Stating and Defending Your Assumptions
A Blue Tokai store manager choosing a new cafe location is not really betting on “coffee demand.” She is betting on assumptions - morning footfall, office density, repeat visits, delivery spillover, rent, staffing and how many people will actually convert after seeing the storefront.
That is the whole game in case interviews too. When data is missing, weak candidates hide guesses; strong candidates state assumptions clearly, explain why they are reasonable and show how the answer changes if they are wrong.
- An assumption is a clearly stated estimate used when exact data is unavailable.
- Good assumptions are explicit, business-backed, easy to calculate with and open to correction.
- Always state assumptions before using them: “I will assume X because Y; we can adjust it if you have better data.”
- Defend with logic, not confidence: use segmentation, benchmarks, constraints, common sense and sanity checks.
- The most important assumptions deserve sensitivity testing: ask, “If this changes by 20 percent, does my answer still hold?”
- Never stack many hidden assumptions inside one number. Break the number into drivers.
- In interviews, the evaluator is often judging your thinking quality more than your exact number.
The Big Picture: Assumptions Are the Bridge Between Unknowns and Decisions
In guesstimates and business cases, you rarely have perfect data. Your job is to make the uncertainty visible, turn it into structured estimates, and still reach a decision that is logically defensible.
Core Explanation: How to State and Defend Assumptions
The first rule is simple: never let an assumption enter your math silently. If you use “10 percent conversion,” “₹300 average order value,” or “2 visits per month,” the interviewer must know where it came from and how much it matters.
Before making assumptions, define the business problem sharply. A vague problem creates vague assumptions. If this step feels shaky, revise defining the problem before solving it before practising harder cases.
The Assumption Funnel: From Many Possible Guesses to a Defensible Number
Start broad, then narrow. The best candidates do not pull one number from thin air; they filter it through context, segmentation and sanity checks.
The Four Tests of a Good Assumption
A good assumption is not necessarily exact. It is useful because it is clear, logical and decision-relevant.
The Assumption Matrix: Where to Spend Your Defence Time
Not every assumption deserves the same attention. Defend the assumptions that are both uncertain and high-impact; do not waste time debating minor rounding choices.
Definitions You Can Say in One Breath
- Assumption: a stated estimate used to proceed when exact data is unavailable.
- Defending an assumption: explaining why an estimate is reasonable using business logic, benchmarks, segmentation or constraints.
- Sensitivity analysis: testing how the answer changes when a key assumption changes.
- Sanity check: a quick reality test that compares your output with common sense or known constraints.
Worked Example: Estimating Daily Orders for a Food Stall Near a Metro Station
Suppose you are estimating daily lunch orders for a food stall near a busy Indian metro station. You do not know exact footfall, so you need assumptions.
Answer: about 90 lunch orders per day. A good defence would add: “The most sensitive assumptions are lunch-time share and stall conversion. If conversion is 20 percent instead of 30 percent, orders fall to 60; if it is 40 percent, orders rise to 120. So I would give a practical range of 60-120 daily lunch orders.”
You did not pretend the number was exact. You decomposed it, defended each layer and showed which assumptions matter most.
Case Study: Blue Tokai and the Discipline of Store-Level Assumptions
Blue Tokai shows why expansion decisions depend on explicit assumptions about micro-markets, not broad statements like “premium coffee is growing.”
Imagine evaluating a new Blue Tokai cafe in an Indian urban neighbourhood. A shallow answer says: “Young professionals like coffee, so the store should work.” A strong answer breaks the bet into assumptions: nearby office density, morning and evening traffic, weekend catchment, seating turnover, delivery orders, average ticket size, rent burden and staff productivity.
The primary driver is micro-market quality - enough customers with the income, habit and convenience need to buy specialty coffee repeatedly. Supporting drivers include a consistent product experience, visible storefront access, delivery platform demand, menu mix beyond coffee, and disciplined fixed costs such as rent and staffing.
The lesson is simple: expansion is not won by one magic assumption. It is won when the biggest assumption - local repeatable demand - is supported by smaller assumptions on access, operations, pricing and cost discipline.
How AI Changes Stating and Defending Your Assumptions
AI does not remove the need for assumptions. It makes weak assumptions easier to expose.
- Faster benchmark discovery: tools like Perplexity can help you find public context on industry structure, customer behaviour or company filings before practice. Use it to sharpen assumptions, not to copy numbers blindly.
- Better sensitivity practice: ChatGPT or Claude can generate alternate scenarios: “What happens if conversion is 20 percent lower?” This trains you to defend ranges instead of single-point answers.
- Cleaner mock interviews: AI can challenge your assumptions in real time: “Why 10 percent and not 2 percent?” That is exactly the pressure you need before a live case.
Practical workflow: load your case prompt, your assumptions and your final answer into ChatGPT or Claude. Ask: “Identify my three weakest assumptions, suggest better driver-based alternatives, and cross-question me like a consulting interviewer.” For structured practice, use AI as a mock interviewer for practising cases.
Interview Relevance
“Estimate the monthly revenue of a quick-service restaurant inside a large mall. Please state your assumptions clearly.”
When assumptions affect profitability, connect them to unit economics. For example, revenue assumptions become far more powerful when paired with contribution margin and break-even analysis in cases.
Common Mistake
The biggest mistake is defending every assumption equally. Candidates waste time justifying tiny rounding choices while leaving the main driver unsupported. The fix: identify the two assumptions that most affect the final answer, defend those deeply, and move quickly through the rest.