Your download
The Hidden Margin Teardown Kit
<!-- 한국어 메모: 이 문서는 Notion 에 그대로 붙여넣는 무료 리드마그넷 본문(영어)이다. 사실 주장이 아니라 "분해 프레임워크 + 빈 워크시트"이므로 날조 표면이 없다. 브랜드/기업 실명은 등장하지 않는다 — 사용자가 자기가 고른 대상에 프레임워크를 적용한다. -->
The Hidden Margin Teardown Kit
A 7-question framework for figuring out where any company, product, or subscription actually makes its money — not where it says it does.
Most pricing pages describe what you're buying. They almost never describe what you're funding. This kit gives you a repeatable way to close that gap, using only information you can find or observe yourself. No insider knowledge required, none assumed.
Before you start
Pick one target: a company, a product, a subscription, or a service you already use or can research publicly (its website, pricing page, app store listing, packaging, terms of service, or public filings if it's a public company). Keep the target narrow — "this specific plan" or "this specific product line," not "this entire company."
This is an observation exercise, not a verdict. You're building a hypothesis about how the money moves, not proving it. Where you can't find evidence, the worksheet asks you to say so — an honest "unknown" is a better answer than a guessed number.
The 7 questions
- What does the customer think they're paying for? Write the product/service in the customer's own language — the thing on the label, not the thing underneath it.
- What's the cheapest version of this that would technically still work? Strip the packaging, the brand, the convenience. What's left is closer to the underlying cost structure than the price tag is.
- Where does money keep coming in after the first purchase? Subscriptions, refills, replacement parts, warranties, financing, storage fees, in-app purchases. A one-time sale and a recurring relationship are different businesses wearing the same product.
- What's priced to lose money, and what's priced to make it back? Some items exist to get you in the door. Find the item that looks generous and ask what it's next to.
- What are you paying extra for that isn't the product itself? Convenience, speed, status, guilt relief, fear reduction, FOMO. Name the emotional or situational premium separately from the functional price.
- Who else pays this company besides you? Advertisers, data buyers, wholesale partners, affiliates, franchisees, landlords collecting a cut, platforms taking a fee. A "free" product usually means you found the wrong customer to look for.
- What single change would break this model? A supplier price hike, a platform policy change, a shift in customer habits, a new regulation. This question tells you how fragile the margin actually is, which is often more interesting than the margin itself.
The worksheet
Copy this into a Notion database (each row = one question is one property set; or run one full pass per target as a page). Columns map directly to Notion property types noted in brackets.
| Question # [Select] | Your Observation [Text] | Evidence / Where You Found It [URL or Text] | Confidence [Select: Low / Medium / High] |
|---|---|---|---|
| 1 | |||
| 2 | |||
| 3 | |||
| 4 | |||
| 5 | |||
| 6 | |||
| 7 |
Target: _______________________ Date: _______________________ One-sentence hypothesis (after all 7 rows): _______________________________________
How to fill it in
- Evidence column is not optional. If your answer is "I think..." write that down honestly in the Confidence column as Low — don't upgrade a guess to High because it sounds right.
- Public sources only. Pricing pages, app store screenshots, packaging photos, annual reports (for public companies), job postings (they reveal what a company is investing in), review sites, and your own receipts are all fair game. Anything you'd have to call an "unnamed source" for doesn't go in this worksheet.
- Low confidence is a valid final answer. The point of the kit is to make your certainty visible, not to force a conclusion. A worksheet with three "Low" rows and one sharp "High" row is more useful than one where everything is guessed at High.
- Re-run it quarterly on the same target if you're tracking a company over time. Pricing pages and fee structures change; a stale teardown is worth re-checking before you repeat it as fact.
What's next
If this 7-question pass was useful, the $19 Vault has the expanded version: a longer worked-example case file per teardown, more granular sub-questions for subscription and marketplace businesses specifically, and a comparison template for running the same 7 questions on two competitors side by side. It's an extension of this exact worksheet, not a different product.
<!-- 한국어 메모: Claude/ChatGPT 에 그대로 붙여넣는 프롬프트 10개(영어). 각 프롬프트는 역할·맥락·출력형식·제약을 포함한다. 수치/매출을 만들어내라고 지시하는 프롬프트는 없다 — 전부 "네가 붙여넣은 실제 텍스트를 분석하라"는 형태로 설계했다. -->
Money & Business Secrets — Prompt Pack
Ten prompts for Claude or ChatGPT, built to pair with the Hidden Margin Teardown Kit. Copy the full prompt, replace the bracketed [ ] sections with your own target, and run it. Every prompt asks the model to reason over information *you* provide — none of them ask it to invent numbers, quote sources, or make claims about a real company's finances it can't verify.
1. Margin Teardown Assistant
What it's for: Runs your target through the 7-question framework and flags where your evidence is thin.
You are helping me apply a 7-question business-model teardown framework to a company
or product I'm researching. Here is the framework:
1. What does the customer think they're paying for?
2. What's the cheapest version of this that would technically still work?
3. Where does money keep coming in after the first purchase?
4. What's priced to lose money, and what's priced to make it back?
5. What am I paying extra for that isn't the product itself?
6. Who else pays this company besides me?
7. What single change would break this model?
My target: [company/product name and a one-sentence description]
My research notes so far: [paste your own notes, pricing page text, or observations]
For each of the 7 questions: give me your best-supported answer based ONLY on what I
provided above, then rate your own confidence (Low/Medium/High) and say explicitly what
additional evidence would raise that confidence. Do not invent numbers, quotes, or
sources I did not give you. If you don't have enough information for a question, say
"insufficient information" rather than guessing.Swap: the target description and your research notes. The more raw material you paste in, the less the model has to guess.
2. Loss Leader Finder
What it's for: Helps you spot which product in a lineup is priced to lose money on purpose.
I'm going to describe a company's product lineup. Help me identify which item(s) are
most likely priced as a "loss leader" (priced at or below cost to drive a different,
more profitable purchase) and which item(s) are most likely the actual profit center.
Product lineup: [list every product/tier/plan and its price]
Walk through your reasoning for each item, then give me a ranked hypothesis (most likely
loss leader → most likely profit center). Label this as a hypothesis, not a fact — I
have not given you their actual cost structure, only prices.Swap: the product lineup and prices.
3. Recurring Revenue Detector
What it's for: Finds the parts of a purchase that create ongoing payments, not just a one-time sale.
Here is a product or service description: [paste product page, packaging text, or your
own summary]
Identify every mechanism in this description that could create recurring or repeat
revenue after the first purchase — subscriptions, consumables, refills, warranties,
storage fees, in-app purchases, financing plans, membership renewals. For each one,
quote the exact phrase from my text that signals it. If none are present, say so plainly
instead of inventing one.Swap: the product/service description.
4. Pricing Psychology Decoder
What it's for: Explains which known pricing-psychology techniques (anchoring, decoy pricing, charm pricing) a specific price list is using.
Here is a pricing page or price list: [paste the tiers/prices/labels exactly as shown]
Identify which well-documented pricing psychology techniques this appears to use —
for example: anchoring (a high-priced option shown first), decoy pricing (a deliberately
unattractive middle option that makes another option look better), charm pricing
(prices ending in .99 or .97), or bundling. For each technique you flag, point to the
specific price or label that supports it. Don't claim a technique is present if the
evidence is weak — say "possible but not clear" instead.Swap: the pricing page content.
5. Two-Company Business Model Comparison
What it's for: Compares how two competitors in the same space seem to make money differently.
I'm comparing two companies in the same market. Here's what I know about each:
Company A: [name + everything you know: pricing, product, positioning]
Company B: [name + everything you know: pricing, product, positioning]
Based only on what I've given you, compare their apparent business models: pricing
structure, what seems to be the profit center for each, and where they seem to be
optimizing for different customers. Flag anywhere you're inferring rather than working
from something I stated directly.Swap: both company descriptions.
6. "What Would Break This" Stress Test
What it's for: Pressure-tests how fragile a business model is to a single external change.
Here is my understanding of how [company/product] makes money: [paste your teardown
notes or a summary].
Give me 5 plausible external changes (supplier cost shift, platform policy change,
regulation, customer habit shift, new competitor) that would most damage this model if
they happened. For each one, explain the mechanism — why that specific change would hurt
this specific model, not businesses in general. Rank them by how much damage they'd
plausibly cause.Swap: your own teardown notes or summary.
7. Public Filing / Annual Report Digest
What it's for: Summarizes a public company's own disclosures for margin-relevant signals, using text you supply (not the model's memory of the company).
I'm going to paste an excerpt from a public company's annual report, investor letter, or
earnings call transcript. Read only the text I provide — do not use outside knowledge
about this company to fill gaps.
Excerpt: [paste the text]
Summarize what this excerpt reveals about: where revenue actually comes from, what
management says is growing or shrinking, and any risk factors they disclose themselves.
Quote the specific sentences you're basing each point on.Swap: the actual filing excerpt. This only works well with real pasted text — if you don't have the filing, skip this prompt rather than asking the model to recall figures from memory.
8. Hidden Fee Auditor
What it's for: Reviews a service's terms of service or checkout flow for you to spot fees that aren't in the headline price.
Here is the checkout flow / terms of service / fee schedule for a service I use: [paste
the text]
List every fee, charge, or cost mentioned that is not part of the advertised headline
price — cancellation fees, currency conversion, service fees, restocking fees, minimum
purchase requirements, auto-renewal terms. Quote the exact clause for each one.Swap: the checkout or terms-of-service text.
9. Competitor Positioning Map
What it's for: Turns a list of competitors into a simple 2-axis positioning description you can sketch out.
Here are [N] competitors in the same space, with what I know about each one's price
point and target customer: [list them]
Suggest two axes that would meaningfully separate these companies on a positioning map
(for example: price vs. convenience, or DIY vs. done-for-you), then place each company
on that map with a one-sentence justification. Base every placement only on the
information I gave you.Swap: your competitor list and what you know about each.
10. Weekly Teardown Digest Generator
What it's for: Turns a completed teardown worksheet into a clean written summary you can save to Notion or use as a personal note — not a public claim, a private digest.
Here is my completed 7-question teardown worksheet for [target]: [paste your filled-in
answers and confidence ratings from the worksheet]
Turn this into a clean 150-word summary written for my own notes, written entirely from
what I gave you. Preserve my confidence ratings — if I marked something Low confidence,
phrase that part of the summary as a hypothesis ("this appears to be...") rather than as
a settled fact.Swap: your filled-in worksheet answers.