the stated payments before customer receipts, assuming no other cash flows.

Cash sequence. Starting cash is $250. On 1 October, the $400 supplier payment would move cash to −$150 if made without additional funds. On 10 October, another $100 is due, taking the cumulative gap to −$250. Customer cash of $720 arrives only on 20 October. Therefore the simplified peak cash shortfall is $250 before receipts, assuming all payments occur as stated and no credit terms exist.

The total-period model shows $720 revenue against $500 of stated costs, a $220 simplified surplus. Yet the team cannot pay $500 of early obligations from $250 of starting cash. Profitability over a period and liquidity at a particular date are different questions. Working capital concerns resources needed to bridge this timing gap.

Decision options. In a real business, options might include negotiating payment terms, requesting deposits, changing order timing or arranging appropriate financing. This classroom manual does not recommend a financial product. The educational decision is to redesign the cash plan before committing. A project should not sign or promise obligations it cannot responsibly meet.

Forecast lesson. A monthly table with one revenue total and one cost total would hide the problem. Tricia changes the model to dated inflows and outflows. Alicia adds a downside case where some customer payments arrive late. Kai Kai learns why runway and cash timing are operational metrics, not gloomy distractions from a profitable idea.

Investigation 9: the faster printer that does not increase output

Packet. A fictional four-step production line handles planner kits. Printing can process sixty kits per hour. Cutting can process forty. Binding can process twenty. Packing can process thirty. Demand is expected at twenty-five kits per hour during a short peak. Kai Kai proposes buying access to a faster printer that can handle one hundred kits per hour because printing has the most visible queue of paper. The upgrade costs resources that could instead improve binding to thirty kits per hour.

Task. Identify current system capacity and bottleneck under the simplified assumption that every kit must pass through all four steps and no other losses occur. Compare the two proposed upgrades. Discuss utilisation and queue formation when demand is twenty-five per hour.

Current capacity. The slowest required step is binding at twenty kits per hour, so the line cannot sustainably produce more than twenty completed kits per hour under the simplified serial-process model. The bottleneck is binding. Increasing printer capacity from sixty to one hundred does not change the twenty-kit system capacity because binding remains the constraint. Increasing binding to thirty makes packing, at thirty, the next joint constraint; system capacity becomes thirty kits per hour.

At expected demand of twenty-five per hour, the current system is overloaded because demand exceeds the twenty-unit capacity. A queue will accumulate before or at the bottleneck unless demand is smoothed, work is delayed or another capacity response occurs. If binding capacity rises to thirty, expected demand uses about 83.3% of that step’s capacity and packing also sits at 83.3%, creating some headroom. Printing at sixty operates far below its theoretical maximum for this flow, so making it faster does not address the current output constraint.

Decision. Improve binding first if the objective is completed throughput and the simplified model is accurate. Then remeasure the line because the bottleneck moves. Tricia also checks quality: if faster binding doubles defect rate, nominal capacity does not equal usable throughput. Alicia asks whether the demand forecast itself is reliable. Operations should not be optimised around an untested market forecast without a plan for revising capacity.

Vocabulary transfer. This case shows why local efficiency can conflict with system performance. Printing is already faster than the line needs. A metric reporting printer utilisation could make idle capacity look wasteful even when that spare capacity is harmless. The decision should be based on system throughput, queue and customer promise, not on making every resource equally busy.

Investigation 10: the forecast that changes when one assumption moves

Packet. A simplified fictional forecast assumes 1,000 qualified visits, 8% purchase conversion, a twelve-dollar price, five-dollar variable cost per unit and $300 fixed cost. Base-case expected orders are therefore eighty. The team wants to know which assumption deserves the next experiment. Three plausible alternative values are supplied one at a time: visits fall to 800, conversion falls to 5%, or variable cost rises to seven dollars. Hold all other base assumptions constant for each sensitivity check.

Task. Calculate base-case contribution after fixed cost. Then calculate the result under each one-variable change. Compare the magnitude of each change and identify which assumption appears most economically sensitive within the supplied ranges. Explain why sensitivity is not the same as probability.

Base case. Orders = 1,000 × 8% = 80. Contribution per unit = $12 − $5 = $7. Total contribution = 80 × 7 = $560. After $300 fixed cost, the simplified result is $260.

Visits at 800. Orders = 800 × 8% = 64. Contribution = 64 × 7 = $448. After fixed cost = $148. The result falls by $112. Conversion at 5%. Orders = 1,000 × 5% = 50. Contribution = 50 × 7 = $350. After fixed cost = $50. The result falls by $210. Variable cost at seven dollars. Orders remain eighty, contribution per unit becomes $5, total contribution $400, and the result after fixed cost is $100. The fall is $160.

Within these supplied ranges, the result is most sensitive to the conversion-rate change, followed by the cost change and then the visit change. This does not mean conversion is most likely to deteriorate. Sensitivity concerns effect size when an assumption changes; risk also needs the plausibility of that change. If the variable-cost increase is almost certain while the conversion drop is unlikely, the risk priority might differ.

Decision. The next experiment should consider both uncertainty and sensitivity. If conversion is highly uncertain as well as highly influential, it is a strong candidate for testing. If conversion is already well established while supplier cost is poorly understood, cost evidence may deserve priority. Tricia updates the assumption register rather than letting one sensitivity table dictate the entire plan. Alicia creates a downside scenario combining weaker conversion and higher cost to inspect cash resilience. Kai Kai sees that a forecast is not one number; it is a map of relationships.

Business report workshop: from pitch language to decision language

Weak draft. “Our planner is a huge success because 75% of students said they would use it. The market is worth one billion dollars, and we only need a tiny share. Our campaign reached ten thousand impressions and therefore proved demand. The seven-dollar price sells more units, so it creates more profit. Referrals give us cheaper customers, making them the best channel. We can easily scale because our printer handles sixty kits per hour. The business is low risk and should launch immediately.”

The paragraph contains several individually plausible facts joined into unsupported conclusions. Seventy-five per cent refers to stated intention in one leading-question sample. The billion-dollar figure is an abstract TAM, not reachable revenue. Impressions measure distribution, not demand. Seven-dollar pricing produces more forecast units but less contribution in the supplied case. Referrals have lower CAC but weaker observed retention. Printer capacity does not determine line throughput because binding is the bottleneck. “Low risk” ignores uncertainty in conversion, cost, capacity and cash timing.

Step 1: restore the unit of evidence. Replace “students” with “nine of twelve classmates in the initial interview.” Replace “market” with the relevant global TAM estimate or local obtainable event forecast. Replace “campaign success” with the actual funnel stages. The nouns in the sentence control the claim’s scope before any hedge is added.

Step 2: restore the denominator. Every percentage should tell the reader what it divides. Stated use intention is nine of twelve interviewees. New-campaign visit-to-order conversion is sixteen of four hundred visits. Referral CAC is sixty dollars divided by thirty acquired customers. Once the denominator is visible, several dramatic claims become naturally narrower.

Step 3: restore time. Retention belongs to a cohort age. Cash belongs to a date. Payback belongs to a period. Inventory and capacity belong to an operating window. “Profitable” without time can hide a cash shortage before receipts arrive. “Retained” without a cohort age can combine customers who have had unequal opportunity to leave.

Step 4: restore the decision criterion. More sales are not automatically the objective if contribution falls below the level needed to cover fixed cost. Higher utilisation is not automatically desirable if queues become unstable. More interviews are not automatically better if the sample remains narrow and questions remain leading. The criterion should match the actual decision: customer learning, economic contribution, service performance or another explicitly stated objective.

Revised model. “The current evidence supports continued testing rather than immediate launch. Nine of twelve classmates in the initial interview said they would use the planner, but the questions were leading and measured stated intention rather than adoption. A later simulated price test showed lower choice at twelve dollars than six, while the unit-economics model produced stronger contribution at twelve under the supplied sales forecasts. The most realistic event forecast is approximately thirty orders, far below the global TAM calculation and below current capacity after the binding constraint is improved. Referral acquisition is cheaper in the observed cohort, while paid acquisition retains better over two months. The next decision should therefore test conversion and retention with a cleaner design, confirm supplier cost and complete a dated cash plan before any real commitment.”

The revised report is not anti-entrepreneurial. It still identifies an opportunity and a path forward. Its confidence is selective. It treats customer enthusiasm as a research signal, pricing as a joint market-and-economics problem, market sizing as a hierarchy and operations as a system. A reader can see what would change the recommendation. That traceability is more useful than a pitch that sounds certain but gives the team no way to recognise when it is wrong.

Independent assessment: the StudyStack launch decision

The entire packet below is fictional. It is designed to test whether the learner can integrate customer evidence, unit economics, funnel metrics, operations and uncertainty without being told which vocabulary term to use first. Attempt the questions before reading the answer discussion. The strongest responses will make several correct distinctions in ordinary language and then introduce technical terms where those terms genuinely sharpen the explanation.

Fictional brief. StudyStack is a paper-based planning kit for students whose weekly schedules change often. The team defines its initial beachhead market as students participating in at least two recurring activities outside ordinary lessons and who report using at least two separate planning tools. The prototype contains a weekly board, removable task cards and a deadline-change strip. The team has not launched publicly. It has permission only to analyse the supplied classroom data and make a recommendation for the next supervised test.

Discovery packet. Twenty-four target participants complete neutral problem interviews. Eighteen describe at least one recent occasion in which a changed deadline or activity schedule required them to check multiple sources. Four say their current system works well enough that the problem is minor. Two report no relevant recent problem. Of the eighteen who report the problem, fourteen describe a workaround involving screenshots, message searches or rewriting deadlines. Separately, a broad social poll of two hundred followers asks, “Would you love one perfect planner for everything?” and receives one hundred and sixty yes responses.

Prototype packet. Twelve of the eighteen problem-positive participants use their usual method on a fictional changing-deadline task and then use StudyStack on a comparable task with order balanced across participants. Under the usual method, the group records twenty-seven missed or misplaced task updates across all trials. Under StudyStack, it records eleven. Nine of the twelve complete the StudyStack task without facilitator help. Three require explanation of the removable-card system. The exercise does not measure long-term use or academic results.

Price packet. In a separate token-based simulation with sixty comparable target participants, thirty are randomly assigned a six-dollar equivalent condition and thirty a ten-dollar condition. At six dollars, eighteen choose StudyStack over the generic alternative. At ten dollars, twelve choose it. The tokens have a stated opportunity cost inside the exercise but are not real money. The team therefore labels the result simulated price-choice evidence rather than actual willingness to pay.

Economics packet. Variable cost is estimated at four dollars per complete kit under the current small-batch process. Fixed event and setup cost is ninety-six dollars. At a six-dollar selling price, contribution per unit is two dollars. At ten dollars, contribution per unit is six dollars. The team forecasts forty-five sales at six dollars or thirty sales at ten dollars during the proposed event. The event can reach approximately two hundred target visitors under the base case. The forecasts are derived from earlier simulations and are not guaranteed.

Funnel packet. A previous supervised simulation produced 1,200 impressions, 600 reached accounts, 180 landing-page visits, 90 task starts, 63 task completions and 18 simulated orders. A proposed paid channel would cost seventy-two dollars under the classroom model. A referral simulation would cost thirty dollars in rewards and is expected to produce ten new customers. No retention data exist yet because StudyStack is currently modelled as a one-time kit rather than a recurring subscription.

Operations packet. Printing capacity is sixty kits per hour, cutting forty, binding twenty-four and packing thirty. The event forecast at the ten-dollar price is thirty units over a two-hour production window. A supplier quote lowers variable material cost by fifty cents per kit if the team orders one hundred units in advance, but unsold units cannot be returned in the model. The supplier requires payment before the event. Starting project cash is three hundred dollars.

Decision questions. 1) Which discovery evidence is stronger for the problem hypothesis: the social poll or the neutral interviews, and why? 2) Calculate the share of interviewed target participants reporting the problem. 3) What does the prototype task support, and what does it not support? 4) Calculate simulated choice rates at six and ten dollars. 5) Calculate contribution margin per unit at both prices. 6) Calculate break-even units for each price. 7) Calculate simplified event contribution after fixed cost under both forecast volumes. 8) Calculate visit-to-start, start-to-completion and visit-to-order conversion in the funnel. 9) Calculate the proposed paid-channel CAC if eighteen orders resulted and the referral CAC if ten customers resulted. 10) Explain why comparing those CAC figures alone is incomplete. 11) Identify the current bottleneck and two-hour practical system capacity under the simple serial-process assumption. 12) Evaluate the one-hundred-unit supplier discount. 13) Identify the most important missing evidence before a real launch. 14) Write a decision rule for the next test. 15) Write a 250–350-word recommendation.

Answer discussion: discovery and prototype evidence

The neutral interviews are better matched to the problem hypothesis because they ask target participants about recent behaviour and current workarounds. Eighteen of twenty-four report a relevant recent problem: 75% of this interviewed target sample. That percentage must not be silently converted into “75% of all students.” The sample is defined by the project’s target criteria and contains only twenty-four people. The social poll receives 80% positive responses, but the question is leading, hypothetical and shown to a broad follower group that may not match the target segment. It is weaker evidence for problem prevalence.

The prototype task records twenty-seven errors under the usual methods and eleven under StudyStack across comparable fictional tasks. That is sixteen fewer recorded errors in the supplied set, a reduction of about 59.3% relative to twenty-seven. The exercise also shows that nine of twelve participants complete the StudyStack task without help while three need explanation. These observations support further usability investigation and a hypothesis that the prototype can improve task handling under the exercise conditions. They do not establish improved grades, long-term adoption, general student productivity or willingness to pay.

The three participants needing explanation are especially useful. The team should not delete them as outliers merely because most participants succeeded. The confusion may identify a feature boundary or onboarding need. A next iteration could simplify the removable-card interaction and predefine a task-completion criterion. If performance improves without adding facilitator help, the result would support the design change more directly.

Answer discussion: pricing, contribution and break-even

At six dollars, eighteen of thirty participants choose StudyStack, giving a 60% simulated choice rate. At ten dollars, twelve of thirty choose it, giving 40%. The observed simulation is consistent with price sensitivity. It does not establish a complete demand curve or actual market willingness to pay because the exercise uses tokens and only two prices.

Contribution per unit at six dollars is $6 − $4 = $2. At ten dollars it is $10 − $4 = $6. Break-even units are $96 ÷ $2 = 48 units at six dollars and $96 ÷ $6 = 16 units at ten dollars. Under the forecast volumes, six-dollar sales of forty-five units generate $90 total contribution, which is six dollars below the fixed cost. The simplified event result is therefore −$6. Ten-dollar sales of thirty units generate $180 contribution; after the $96 fixed cost, the simplified result is $84.

This does not prove ten dollars is the correct launch price. The forty-five- and thirty-unit forecasts may be wrong, and the simulated choice rates come from a different context. It does show that the lower-price model is fragile under the supplied forecast because its contribution is too small to cover fixed cost. The next pricing test should therefore examine both conversion and contribution rather than maximise choice rate alone.

Answer discussion: funnel and acquisition

Visit-to-start conversion is 90 ÷ 180 = 50%. Start-to-completion is 63 ÷ 90 = 70%. Visit-to-order is 18 ÷ 180 = 10%. Impressions and reached accounts describe upper-funnel distribution but should not be reported as customers. If the proposed paid channel costs seventy-two dollars and eighteen new customers are attributed under the model, CAC is $4. The referral simulation costs thirty dollars and produces ten customers, giving CAC of $3.

Referral CAC is lower by one dollar in the supplied model, but CAC alone is not enough. The project needs customer quality, repeat or referral behaviour where relevant, channel capacity and contribution after acquisition cost. Because StudyStack is modelled as a one-time purchase, an LTV calculation should not pretend a recurring stream exists. A simple customer contribution after acquisition might subtract the CAC from the unit contribution while acknowledging channel-specific costs and refunds. The team should not manufacture retention just because the vocabulary list contains retention.

At the ten-dollar price and four-dollar variable cost, six dollars of unit contribution minus a four-dollar paid-channel CAC leaves two dollars before fixed cost and other excluded expenses under the simple attribution model. Referral leaves three dollars after its three-dollar CAC. Those figures show why acquisition spending interacts with pricing. They still do not include every business cost or prove the channel can scale at the same CAC.

Answer discussion: operations and inventory

Binding at twenty-four kits per hour is the bottleneck. Under the simple serial-process assumption, system capacity is twenty-four finished kits per hour. Over two hours, theoretical system capacity is forty-eight kits, before breaks, defects or setup losses. The thirty-unit ten-dollar forecast therefore fits within the nominal two-hour capacity, while leaving limited but useful headroom. Buying a faster printer would not raise that system capacity because printing is already far above the binding rate.

The supplier discount reduces variable material cost by fifty cents per kit if one hundred units are ordered. The maximum direct savings compared with the current cost is fifty dollars across all one hundred units. But the base ten-dollar forecast is only thirty sales. Buying one hundred units means cash and inventory are committed for seventy forecast-unsold units. If the full current four-dollar variable cost represented material cost for this simplified comparison, one hundred discounted units would require approximately $350 instead of $400, already exceeding the stated $300 starting cash before fixed event costs. The actual packet describes variable cost more broadly, so the precise supplier payment requires a cost breakdown not provided. That missing detail itself is part of the answer.

The discount therefore should not be accepted merely because unit cost falls. The team needs the supplier-payment amount, demand range, storage or obsolescence risk and cash timing. A smaller order at higher unit cost may be economically sensible when demand uncertainty is large. Inventory decisions should optimise the system, not the purchase-price line in isolation.

Answer discussion: decision rule and recommendation

The most important missing evidence depends on which decision the team plans to make. For a real launch decision, simulated price choice should be strengthened, the production cost breakdown confirmed, the three usability failures repaired and the cash requirement mapped. A strong next test could focus on the riskiest assumption that customers will choose the product at a price producing sustainable contribution.

Example decision rule. “Proceed to a supervised small event test if at least eight of ten comparable target participants complete the revised core task without facilitator help, and if at least four of ten choose StudyStack at the ten-dollar-equivalent price in the pre-defined simulation while the confirmed variable cost remains no more than four dollars per kit. If either usability or economics criterion fails, revise before committing inventory.” This is an example teaching rule, not a universal business threshold. Its advantage is that it names the evidence that would change the next action.

Model recommendation. “StudyStack has enough evidence to justify another supervised test but not a full launch. Eighteen of twenty-four target interviewees described a recent scheduling problem, and the prototype exercise recorded fewer task-update errors than participants’ usual methods. However, three of twelve prototype users required explanation, so the interaction still needs repair. The price simulation shows lower choice at ten dollars than six, but the ten-dollar model produces a six-dollar unit contribution and a sixteen-unit break-even threshold, compared with two dollars and forty-eight units at six dollars. Under the supplied event forecasts, the ten-dollar scenario covers the fixed cost while the six-dollar scenario does not. The current production bottleneck is binding at twenty-four kits per hour, which is sufficient for the thirty-unit base forecast over two hours. The proposed one-hundred-unit supplier discount should not be accepted until the team confirms the cash requirement and downside demand because the order greatly exceeds the base forecast. The next step should test the revised prototype and ten-dollar-equivalent choice under a pre-defined decision rule, then update the cash and inventory plan using actual results.”

Fifty retrieval and transfer challenges

1. What is customer discovery, and how is it different from asking customers to approve a finished idea? 2. Give one problem-interview question anchored in a recent event. 3. Rewrite a leading question about a product into a neutral question. 4. Distinguish stated preference from behavioural evidence. 5. What makes a hypothesis falsifiable? 6. Why can a large sample still be biased? 7. Distinguish sampling bias from response bias. 8. What turns an observation into a customer insight? 9. What evidence would support problem–solution fit but not product–market fit? 10. Give one example of a switching cost that is not money.

11. Why can an early adopter be useful but unrepresentative of later customers? 12. Distinguish beachhead market from SAM. 13. What makes differentiation meaningful rather than merely different? 14. Give one substitute outside the apparent product category. 15. Why does a competitor sometimes support rather than destroy an opportunity hypothesis? 16. Explain TAM without turning it into a sales forecast. 17. Build a bottom-up SOM from reach, conversion and capacity. 18. What denominator belongs in market share by revenue? 19. Why is social interest not the same as demand? 20. What would be needed to interpret price sensitivity fairly?

21. Distinguish business model from revenue model. 22. Give a fixed cost that can change when scale changes. 23. Why can marginal cost differ from average variable cost? 24. Explain sunk cost without saying past spending is always irrelevant. 25. Give one opportunity cost in a school enterprise. 26. Calculate contribution on a twelve-dollar price and five-dollar variable cost. 27. Explain why a 50% markup is not a 50% margin. 28. Calculate break-even when fixed costs are eighty dollars and contribution is four dollars per unit. 29. Choose the appropriate unit for unit economics in a physical product versus subscription. 30. Explain why CAC requires a clear definition of acquisition cost and new customer.

31. Why should LTV be labelled as an estimate? 32. Explain why a favourable CAC-to-LTV relationship can coexist with cash pressure. 33. Calculate simple payback on six-dollar CAC and two dollars monthly contribution. 34. Define conversion from one funnel stage to another. 35. Why can retention and satisfaction move differently? 36. When might churn be an irrelevant metric? 37. Why analyse cohorts of equal age? 38. Give an example of a channel-mix change that alters aggregate CAC. 39. Why can AOV rise while profit falls? 40. What natural purchase cycle should be considered before judging repeat purchase?

41. Explain how a profitable project can run out of cash. 42. What is runway, and which assumption can make it change immediately? 43. Why can buying more inventory improve unit cost while worsening working capital? 44. Identify the bottleneck in a four-step process with capacities 60, 45, 22 and 30. 45. Why does improving a non-bottleneck sometimes fail to raise throughput? 46. What is the difference between a forecast and a scenario? 47. What does sensitivity analysis tell you that probability does not? 48. Write a decision rule with a measurable trigger. 49. Convert an impressive vanity metric into an actionable metric for the same objective. 50. What should appear in a decision log so a future reviewer can reconstruct the reasoning?

A six-week advanced entrepreneurship vocabulary sequence

Week 1: discovery before solution. Use Terms 1–20 with one safe everyday problem. Students conduct or analyse fictional problem interviews, identify leading wording and distinguish stated from behavioural evidence. Finish the week with an assumption register: belief, current evidence, consequence if wrong and next test. The learner should be able to explain why compliments do not establish adoption.

Week 2: value and market boundaries. Use Terms 21–40. Write one value proposition, identify the customer’s job, current substitute and switching cost, then build a market hierarchy from TAM to an immediate obtainable test. Compare two competitors without declaring an overall winner. Finish by explaining which meaningful differentiation the target user appears to value and what evidence still needs collecting.

Week 3: unit economics. Use Terms 41–60. Build a simplified price–cost model with fixed cost, variable cost, contribution, markup, margin and break-even. Introduce CAC only after defining what counts as a new customer and which acquisition costs are included. Model LTV as an estimate with visible assumptions rather than a fact. Change one assumption and explain the result in a sentence.

Week 4: growth and operations. Use Terms 61–80. Build a small funnel, calculate stage conversions, compare channels and track a cohort. Then map a four-step operation and identify the bottleneck. Add a dated cash timeline and one inventory choice. The goal is to see a business as a connected system where marketing, cash and operations constrain one another.

Week 5: uncertainty and metrics. Use Terms 81–100. Build a base case, downside case and one-variable sensitivity table. Distinguish a KPI from every number on the dashboard and a leading indicator from a lagging outcome. Write one decision rule and one stop criterion before reading the fictional result. End with an ethical-marketing rewrite and a privacy-by-design check.

Week 6: integrated decision. Complete the StudyStack independent assessment with no glossary visible at first. Require the learner to calculate, interpret and recommend. Afterwards, reopen the terms and identify which labels improved the explanation. The final product is a two-page evidence memo: decision, evidence, economics, operational constraint, uncertainty, next experiment and decision log entry.

Between sessions, retrieve a small number of terms from memory and create a new example in a different enterprise context. A café queue, club event, tutoring resource, digital tool or fictional delivery service can all test transfer without requiring a real business. The article does not promise mastery after six weeks. Pace should follow the learner’s ability to distinguish, calculate and transfer the concepts.

Assessment rubric: from word recognition to business judgement

Level 1 — Recognition. The learner matches a term to a familiar definition or example. They know that CAC concerns acquisition cost and that break-even concerns revenue and cost. This is useful but insufficient for advanced performance.

Level 2 — Boundary. The learner distinguishes near concepts: stated versus revealed preference, markup versus margin, profit versus cash, throughput versus capacity, TAM versus SOM, and forecast versus scenario. They can explain why the words cannot be swapped without changing the meaning.

Level 3 — Calculation and application. The learner uses the right numerator, denominator and unit in a new packet. They connect the calculation to a sentence that states what the result describes. A correct number with a wrong interpretation does not meet this level.

Level 4 — Diagnosis. The learner identifies why evidence is weak or a system underperforms. They can recognise a leading question, an unrealistic market-size shortcut, a weak denominator, a cash-timing problem or a bottleneck. Their diagnosis suggests a test or process change.

Level 5 — Decision and transfer. The learner integrates customer evidence, economics, operations and uncertainty into a bounded recommendation. They specify what would change the decision, protect privacy, avoid unsupported marketing claims and transfer the reasoning into an unfamiliar enterprise case.

Teacher and parent guidance

Use fictional cases freely when real enterprise activity would create unnecessary pressure, spending, privacy collection or safety issues. The reasoning does not require students to operate a company. A well-designed dataset can teach customer discovery, margins, cash flow and decision rules without asking a child to negotiate contracts or expose personal finances.

When a learner makes a numerical mistake, determine whether the problem is arithmetic or concept selection. A student who divides acquisition spending by website visits may calculate perfectly while answering the wrong question. Ask them to name the numerator, denominator and unit before redoing the arithmetic. This often repairs the conceptual layer more efficiently than another page of calculations.

When a learner uses a business term as decoration, remove the term temporarily. Ask for the idea in plain language. If the reasoning is correct, restore the technical word and teach its natural collocations. If the reasoning is wrong, return to the evidence relationship. Vocabulary should compress an understood idea, not hide an unclear one.

Parents can support entrepreneurial thinking without turning every hobby into a commercial project. Ask: “What problem are you trying to solve?” “How do you know it matters?” “What are people doing now?” “What is the cheapest safe way to learn whether your assumption is wrong?” “What would make you stop or change the idea?” These questions build agency and evidence habits whether or not the learner ever starts a business.

Frequently asked questions

Is this an official Secondary 2 or Grade 8 entrepreneurship syllabus? No. It is an advanced interdisciplinary vocabulary collection. School programmes differ in whether they teach enterprise, economics, business studies, design, personal finance or project work at this stage.

Do students need to memorise all one hundred terms? No. Learn the terms required by the current problem, retrieve them later and transfer them into new cases. Depth and correct boundaries matter more than being able to recite the list from 1 to 100.

Why does the manual use so much arithmetic? Business vocabulary often hides a denominator. Margin, markup, conversion, retention, churn, CAC and break-even all depend on relationships among quantities. Calculating them reveals whether the learner truly understands the term.

Why are customer interviews not enough? Interviews can reveal experience and reasoning, but people cannot always predict their future behaviour accurately. Combine stated responses with appropriate behavioural evidence, product tests and actual outcomes where safe and permitted.

What if the data contradict the original idea? That is useful learning. Revise the assumption, change the solution, narrow the segment or stop the path. Entrepreneurship becomes wasteful when every result is forced to confirm the founder’s first belief.

Is a large TAM a sign that the idea is good? No. Market size does not establish product fit, reachable demand, economics or execution capability. A smaller but well-understood beachhead can produce more useful learning than a giant abstract market.

Is lower CAC always better? No. Customer contribution, retention, channel capacity and time to payback also matter. A cheaper channel can acquire customers who are less suitable or less valuable under the business model.

Is positive unit economics enough to launch? No. The project can still face weak demand, fixed costs, cash timing, capacity, safety, legal or quality problems. Unit economics answer an important but bounded question.

Should students use real money to learn willingness to pay? Not necessarily. A supervised simulation with a meaningful trade-off can teach the distinction while avoiding pressure. Any real transaction must follow school, parental and local requirements.

What makes a metric actionable? The team understands what it measures, how it relates to an objective and what action a change in the metric would reasonably trigger. A large number is not actionable merely because it is easy to report.

Reference shelf and connected eduKate routes

The U.S. Small Business Administration business-planning resources connect market research, competitive analysis, startup costs and break-even planning. Its public break-even materials use the simple relationship fixed costs divided by selling price minus variable cost per unit. The NSF I-Corps programme provides an example of customer discovery as experiential entrepreneurship education. These references support background concepts; they do not validate the fictional datasets, thresholds or recommendations in this article.

For the broader foundation vocabulary, use Secondary 2 Entrepreneurship Vocabulary | 150 Grade 8 Business, Marketing, Finance and Enterprise Terms. For related high-level subject context on eduKateSingapore, read What Is Business Studies?, What Is Marketing? and What Is Economics?. Those pages own their broader disciplinary explanations; this manual owns the Secondary 2 advanced vocabulary application.

Continue through the Secondary 2 Advanced Vocabulary Collection. The media literacy manual develops source evaluation, quantitative claims and accountable publication. The food science manual develops experimental design, ingredient function and product evaluation. The subject matter changes while the evidence discipline transfers.

Final transfer: build a decision somebody else can inspect

Alicia returns to the original planner idea. She still cares about design, but design now follows a defined user job. Tricia can trace each important claim to an interview, task, calculation or explicit assumption. Kai Kai still likes ambitious forecasts, but he now writes the acquisition path, contribution margin and bottleneck beneath them. When the numbers disagree with the story, the team changes the story.

That is the practical purpose of advanced entrepreneurship vocabulary. It gives the learner handles for uncertainty. Name the assumption. Find the customer. Define the denominator. Test the riskiest belief. Calculate the unit. Follow the cash. Find the constraint. Choose the metric. Write the decision rule. Record why the decision was made. When those actions survive transfer into a new case, the words have become a working business language rather than a list of impressive terms.