While Stephen Walt is discussing writing, I think this applies to speaking as well (boldface mine):
A second reason is the failure of many scholars to appreciate the difference between the logic of discovery and the logic of presentation. Specifically, the process by which a scholar figures out the answer to a particular question is rarely if ever the best way to explain that answer to a reader. But all too often articles and manuscripts read a bit like a research narrative: “First we read the literature, then we derived the following hypotheses, then we collected this data or researched these cases, then we analyzed them and got these results, and the next day we performed our robustness checks, and here’s what we’re going to do next.”
The problem is that this narrative form is rarely the best way to make a convincing case. Once you know what your argument is, really effective writing involves sitting down and thinking hard about the best way to present that argument to the reader. The most important part of that process is figuring out the overall structure of the argument — what points need to be developed first, and then what follows naturally or logically from them, and so on. An ideal piece of social science writing should have a built-in sense of logical or structural inevitability so that the reader moves along the argument and supporting evidence as effortlessly as possible.
Achieving this quality requires empathy. You have to be able to step outside your own understanding of the problem at hand and ask how your words are going to affect the thinking of someone who doesn’t already know what you know and may even be inclined to disagree with you at first. Indeed, persuasive writing doesn’t just convince the already-converted, a really well-crafted and well-supported argument will overcome a skeptic’s initial resistance.
This is a hallmark of bad talks: the speaker describes the journey, but without first providing the end result. This isn’t a mystery novel: you have to provide context for the information you’re providing the audience, otherwise it’s just a collection of unassociated facts. Don’t give your talk backwards.