Where the Numbers Live

Every set of fundraising materials is really two key documents having a conversation.

There is the deck, which makes the case. And there is the model, which proves it. They are read by the same person, often in the same hour, and they are judged as one thing even though they are built as two.

Most of the trouble I see starts here. Not with the quality of either document, but with the question of what belongs where.

The instinct to put everything on the slide.

It comes from a good place.

You have done the work. The model is detailed, the assumptions are considered, and you want the reader to see that. So the numbers migrate onto the slides. A revenue build appears. Then a cost breakdown. Then a sensitivity table, shrunk to 8 point because it has to fit.

By the third of these, something has quietly gone wrong. The deck has stopped making an argument and started reproducing a spreadsheet, badly. Nobody can read it. Nobody can check it. And the one thing it was meant to do, which is carry the reader forward, has stalled.

A deck is not a smaller model. It is a different instrument.

What each document is actually for.

The deck exists to make a proposition a reader can hold in their head.

It should tell them what the business is, why the opportunity is real, what you are asking for, and what happens to their money. It should be readable in one sitting by someone who is tired. Every number on it should be there because the argument fails without it.

The model exists to let a reader interrogate that proposition.

It is where the assumptions live, where the mechanics are visible, and where someone can change a figure and watch what happens downstream. It is not a persuasion document. It is a testing instrument, and it should be built to be pulled apart.

Confusing the two produces the 2 failures I see most often: a deck that reads like a printout, and a model that has been tidied up for presentation until it can no longer be tested.

The rule I use.

A number belongs on a slide only if the story breaks without it.

Everything else belongs in the model, referenced but not reproduced.

That sounds austere, and in practice it is liberating. It means the deck can breathe. It means the handful of figures that do appear carry real weight, because they are not competing with 40 others. And it means the reader who wants to go deeper knows exactly where to go.

The lender or investor who wants the detail will always ask for the model. They were going to ask anyway. Putting a compressed version of it on slide 11 does not save them the trip, it just makes slide 11 unreadable.

Where they have to agree.

This is the part that is not optional.

Every figure that appears in both places must be identical. Revenue on the slide is revenue in the model. The ask on the front page is the ask in the sources and uses. The margin you quote is the margin the model produces.

It sounds obvious. It is also, in my experience, the single most common defect in materials that arrive for review, and it happens for an entirely innocent reason: the deck gets built, then the model gets refined, and nobody goes back.

The cost is out of all proportion to the cause. A reader who finds one number that does not tie stops trusting the rest. They are not being unfair. They have no way to know whether they have found the only error or the first one.

The order I build them in.

Model first, wherever the timeline allows it.

Building the model first means the story is shaped by what the numbers actually support, rather than the numbers being bent to fit a story that was written before anyone checked. It also means the deck can quote with confidence, because the figures are settled.

When it happens the other way round, and sometimes it has to, the discipline is to treat the deck as provisional until the model is finished, then go back through it line by line. That reconciliation pass is unglamorous and it is where a great many problems get caught.

The walkthrough matters more than either.

There is a third document nobody talks about, and it is the one in the reader's head.

A model that only its author can navigate is not much use to a credit team. Neither is a deck that needs you standing next to it. Part of the job is making sure both documents can be handed over and understood without their author in the room, because that is exactly what happens: they get forwarded.

That means labelled tabs, visible assumptions, and a summary that lets someone who will never open the detailed sheets take away the right conclusion. It means a deck that survives being read cold on a phone.

Closing thoughts.

I think of it as one job with 2 outputs.

The deck earns the reader's attention and tells them what you are proposing. The model gives them the means to satisfy themselves it holds up. Neither works alone. A beautiful deck with nothing underneath it gets found out at the first serious question, and a rigorous model with no story around it never gets read at all.

Get the division right and both documents get easier to build. The deck stops trying to be a proof, and the model stops trying to be a pitch. Each does the thing it is good at, and the reader moves between them without friction.

That is really all any of this is about. Making it possible for someone to understand your business, check your reasoning, and reach a decision without having to work for it.

We don't write to go viral. If you're reading this, you've probably already heard of Decksadu. These short articles are simply a place for us to think out loud - about the art of helping extraordinary people build exceptional businesses.