The memo you rewrite every weekend

Open the last deal you looked at seriously and find the notes you took. If you are like most angels, they are in a doc that starts fresh every time: what does this company do, is the market real, do the numbers hold up, what would kill it, what do I ask the founder. The same five headings, retyped, deal after deal.

That structure is the right structure. The waste is doing it by hand at the speed of reading, on a weekend, and abandoning it the moment a deal looks unpromising, which is exactly the deal where a fast memo would have saved you the most time.

The five things a first-pass memo is for

A first-pass memo is not the investment decision. It is the thing that tells you whether the decision is worth an hour of your attention. It answers five questions and stops.

• What is the business, in plain terms, stripped of the deck's own adjectives.
• What is the market, and is the size claim built up or handed down.
• What are the numbers that matter, taken from the deck rather than remembered.
• What are the risks, ranked, with the one that could actually sink it at the top.
• What do I ask the founder, in priority order, before I spend more time.

None of that is a recommendation. A good first-pass memo never says invest. It says here is the deal, here is the sharp edge, here is the first question. The judgement is yours and it should stay yours.

The part a model should do

Reading forty slides and pulling out the numbers, noticing that one customer is a fifth of revenue, that a fourteen billion dollar market was derived from a premium pool, that the growth chart and the churn line disagree, is exactly the kind of patient reading a language model is good at and a tired human is not at 11pm on a Sunday.

A human skims that and files it under strong growth. A memo does the arithmetic: 3.1% a month is about 31% of customers a year, so a lot of the growth is replacing churned logos rather than compounding, and the 104% retention is being carried by the accounts that stay. Same four numbers, a different read.

The memo is not the decision. It is the thing that makes the first hour of the decision fast enough that you actually spend it.

What you keep doing yourself

The reference calls. The founder meeting. The gut read on whether these two people are the ones to build this. The memo does not replace any of that, and it should not pretend to. It clears the reading so you arrive at the human part already knowing which three things to press on.

That is the whole idea behind Meridian: same five headings you already use, filled in from the deck and the data room in about a minute, with the numbers grounded back in the document so you can check them. First-pass diligence, in an hour, not a weekend.