Pitch Deck Design: The Structure the Conversation Rests On

The deck is read without you

The defining property of a pitch deck is almost always overlooked: most of the time it is opened when you are not in the room.

It arrives by email, gets forwarded inside a fund, opened on a phone between meetings and skimmed diagonally. The spoken defence comes later, and not always.

From which follows the one requirement everything else derives from: the document has to work alone. If it only makes sense with your commentary, it is not finished.

One slide, one claim

The most effective edit to any deck takes an hour and needs no designer.

Put the slide's conclusion in its headline. Not "Market" but "The market doubled in two years". Not "Team" but "We assembled people who have done this twice".

Someone who flicks through the headlines alone should receive the whole story. Everything else on the slide is evidence for that headline, not a second claim.

What the main body consists of

The set of slides is not dogma, but there is a sequence an investor expects, and departing from it is worth explaining.

The problem and why it exists. The solution and how it differs. Market: size, movement, where the figures came from. Product: what already works. Model: how you earn. Traction: what has already happened. Team: why you specifically. Competitors. The plan and what the money is for.

Ten to fifteen slides in the main body, then an appendix of any length. The main body has to read in three minutes without your voice.

Each slide answers exactly one question. A slide answering two reads as answering neither, because the reader does not know where to concentrate.

And order matters: traction after the model reads as proof; traction before the model reads as numbers without meaning.

What to do with the team slide: Design here governs reading speed, not impression.

This slide is most often made perfunctorily and most often read carefully, because at an early stage people invest in people.

What does not work: a list of job titles with photographs and the words "ten years of experience". That is identical everywhere and communicates nothing checkable.

What works: what each person has specifically done that bears on this task. Not "strong marketer" but "launched sales in this category in three countries". Verifiability beats resonance.

It is worth showing the gaps honestly. A slide saying "no technical lead yet, hiring" earns more trust than a perfect picture nobody believes.

And the order of people on the slide: first the one responsible for the riskiest part. That is usually not the person speaking in the meeting.

Figures instead of adjectives

"Fast-growing market" and "unique technology" mean nothing, because nobody writes the opposite.

What works is checkable: how much, over what period, from which source. Even a modest figure with a citation persuades more than a large adjective.

The same applies to risk. A slide that names the main obstacle honestly earns more trust than a deck in which everything is smooth.

How to present figures

Figures are the most common place where a deck loses trust, and almost always for one of three reasons.

First, no source. Any number without an indication of where it came from reads as the author's estimate. One line of small type under the chart settles it completely.

Second, the chosen window. A chart starting from a convenient month is recognised instantly and raises questions about everything else. Show the whole period you have, even where it is uneven.

Third, percentages without a base. "Three hundred per cent growth" without absolute values means the absolute values are small. The reader knows that, and the omission works against you.

The general rule is simple: present a figure the way you would to someone who will later check it. They will.

What design actually does here

In a presentation, design governs reading speed rather than impression.

It sets priority inside a slide: what is read first, second, and what can be skipped. It holds one type scale so the eye does not rebuild the hierarchy on every slide.

And it removes what does not earn its place: frames, shadows, decorative icons, background images under text. Each of those takes attention away from the figure the slide exists for.

Why a deck has two layers

A deck serves two readers at once, and that is its central structural difficulty. The first decides in three minutes whether a meeting is worth it. The second has already agreed and wants to check what the conclusions rest on.

Writing one document that satisfies both is impossible: a short one reads as shallow, a detailed one does not get finished. So the deck splits into two layers — the main part and the appendix.

The split is not cosmetic. It decides what reaches a slide: what changes a decision goes into the main part, and everything else goes downstairs and stays there until the first specific question.

The appendix: what to move there

An appendix is not a dumping ground but an instrument for managing attention. The conclusion goes in the main body; what supports it goes in the appendix.

That means the market sizing method, the detailed financial model, full team biographies, technical architecture, legal structure, competitor detail and product history.

Referring to the appendix directly from a slide helps: "methodology on slide 21". It shows the detail exists while not requiring everyone to read it.

An appendix can comfortably run twice the length of the main body. It is read selectively and only by people who have reached serious interest.

Which is exactly why nothing the main body depends on can be moved there: not everyone reads the appendix, and nobody reads it first.

One deck or several

A deck is often pushed toward being universal: suitable for sending, for presenting and for clients alike. The result is a document that works poorly in all three modes at once.

The differences are substantial. A deck that gets emailed carries more text, because you are not there. A deck for speaking carries the minimum, because you are talking and the slide supports you. A client deck has entirely different content: not market and model, but what the client receives and on what terms.

The practical answer is one source and two or three builds. The bulk of the slides is shared, while each version hides some and rewrites some headlines for its mode.

That costs less than maintaining three separate files, which inevitably diverge: a number gets corrected in one and stays stale in the other two.

And it is easier for a team: with three versions but one source, nobody sends outdated figures by accident.

Checking a deck before sending

Four checks, a few minutes each, catch most of the problems.

First: read only the headlines in order. If they form a coherent story, the deck works without you. If they form a list of topics, the headlines need rewriting into conclusions.

Second: open it on a phone. A significant share of first views happen there, and text that looked fine on a monitor is often unreadable.

Third: print it in black and white. That catches slides whose meaning rests on colour: two colours of similar lightness merge on a chart, and the chart stops showing anything.

Fourth: give the deck to somebody outside the project and ask them to retell it. The gap between the retelling and the intention is exactly what an investor will see.

How to send the deck

A file that did not open equals a deck that does not exist. Send one document in a format that opens everywhere without installing anything, and do not assume the recipient runs your editor.

The file name is what the person will see in their downloads a month later. "Company name — deck — month and year" is findable by search; "presentation_final_v7" never is.

Weight matters: heavy attachments get cut by mail filters and will not open on a phone. If the deck does not fit in an email, a link beats an archive — but the link has to work without a sign-up and must not expire in a week.