Memo structure

The 8-section IC memo, section by section

Almost every institutional investment memo — a16z, YC, NVCA-style, or a first-time solo GP's — covers the same eight things in roughly the same order. Not because someone standardized it, but because a partnership asks the same eight questions before it commits capital. Here's each section, what it has to accomplish, and the single mistake that guts it.

Structure re-derived from public institutional-memo norms. Nothing below is copied from a course or an employer's template.

Why eight sections, and why this order

An IC memo isn't a pitch deck retyped. A deck sells; a memo decides. Its job is to let a reader who has never met the founder reconstruct your judgment and either agree or find the hole. The order matters: you state the thesis, then spend the rest of the memo testing whether it survives contact with the market, the product, the team, the price, the return math, and the risks — and only then make the call.

Read them as a chain. Each section answers one question; if any link is weak, the memo's conclusion is weak, and a good IC will find exactly that link. Below, each section lists its job, the questions to answer, and the killer — the mistake that most often sinks it.

1 · Target Overview & Thesis

Job: state what the company does and the one-sentence reason you'd write the check — before any detail.

• In one sentence a stranger understands: what does the company sell, to whom, for what?
• Round mechanics: stage, amount, instrument (SAFE / priced), pre/post, your proposed check.
• The thesis: "We invest because ____, and this returns the fund if ____."
• What has to be true for this to be a fund-returner? Name it now — the rest of the memo tests it.

2 · Market Sizing & Timing

Job: prove there's a large, reachable, inflecting market — bottom-up, with a worked correction of the founder's number.

• Bottom-up SOM: reachable buyers × realistic ACV × attach rate. Show every input and its source.
• Restate the founder's TAM, then correct it — what did they overcount, and what's the honest number?
• Why now? The specific change (cost, regulation, behavior) that makes this buildable today.
• Market structure: fragmented or consolidated? Where does the wedge expand next?

There's a whole guide on doing this correctly — see stage-weighted screening for how the Market dimension is weighted and scored.

3 · Product / Solution Evaluation

Job: establish that the product is real, better on an axis buyers pay for, and building a compounding moat.

• What exists today (live, sandbox, demo) vs. roadmap? Be explicit about the line between the two.
• Better on what dimension buyers actually pay for — and by roughly how much?
• The moat: what compounds (data, network, switching cost, distribution) so year-3 is harder to copy than year-1?
• Build-vs-buy: could a funded incumbent ship this in a quarter? If so, what's the defensible head start?

4 · Team Evaluation

Job: answer why THIS team wins THIS specific race — with evidence, not adjectives.

• The unfair insight: what do these founders know or have that outsiders don't? Tie it to lived experience.
• Evidence of execution: what have they already shipped, sold, or built with these hands?
• Can they hire ahead of the curve? Who's said yes, and who's the next critical hire?
• Alignment: equity split, vesting, decision-making — resolved and documented? Any departed-founder overhang?

5 · Valuation & Comparables

Job: anchor the price to comparable financings and a defensible method, not to the founder's ask.

• Comparable rounds: 3–5 recent financings at this stage/sector — who raised, how much, at what range?
• Where does THIS round sit vs. comps — cheap, in-line, or rich — and why is that justified?
• Cross-check with a forward method (plausible next-round revenue × a sane multiple, discounted) — do they meet in the middle?
• Entry ownership: at this price and check, what % do you own, and does it clear your fund's ownership floor?

6 · Cap Table & Exit-Return Waterfall

Job: show what YOUR position is worth across exit scenarios, after dilution and liquidation preferences — the math cheap templates get wrong.

• Model dilution: your entry %, then the effect of the next 1–2 rounds and option-pool top-ups.
• Exit-multiple grid: at conservative / base / upside exits, your gross return (MOIC) and rough IRR.
• Liquidation preferences: whose 1× sits ahead of you, and how it changes proceeds at a low exit.
• Fund math: does the base case return a meaningful fraction of the fund; does the upside case return it outright?

This is the section most first memos get wrong. Two guides go deep: dilution + liquidation preference and MOIC vs IRR.

7 · Risks & Red Flags

Job: list the ways this loses money, honestly ranked — the section that earns the memo's credibility.

• Top 3 risks that actually kill the return (market, execution, regulatory, competitive) — ranked by damage × likelihood.
• For each: what would you need to see to get comfortable, and is that observable pre-investment?
• Diligence red flags from the screening scorecard (cap table, metrics, unit economics, IP) — status of each.
• The bear case in the founder's strongest words: if this fails, what will the post-mortem say?

8 · Recommendation & Terms

Job: state the decision, the check, the conditions, and what would change your mind — unambiguously.

• The call: Invest / Pass / Track — one word, up front, no hedging.
• If invest: check size, ownership, instrument, and any conditions before wiring.
• The kill-switches: what during confirmatory diligence would flip this to a pass?
• Post-investment: the first thing you'd help with, and how you'll know in 6–12 months if the thesis is tracking.

How the sections fit together

Section 1 makes a claim ("this returns the fund if X"). Sections 2–5 test whether X is plausible — is the market real, the product real, the team able, the price fair. Section 6 turns "plausible" into dollars: even if everything works, does your diluted, post-preference position actually return capital? Section 7 is the honesty check. Section 8 forces a decision. Skip any one and a sharp reader will feel the gap.