The deck is the last thing you build.

Ask a founder preparing to raise what they are working on and the answer is almost always the deck. It is the artefact the whole process is named after, the thing investors ask for, and the obvious place to start.

It is the wrong place to start, and starting there is the direct cause of the two failures that show up most often in diligence: documents that disagree with each other, and a valuation nobody in the company can actually defend.

A deck is a compression, and compression needs something to compress

A pitch deck is not an argument. It is a compressed presentation of an argument that already exists somewhere else, in longer form, with evidence attached. Purpose, problem, why now, market, product, model, go to market, proof, competition, team, financials, the ask. Every one of those slides is a summary of work.

When the deck is built first, there is nothing underneath it to summarise. So the slides get filled with what feels right at the time. A market figure from a report. A revenue projection built to look like a curve investors like. A valuation someone suggested over coffee. Each is plausible in isolation, and none of them is connected to any of the others.

Then the model gets built, and it disagrees with the deck. Then the memo gets written, and it disagrees with both. Nobody set out to be inconsistent. The documents were simply written at different times against different assumptions, and there was never a single source for any of them to be wrong about.

An investor who finds the model and the deck disagreeing has learned something about the company, and it is not what the company intended to teach.

The order that works

Build the case. Price it. Structure it. Then assemble the pitch. Then raise.

01Build the case

The investment summary, the memo, the market work built bottom-up, the financial model, and an honest stress test of all of it. This is the substance, and it is where essentially all of the thinking happens.

02Price it

Only now is there enough evidence to value the company: real revenue or its absence, a defensible market, a model with assumptions someone has argued with, and comparables researched properly. A valuation set before this point is a guess that everything downstream then has to accommodate.

03Structure it

Round size, instrument, minimum ticket, allocation, dilution, direct or pooled. The ask is a consequence of the case and the price. It cannot precede them, although in practice it usually does.

04Assemble the pitch

Now the deck and the teaser can be built, and every figure on every slide traces back to something defended. The deck becomes the easiest part of the process rather than the hardest, because there is nothing left to decide.

What this changes in the room

A deck built last has a property a deck built first cannot have: every number on it can be defended in the meeting, immediately, by whoever is presenting. The market figure has a build behind it. The valuation has a comparable set. The projection has a model, with assumptions someone has already attacked.

That is what investors are testing when they push on a slide. Not whether the number is impressive, but whether the person presenting knows where it came from. The founder who can answer has demonstrated something no amount of design achieves. The founder who cannot has also demonstrated something, and it is usually the end of the conversation.

This is also why the sequence is not a matter of taste. It is why the deck and the teaser are the final artefacts of a mandate rather than the first, and why a company that arrives asking for a deck is usually asking for the wrong thing at the wrong moment.