How to structure a talk, pitch, or video so it doesn’t ramble

Rambling feels like a discipline problem. “I just need to stay focused.” It’s not. Rambling is almost always a structure problem — you sat down to record without a shape to fill, so your mouth improvised one in real time, and improvised structure sounds exactly like what it is.

The fix isn’t more willpower. It’s picking a shape before you hit record.

One idea, not five

The single most repeated piece of advice from people who study great talks for a living is that a talk should carry one idea. Not one theme with five sub-points that all deserve equal airtime — one idea, built carefully, with everything else in service of it. If you can’t say what your video is about in a single sentence before you script it, that’s the rambling, not the delivery.

The rule of three

Once you have your one idea, you need a small number of supports for it — and three is the number that shows up again and again in speeches, pitches, and product demos alike. Three is enough to feel thorough and few enough to actually remember. Four points start to blur. Two feel thin. If your outline has six bullet points, you don’t have an outline yet — you have a list of everything you could say, which is a different document.

PREP: a shape for each section

Within each of your three points, a simple pattern keeps you from wandering: state your point, give the reason it’s true, back it with an example, then restate the point to close the loop. It sounds almost too simple to need naming, but notice how rarely rambling videos actually do all four steps — most skip straight from point to example and never say, in one clean sentence, why any of it matters.

Structure isn’t a constraint on spontaneity. It’s what makes spontaneity land instead of drift.

Even the talks that feel effortless are built, not improvised

It’s tempting to think the best speakers are just naturally fluent — that structure is a crutch for people who aren’t. Look closely at the talks people actually remember years later, though, and the opposite is true: they tend to have unusually deliberate shapes. A strong opening that earns attention in the first fifteen seconds. A clear throughline. A close that returns to where it started. None of that happens by accident, no matter how relaxed it looks in delivery.

Writing this into your script

The practical version of all this: write your script as sections, one per point, each one following point-reason-example-point. Not a wall of text — three or four labeled chunks you could rearrange if you had to. This is exactly what a section-based script is for. In Reframe, each of those chunks becomes its own recordable section, so you’re never trying to hold a five-minute structure in your head while you talk — you’re just executing one clean point at a time, in order, and the shape takes care of the rest.