Resources · Case study

Case study

Screening 400 inbound decks a quarter

A seed fund with two investors and one analyst, receiving more decks than it could open. Now every deck is scored against the thesis and every founder gets a reply, in the fund's own voice.

Claude Monet, Arrival of the Normandy Train, Gare Saint-Lazare, 1877
Claude Monet, Arrival of the Normandy Train, Gare Saint-Lazare, 1877
400+inbound decks per quarter
100%receive a written reply
2 daysmedian time to first response, down from 19

The situation

Inbound had doubled in a year. The analyst was opening decks in the order they arrived, writing a paragraph on the ones that looked interesting, and quietly never replying to the rest. Two things were going wrong. Good deals were sitting unopened for three weeks. And the fund's reputation among founders was that it did not answer.

The partners had a thesis document. It was four pages long, precise, and nobody scored decks against it, because scoring forty decks a week by hand against four pages is not a job anyone does for long.

What changed

Every inbound deck lands in Navent as a new company on the screening step. The agent reads the deck and the intro email, pulls out the claims a first look turns on, and scores them against the thesis. Each score cites the sentence in the thesis it came from, so a partner who disagrees can see whether the model misread the deck or whether the thesis says something they no longer believe.

Decks are Internal by default; a founder sent them to be read. So screening runs on the default model, in Sweden. The plan for each step shows this, and shows the cost, which for a deck is a few cents.

The thesis went from a document we quoted at offsites to a thing that actually decides what we look at on Monday morning.Partner

Every deck gets a drafted reply: a pass note or a take-a-meeting note, written in the voice of the fund's previous ones. The analyst edits and sends. Nothing goes out unread, but nothing waits three weeks either.

What they found out

  • Around one deck in nine scored higher than a partner's gut had rated it. Two of those became meetings that would not otherwise have happened.
  • The thesis was wrong in one place. Enough decks scored badly on a criterion the partners kept overriding that they rewrote it.
  • Founders noticed the replies. Referrals from passed founders went up, which the partners had not expected and cannot fully explain.

What it did not fix

The meeting is still the meeting. Navent does not make a partner a better judge of a founder in a room. It makes sure the right founders get into the room, and that everyone else hears back.

See it on one of your deals

Bring a data room, a deck and a term sheet. We will walk the process map on your material.