Study how companies make money without inventing the numbers they do not disclose.

Company research dossier: evidence before narrative

A useful company breakdown begins with a dated evidence register rather than a polished description. Record what the company sells, who pays, which actions trigger payment, how the offering is delivered and what external dependencies influence margins. For each assertion, tag its status as filing-supported, company-reported, independently corroborated, analytical inference or unavailable. If a privately held company does not report financial metrics, leave the corresponding cells unresolved. A price shown on a public product page may identify a commercial offer but does not establish customer mix, actual discounting, average contract value or revenue.

Build the dossier around the unit of exchange. In a subscription business that may be a seat, account or period; in a marketplace it may be a completed transaction; for an API provider it may be a measured call or consumption unit. Next ask which parties create value and which party remits payment. Mapping these relationships prevents a common mistake: confusing a platform's transaction volume with its own recognized revenue, or confusing gross receipts with contribution. Where product documentation describes a usage meter but not the company's actual revenue composition, describe the meter without inferring its financial significance.

Comparison matrix: what can safely travel between companies?

Compare companies across buyer, revenue trigger, cost-to-serve driver, acquisition channel, retention mechanism, operational dependency and evidence quality. Avoid ranking them by a single estimated growth or profitability figure. A model with recurring contractual revenue can still have weak renewal economics; an apparently high-margin digital offering can have variable servicing or acquisition costs. The matrix should show trade-offs and missing observations, not award points for opacity. This approach is particularly important for startups whose public statements emphasize opportunity and product development rather than independently reviewed financial performance.

A second matrix tests transferability to the reader's business. Mark each proposed practice as transferable, conditionally transferable or inappropriate and explain why. A pricing tactic may depend on procurement norms; a marketplace incentive may depend on density and network effects; an enterprise sales process may require contract values that support specialist staff. The practical lesson is therefore conditional: copy the underlying mechanism only after matching customers, economics and execution resources. A company case study should help the reader reject unsuitable tactics as readily as identify promising ones.

Falsification questions and update policy

To keep a teardown research-grade, ask what evidence could disprove the current interpretation. Could a headline offer conceal custom enterprise pricing? Might implementation be delivered by partners? Does the claimed audience match independent customer examples? Has the company changed packaging or sales channels since the source was published? Search official disclosures and dated documentation before treating an older analysis as current. For private financial matters, ambiguity is an answer; it is not permission to fill in an attractive number.

Close every company analysis with a short decision record stating the supported facts, key unknowns, business-model hypothesis, alternative interpretation and what additional primary evidence would resolve the uncertainty. This separates observed behaviour from a founder's strategy recommendation. It also makes later comparisons easier: analysts can update one disputed assumption without rewriting the entire company narrative or altering the historical record.