A deck is a list of claims, and everyone in the room knows it. However polished the slides, however confident the founder, a presentation is fundamentally a set of assertions about things that will be true later. A demo is different in kind, not degree. It's a thing that works, right now, in front of you. The gap between a claim and a working thing is enormous, and the strange part is how often founders pour their energy into the wrong side of it.
The reason founders over-invest in decks is that decks are safe. A slide is finished the moment you make it; it can't break in front of an audience, can't reveal that the hard part isn't solved, can't embarrass you with a bug at the worst moment. A demo can do all of those things, which is exactly why it's convincing when it doesn't. The risk you took by showing something real is itself part of the evidence that the thing is real.
There's a tell that everyone in the room registers, even if they couldn't articulate it. When you describe a feature, the listener has to trust you. When you show it working, they don't have to trust you at all — they can see it. Trust is expensive and demos make it unnecessary, which is why a rough working version moves a conversation in a way that a beautiful description of the same thing never quite does. Seeing collapses the distance that telling leaves open.
This generalizes well beyond pitching investors. It's true of customers, of recruits, of anyone you're trying to convince. People have heard a thousand confident descriptions of things that didn't pan out, so descriptions have been discounted in their minds to almost nothing. A working thing hasn't been discounted, because working things are rare and hard to fake. Show the thing, even roughly, even partially, and you've stepped out of the category of people making claims.
The objection is always the same: but I don't have a demo yet, the real thing is too big to show. And the answer is almost always that you don't need the real thing. You need the smallest slice of it that genuinely works — one feature, one flow, one honest demonstration that the core is real. A narrow thing that works beats a broad vision that only exists on slides, because the narrow working thing proves the hard part and the broad vision proves nothing.
Building that narrow slice is harder than making slides, which is the whole reason it's worth more. Anyone can make a deck claiming the thing works. Only someone who actually built it can show it working, and the audience knows this, which is why the demo carries weight the deck can't. The effort you can't fake is precisely the effort that's persuasive, because its presence proves something about you and the thing both.
There's a discipline this imposes that's healthy in its own right. If you commit to demoing rather than describing, you're forced to actually build, early, the part you might otherwise have hidden behind confident language. The deck lets you paper over the unsolved core with a nice sentence. The demo won't let you, because the core either works in the demo or visibly doesn't. So demoing isn't just better persuasion. It's a forcing function for honesty with yourself.
So if you find yourself polishing slides to explain a thing you could instead show, stop and build the smallest version that works. It'll be less impressive than the slides promised and more convincing than they could ever be, because it traded a claim for evidence. Show the thing working, even badly, even briefly. The room changes when the talking stops and the thing actually runs.