# Pitch deck slides, by what each one does

> 24,182 slides from 1,708 pitch decks, sorted by what each argues (problem, market, traction, the ask), with what each kind of slide has to do.

Source: https://millionrounds.com/slides (Millionrounds; every figure on the page links to its filing or article)

Slide Database

# Every slide, by what it *does*.

A deck is the same argument every time, so the useful unit is the slide. Cut apart, a problem slide sits beside hundreds of other problem slides from companies that went on to raise, and the convention becomes visible. So does where breaking it worked.

Open the slide databaseRead the playbook

**Cover** **Product** **Solution** **Validation** **Market** **Problem** **Go to market** **Vision** **Cover** **Product** **Solution** **Validation** **Market** **Problem** **Go to market**

**Cover** *1,771* slides, in 100% of decks **Problem** *893* slides, in 35% of decks **Solution** *791* slides, in 31% of decks **Vision** *548* slides, in 24% of decks **Market** *2,083* slides, in 62% of decks **Product** *1,897* slides, in 52% of decks **Go to market** *448* slides, in 19% of decks **Validation** *979* slides, in 36% of decks

**24,182** slides cut out of 1,708 decks

**14,508** sorted by what they argue

**14** slides in an average deck

**16** kinds of slide, each with a playbook

The playbook

## What each slide has to do

Pick a slide. Written once per kind and adjusted in the app for the size and sector of the company: a problem slide at $300K raised is judged differently from one at $30M.

Cover slide

Say what the company does before anyone has to ask.

**1,771** slides **100%** of decks carry one Where it sits first last

What a reader looks for

- A sentence a stranger could repeat correctly afterwards
- The company name, and what category it is in
- A date, so the reader knows how current this is

The usual mistake

A tagline that sounds like a brand and explains nothing. If the reader has to reach slide three to learn what you sell, the cover has failed.

The gaps

## The slides most decks skip

Of the 1,708 decks we could read, the share that never include these. Each is a question an investor will ask anyway.

81%

**never show the ask**

Say what you are raising and what it buys.

74%

**never show a traction slide**

Show the thing is working, with numbers that cannot be reframed.

84%

**never show a business model slide**

Explain how money arrives, and whether it keeps arriving.

## How the slides are sorted

- 1 **Cut apart** Every deck is split into its pages, one image per slide.
- 2 **Read** Nearly all decks are pictures, so the words are read off each image.
- 3 **Sorted** Each slide is labelled by what it argues. Where the reading comes back as noise, it stays unlabelled rather than guessed: 9,674 slides so far.
- 4 **Judged** Each kind carries its playbook, adjusted for the size and sector of the company that made it.

The slides belong to the companies that made them, so reading one needs an account, and any company can have its deck removed by asking.

## Read them side by side.

Every problem slide from fintech companies that went on to raise, in one screen. Filter 24,182 slides by what they argue, sector and size.

Open the slide database[Or browse the 1714 decks](https://millionrounds.com/decks)