Company Diligence Memo
A company roadmap is a sales document; a diligence memo is what you build after you stop believing it. This capstone turns a company thesis into source-labeled claims, proof gates, kill criteria, and a decision you can defend out loud.
Good diligence is measured by how clearly it states what would change its mind. Certainty is the one thing a memo should never sell.
This capstone turns a quantum company thesis into an inspectable object: claims, source-labeled evidence, proof gates, kill criteria, and a decision label you can defend out loud. It trains judgment — it is not investment advice.
Core concepts: capstone synthesis, company diligence, investment proof gates.
Diligence is managed uncertainty
The company diligence memo converts a company or startup thesis into something another person can audit: claims, evidence with source labels, risks, proof gates, kill criteria, and a decision. It reads clearly enough for an oral defense and stays careful enough to never quote a roadmap as a fact.
That care is what separates the memo from the materials it examines. A pitch sells certainty; a diligence memo manages uncertainty — where it lives, how much it costs, and what would retire it.
The scoring spine
Two scores run through the memo. The proof-gate decision score: . The opportunity value: . Add the teaching mastery check — explanation, worked check, transfer prompt, limitation — so the memo can be defended and taught, not just filed.
The scores are not the decision; they are the argument's skeleton. Claim quality asks whether the technical claim is specific enough to be wrong. Proof progress counts gates already passed. Kill-criteria pressure measures how close the thesis sits to its own tripwires. When two analysts score the same company differently, the formula shows exactly which input they disagree about — and that input becomes the next diligence question.
Anatomy of the memo
State, in order: category, customer, technical claim, source-labeled evidence, baseline, moat, timing, proof gates, risks, kill criteria, decision label, and the evidence that would change the label. Proof gates appear before the conclusion — the reader should meet the tests before the verdict.
The source table is the spine. Separate, at minimum:
- Official company claims — roadmaps, blog posts, keynote numbers.
- Independent technical evidence — reproduced results, third-party benchmarks, peer-reviewed work.
- Customer evidence — pilots, contracts, named workflows.
- Standards and ecosystem evidence — adopted interfaces, published methods.
- Weak sources — funding announcements, press coverage, analyst aggregation.
The table exists because the classic diligence failure is treating a roadmap, a funding round, and a measured result as equally strong. They are not, and the memo must say so explicitly.
Two more sections carry weight beyond their length. Timing states why now — what changed in hardware, cost, or demand that makes this attempt different from the last one. Moat states what survives contact with a well-funded competitor: proprietary fabrication steps, accumulated calibration data, exclusive partnerships, or nothing at all. A memo that writes "team" under moat and stops has answered a different question.
How diligence memos fail
The worst failure is the pitch shape: certainty sold, uncertainty buried. Close behind it: roadmap language quoted without translation into metrics and missing evidence; length that hides the decision (the reader should find the label, the evidence, the risks, and the next proof gate in under a minute); and the vague label. Monitor, wait, partner, build, invest, and avoid are different actions with different costs — pick one, and attach a review date or an evidence trigger.
The engineer's analogy
This is a design review. Name the system, the assumptions, the alternatives, the failure modes, the tests, and the rollback conditions. The company is the system, embedded in technical and market dependencies you do not control. And like a good design document, the memo must be falsifiable: if no conceivable evidence could change the answer, whatever you wrote, it was never diligence.
Who uses this memo
A builder turns it into a startup thesis. A partner uses the proof gates to decide whether to commit engineering time. A learner walks out with an artifact that proves they can connect hardware, software, applications, and strategy — the actual point of the capstone. The memo can also become a playground template route; keep any public version anonymized or dated, and keep the teaching goal separate from anyone's real money.
Exercise
Write the memo. Pick one company or startup thesis and produce the full document.
- Submit: two pages plus oral-defense notes, with decision label, evidence, risks, proof gates, kill criteria, and the source table.
- Check: compute the decision score and the EV; verify the proof gates appear before the conclusion.
- Transfer: rewrite the memo for a company in a different category — hardware, software, or application — and note what moves.
- Repair: if the draft reads like a pitch or a summary of sources, rework the evidence discipline after Chapter 82, Investment Memo and Startup Thesis Templates.
Check your understanding
Defend the memo orally, including what evidence would change the decision and what would kill the thesis outright.
A passing defense quotes no roadmap without translating it into metrics, missing evidence, and a proof gate.
If you get stuck
If the memo hides uncertainty or treats roadmap language as evidence, revisit Chapter 81 (Reading Roadmaps Like an Operator) for claim translation and Chapter 82 (Investment Memo and Startup Thesis Templates) for the memo skeleton — then rewrite the source table first.