Meeting notes that matter three weeks later

Transcript notes versus decision notes

A transcript note answers "what was said." It is a chronological dump: they asked about churn, I explained the cohort thing, they liked the design partner story, we talked about the round, they have to run. It feels thorough because it is faithful to the conversation. It is also nearly worthless three weeks later, because the conversation is not what you need. You need the decisions the conversation should drive.

A decision note answers a different question: "what does my future self need to know to make the right next move on this thread?" That is a small, specific set of things, and almost none of it is the chronological flow of who said what. The partner's exact wording on churn does not matter. Whether their churn question was a real objection or polite curiosity matters enormously, because one means you owe them a proof point and the other means you do not. Transcript notes capture the words and lose the signal. Decision notes throw away the words and keep the signal.

The test for whether you are taking the right kind of note is simple. Open a note from two weeks ago, a thread you have not touched since. Can you, in sixty seconds, answer: is the ball in my court or theirs, what is the one thing that would move this forward, and what did I promise? If you can't, you took a transcript.

The seven fields a fundraising note actually needs

A useful note is not longer than a bad one. It is structured differently. After running the same call dozens of times, the same seven slots are the ones that carry the next move. Everything else is noise you can drop.

The first is signal: your honest read on whether this is real interest, polite interest, or a courtesy meeting for your connector. Write it as a verdict, not a feeling. "Real, moving toward process" is useful. "Went well" is not.

The second is objections: the specific concerns they raised, separated into ones that are real blockers and ones that were curiosity. This is the single highest-value field and the one transcript notes destroy, because in the moment every question feels equal. Mark which ones, if answered, actually change their decision.

The third is their thesis: why this investor, specifically, might care about your company. Not their fund's generic stage and sector, but the particular angle that lit them up. The partner who leaned in when you mentioned your go-to-market is telling you what to lead with next time.

The fourth is promised action, with an owner and a clock. Every meeting ends with something owed by someone, or it ends dead. Write who owes what and by when. "She intros me to the Acme VP head of growth, this week" is a note. "Talked about intros" is a thread you will lose.

The fifth is proof requested: the concrete thing they want to see before they move. A metric, a reference, a cohort chart, a customer call. This is different from an objection. It is the homework that closes the objection.

The sixth is personal context: the one or two human details that make your next message land as a person, not a pipeline entry. They are about to go on paternity leave. They invested in a competitor's adjacent space. They came up as an operator at a company you respect. This is what keeps follow-ups from reading like mail merge.

The seventh is next move: given all of the above, the single most useful thing you can do next, written as an action you could execute today. This is the field that turns the note from a record into an instruction.

Bad note, good note

Here is the same meeting captured both ways. The transcript version is what most founders write. The decision version is what the thread needs three weeks later.

TRANSCRIPT NOTE — Priya, Northgate Ventures
  Good call, 45 min. Asked a lot about churn and the
  enterprise motion. Liked the design partner story.
  Talked about the round, said they usually do seed.
  Mentioned someone at a portfolio co I should meet.
  Send deck + metrics. Follow up next week.

DECISION NOTE — Priya, Northgate Ventures        [2026-05-22]

  SIGNAL        Real interest, but gated. Moved toward process
                (asked round size, timeline). Not a courtesy take.

  OBJECTIONS    [BLOCKER]  Enterprise churn: she's seen this
                           category churn hard post-pilot. THE gate.
                [curiosity] Founder-market fit. Satisfied in call.

  THEIR THESIS  Believes the wedge is the design-partner motion,
                not the product. Lead future updates with that.

  PROMISED      → SHE intros me to Marcus (VP Eng, Lumen) for a
                   reference on the enterprise motion. "This week."
                → I owe: deck + retention cohort chart.

  PROOF NEEDED  Net retention by cohort, 12-mo, enterprise only.
                Until she sees it staying flat, churn = open.

  PERSONAL      Ex-operator, scaled Lumen's GTM. Speaks ops, not
                hype. Skeptical of "growth" framed as vanity.

  NEXT MOVE     Send cohort chart framed AT the churn fear, not as
                a brag. One line: "retention held through pilot
                conversion. Here's the cohort view." Wait on her
                Marcus intro before nudging.

The two notes took about the same time to write. One of them tells you, weeks later, exactly what to send and why. The other tells you a meeting happened.

A tagging system so notes are findable, not just written

Writing good notes solves nothing if you can't find the right ones when a thread wakes up. Across thirty conversations you need to sort by state, not scroll by date. A light tag line at the top of each note does this. Four tags, fixed vocabulary, so they stay sortable:

A header like `#stage:partner-mtg #ball:mine #temp:hot #owe:cohort-chart` means that when you sit down for a fundraising work block, you filter to `#ball:mine` and you instantly have your to-do list across every thread, ranked by `#temp`. No re-reading, no guessing, no thread quietly dying because it never surfaced. The tags turn a pile of notes into a queue.

The fundraising meeting note template

Copy this into your notes app. One block per investor meeting.

INVESTOR / FUND — [name, fund]                    [date]
#stage:______  #ball:mine/theirs  #temp:hot/warm/cool/dead  #owe:______

SIGNAL        [real / polite / courtesy — one-line verdict]
OBJECTIONS    [BLOCKER]   ...
              [curiosity] ...
THEIR THESIS  [why THIS investor, specifically, might care]
PROMISED      → THEY owe:  [what / by when]
              → I owe:     [what / by when]
PROOF NEEDED  [the concrete thing that closes the blocker]
PERSONAL      [1–2 human details for the next message]
NEXT MOVE     [single action, executable today]

Where this connects to RoundOS

The manual version of this works and you should run it whether or not you ever touch a tool. But it has a known failure mode: it depends on you writing the decision note right after every call, every time, for weeks, while exhausted. Miss a few and the system has holes exactly where the threads are most active.

RoundOS picks up the call note where it already lives. You paste or upload the raw notes from a meeting, the messy transcript-style dump you actually wrote, and it pulls them apart into the structured fields: the objection that was a blocker versus the one that was curiosity, the promised action and who owns it, the proof point requested, the next move. It connects that note back to the investor and fund context it already holds, so the objection lands against what you know about how that partner thinks. Then it puts the thread in your decision queue with the ball clearly in someone's court, and when you ask for it, drafts the next move aimed at the actual concern rather than a generic check-in. The note stops being a thing you have to discipline yourself to write well and becomes an input the system reads for you.