The twelve questions are the five-questions spine (Section 2.1 of the master map) unfolded into a working sequence. Q1–Q4 answer what does it do and how does it make money; Q5–Q8 answer is it a good business; Q9 answers is it run by good, aligned people; Q10 answers what can kill it; Q11–Q12 answer what is the price implying, and is that beatable. Author's guidance for each follows: what the question is really asking, where in the source stack to find the answer, the number or judgment to produce, and the trap that eats beginners.
What does the business do and how does it make money?
Q1 — What does it do, and how does it make money? (the one-line-per-revenue-line question)
Write one line for each revenue line, then find the real profit engine, which is often not the biggest line. Source: the segment note (mandatory, because a company with segments that won't segment its profits is hiding something) and the "what we do" of Item 1 / MD&A / the investor deck. The deliverable is two sentences: what the company sells and to whom, and which line actually makes the money. The engine is the segment or product carrying the operating profit and the returns, not the revenue headline.
Advertisement
Now the archetypal cases you must internalize. ITC: cigarettes are a minority and shrinking share of revenue but ~75–80% of segment EBIT (illustrative); the FMCG, hotels, paper and agri lines are the revenue story, tobacco is the profit story. Costco: merchandise is ~98% of revenue at a razor-thin margin, and the ~2% membership-fee line is roughly half of operating income. The trap is describing the company by its revenue mix and stopping at "it's a diversified consumer company" when the entire economic truth is that one line subsidizes or carries the rest. Always ask: if I could keep only one revenue line, which one keeps the lights on and the returns high? That is the business.
Q2 — Who are the customers, and why do they buy? (concentration · must-have vs nice-to-have · recurring?)
Three sub-questions, each a risk dial. Concentration: is revenue spread across millions of customers or a handful? (10-K Item 1 and the risk factors disclose customer concentration; Indian ARs disclose it unevenly, so check the MD&A and related-party note.) One customer at 30% of revenue is a different, more fragile business than a million customers at 0.0001% each. Must-have vs nice-to-have: is the product something the customer's own operation or life cannot function without (a bank's core software, a diabetic's insulin, the electricity bill) or something deferrable in a bad month (a new watch, a gym upgrade, discretionary SaaS)? Must-haves have pricing power and recession resilience; nice-to-haves have neither. Recurring? does the customer come back by contract, by consumable, by habit, or by subscription, or is every sale a fresh hunt? Recurring revenue (Costco's annual membership, a pharma refill, an FMCG staple) is worth multiples of the same rupee of one-off revenue (a wedding purchase, a capital-goods order). The deliverable is one sentence: "[millions of / concentrated] customers buy [a must-have / a discretionary want] on a [recurring / one-off] basis." The trap: accepting "sticky" or "loyal" as an answer without a mechanism. Stickiness must cash out as a switching cost, a habit, or a contract, or it's a hope (M4.05's discipline, at speed).
Q3 — How does ₹1 (or $1) of revenue flow to profit? (walk the P&L to FCF; where does the money leak?)
Take one unit of revenue and walk it down the income statement to free cash flow, naming the big deductions and where the leaks are. Revenue → gross profit (what does COGS eat?) → EBIT (what do SG&A, R&D, A&P eat?) → after-tax operating profit → and then the two cash leaks the P&L hides: working capital (does growth swallow cash into receivables and inventory, or does the model release cash via customer advances and supplier float?) and capex (how much reinvestment does a rupee of growth demand?). The deliverable is a mental waterfall: "of ₹1 of revenue, ~₹X is gross profit, ~₹Y is EBIT, ~₹Z survives as free cash after working capital and capex — and the biggest leak is ___." Source: the P&L for the margin stack, the cash-flow statement for the working-capital and capex reality (this is why M8.01 makes you find the cash-flow statement fast). The trap: stopping at net margin. A 10%-net-margin business that turns 100% of profit into cash beats a 20%-net-margin business that leaks it all into inventory and capex. The P&L flatters; the cash-flow statement tells the truth. You already know this from M2.04; here you do it in ninety seconds, in your head.
Q4 — What are the unit economics? (smallest repeatable unit · contribution margin · CAC/LTV · payback)
Find the smallest repeatable unit the business runs on (one store, one customer, one subscriber, one warehouse, one loan, one flight, one bed) and ask whether that one unit makes money, and how fast it pays back its cost to acquire or build. The four numbers, where the model has them: contribution margin (revenue per unit minus the variable cost to serve it: what each incremental unit adds before fixed costs); CAC (cost to acquire one customer/unit); LTV (lifetime value: contribution × how long the unit lasts, sometimes discounted); and payback (months or years for a unit to return its CAC). Healthy patterns you learned in M4.02: LTV/CAC > 3×, CAC payback under ~12 months for subscription, contribution margin positive and rising with scale. The deliverable is a judgment: does a single unit earn a good return on what it costs to create, and does that improve or decay with scale? Source: the investor deck (management often gives cohort or per-store economics), the segment note, and your own division of the totals (revenue ÷ stores, gross profit ÷ subscribers). The trap: accepting company-defined "contribution margin" that excludes real costs (a favourite of loss-making platforms, the CM1/CM2/CM3 game from M5.05); rebuild it with the costs the business actually can't avoid.
Is it a good business?
Q5 — Who competes, and does this company have the right to win? (share gains or losses)
Name the two or three real competitors and ask the single most diagnostic question in competitive analysis: is this company gaining or losing share, and at what cost? Share direction over five to ten years is the cleanest read on competitive position you can get in a teardown (M4.05's market-share-stability test, compressed). Gaining share while margins hold or rise = a genuine right to win. Gaining share while margins fall = buying growth (a warning). Losing share slowly while protecting margin = harvesting, sometimes rationally. Stable shares among a few players earning good returns = a protected oligopoly (a moat signal; confirm in Q6). Source: the deck's market-share claims (and their CRISIL/Frost citations for India, or 10-K competition section for the US), and DRHP industry chapters for Indian share data. The deliverable is a direction and a cost: "gaining/holding/losing share, at improving/stable/deteriorating margin." The trap: taking a company's self-reported share crown at face value. Always ask "share of what market?" (M4.05's market-definition lesson; a company that has to gerrymander the market to claim leadership is telling you something).
Q6 — Is there a moat? (ROIC durability + which of the five sources). The moat quick-test lives in its own section; run it here.
Q7 — What are the economics? (growth = volume vs price · margins · ROCE/ROIC vs the ~12% hurdle · balance sheet · cash conversion)
This is the quantitative heart of "is it a good business," and it has five reads, all from the ten-year panel M8.01 put on your screen. (a) Decompose growth into volume vs price. Revenue up 15%: is that 12% more units or 12% higher prices, or some mix? Volume growth is healthier and more durable; pure price growth eventually meets elasticity, and pure price growth in a commodity is just the cycle. (FMCG's cardinal rule from M5.05: you want mid-single-digit volume growth; value-only growth on flat volume is a warning.) (b) Margin structure and trend: gross, EBIT, net; stable, rising, or eroding, and why. (c) ROCE/ROIC vs the hurdle, the master test (M2.03): is the return on capital durably above the ~12% post-tax INR / ~8–9% USD cost-of-capital hurdle (as of mid-2026, check against your own M3.03 builds)? A business at 25% ROIC for a decade is a different animal from one oscillating around 10%. (d) Balance sheet: net debt/EBITDA, and any of the M8.01 Ctrl-F landmines (pledges, contingent liabilities, going-concern). (e) Cash conversion: CFO/EBITDA and FCF/PAT over several years, where healthy is CFO/EBITDA broadly ~80%+ and FCF/PAT ~80%+ across a cycle; persistent shortfalls mean the profit is being eaten by working capital or capex, or isn't real (M2.04, M2.07).
The deliverable is the economics verdict: good / average / poor. This single call, more than any other, is what the teardown exists to make. The trap: reading one year. A good business is one that earns high returns across a cycle; a single great year is a data point, not a verdict.
Q8 — How is capital allocated? (where did cash go over 5–10 years; incremental returns)
Over the last five to ten years, the company generated a pile of operating cash. Where did it go: reinvestment in the core, acquisitions, dividends, buybacks, debt paydown, or diversification into unrelated ventures? And the sharper question: what return did the reinvested capital earn? The proxy you can compute in a teardown is incremental ROIC, roughly the change in operating profit over the change in invested capital across the period (M2.03/M3.07). A company retaining most of its earnings at a high incremental ROIC is a compounder; one retaining earnings at a low incremental ROIC is destroying value however good its legacy business (M4.05's I × (ROIC − WACC), since growth only helps when the spread is positive). Source: the cash-flow statement over five to ten years, whose financing and investing sections tell you where cash went, plus the returns trend. The deliverable is one sentence: "management has [reinvested at high returns / diworsified into low-return adjacencies / returned cash sensibly / bought back stock above or below value]." The trap: praising a dividend or buyback without checking the price it was done at (buybacks above intrinsic value destroy value, per Thorndike's Outsiders discipline) or praising "growth" that was capital retained at sub-hurdle returns.
Who runs it, and what could kill it?
Q9 — Management and incentives, a quick pass. (the fast version; the deep version is M8.04)
You are not doing a full governance workup in a teardown; that's Day 3 of the deep dive. You are taking three fast reads. (a) Skin in the game and alignment: do the people running it own meaningful stock, and are they paid for the right things (returns, per-share value) or the wrong ones (revenue, size, adjusted EBITDA)? US: the DEF 14A proxy pay-metrics table. India: the Board's Report, the managerial remuneration note, and, critically, the promoter shareholding and pledge percentage (a high or rising pledge is a five-alarm governance flag from Phase 2). (b) Candour: in the one or two transcripts you read (M8.01), do they answer analysts straight, own mistakes, and match prior promises, or do they dodge, blame macro, and use adjusted metrics to bury bad news? (c) The related-party and contingent-liability landmines you Ctrl-F'd in M8.01: a quick scan for value leaking to insiders. The deliverable is one line: "aligned and candid / capable but empire-building / red flags — [name them]." The trap: being charmed by an articulate CEO on a podcast; charisma is not alignment, and the transcript plus the pledge percentage outrank the narrative every time.
Q10 — What are the three key risks that actually matter? (not the risk-factor boilerplate, but the real ones)
The 10-K's Item 1A lists forty risks, thirty-eight of which are lawyer-generated boilerplate ("economic conditions may affect us"). Your job is to name the two or three risks that could genuinely impair this specific business: the ones that, if they happened, would break the thesis, not merely dent a quarter. They usually come from customer concentration (Q2), a single regulatory or key-plant dependency (a USFDA import alert for a pharma; an AGR ruling for a telecom), a technology or business-model disruption (counter-positioning against the incumbent, M4.05), a balance-sheet fragility (an NBFC's ALM mismatch, M5.02), commodity/input exposure, or a moat under active attack (Asian Paints vs Birla Opus). The deliverable is a short list of specific, company-killing risks, each with a one-line mechanism. The trap: listing generic risks (competition, regulation, macro) that apply to every company and therefore inform nothing. A real risk is one you could imagine reading about in this company's obituary.
What does the price imply, and is it fair?
Q11 — What is the market implying? (reverse the multiple: what growth/margin does today's P/E, EV/EBITDA, or P/B assume?) The machinery is reversing the multiple.
Q12 — Is it cheap or expensive versus its quality? (the demanding-or-generous verdict)
The final synthesis: put the price (Q11's implied story) next to the quality (Q6–Q8's verdict) and judge whether the multiple is demanding or generous relative to what the business deserves. A great business at a demanding price and a poor business at a generous price can be the same investment mistake; the teardown's price verdict is always relative to quality, never absolute. A 55× P/E is generous for a cyclical and possibly fair for a franchise compounding at 18% with a twenty-year moat; a 9× P/E is expensive for a melting ice cube and cheap for a durable mid-teens grower. The deliverable is one of three verdicts: the price is demanding (asking more than the evidence supports) / roughly fair / generous (asking less), together with the honest confession that goes with a teardown, which is that this is a first-pass judgment, and if the answer matters, the variant question tells you what to research before acting. The trap: anchoring on the raw multiple ("50× is expensive," "8× is cheap") without the quality on the other side of the scale. The entire discipline of Q12 is refusing to say "expensive" or "cheap" as a standalone word.