The Thursday read

What the data says about founders who won.

Latest · Issue 07 · 2026-07-23

Most why-now slides say nothing.

Every deck has a why-now slide, and most just say the market is growing. In our set, 6 of 8 failures gave a reason that could describe fifty other companies.

Read the issue →

Issue 06 · 2026-07-09

The best co-founders used to be coworkers.

The best co-founders had worked together before. 28 of 35 winners had a relationship that predated the company.

Issue 05 · 2026-07-02

The demo was better than the product.

In every fraud in the data, the demo showed something the product could not do. Theranos and Magic Leap both collapsed.

Issue 04 · 2026-06-18

More money, worse odds.

The four largest raises in our data all failed, roughly $18B combined. The burn multiple catches it while a company is still private.

Issue 03 · 2026-05-19

Winners show how many customers stay.

Every public SaaS winner disclosed a retention metric in their S-1. The failures that filed publicly chose not to.

Issue 02 · 2026-05-14

Ads don't save startups.

73% of failures bet on marketing. 18% of winners did. The single biggest gap in the dataset.

Issue 01 · 2026-05-05

One moat is never enough.

Every failure claimed one moat. Every winner claimed two or more.

Every pattern here came from reading real founders and real outcomes.

Find out where yours stands.

Apply