Suspense is a powerful tool. In story-telling, not knowing what is going to happen or how the story is going to end can grab attention and sustain interest. In giving a research presentation, one is often encouraged to "tell a story". One could certainly make use of suspense to do so.
We are interested in how best to predict X. Previous research has indicated that P Q and R could be crucial here. [much later] It turned out that... [pause for drama]... variable K was surprisingly the strongest predictor of X!
However, most presentations take a more up-front approach.
Welcome to my presentation entitled "How K can be used to predict X"...
The latter approach follows the advice that I have often received, which can be summarised by the maxim:
Tell them what you are going to tell them, then tell them, then tell them what you have told them.
So, is it a bad idea to use suspense with the intention of making the "story" of your presentation more engaging?
[This question was inspired by an answer given elsewhere on Academia.SE.]
I believe that using suspense in a talk is a terrible idea on any but the shortest rhetorical time scales.
The reason is that, at any given time, a significant fraction of your audience will not be paying attention. It doesn't matter whether they're twiddling with their phone or just thinking about what you said two slides ago: the point is, it's terribly easy for people to miss bits of a talk. As a result, if you build your talk as a suspenseful mystery where somebody has to be paying attention at the right times, then you'll probably fail to convey key information to a significant fraction of your audience.
Now, there's nothing wrong with using it in the short range as a bit of rhetorical flourish, e.g., "And what did we find? The wires were covered with purple spots!" Once you start putting long-range dependencies in your talk, however, you're gambling with attention. Likewise, even without putting in a dependency, if you say your key idea only once in your talk, you're decreasing the number of people who will hear and remember it, compared to if you say it in "preview" and again when you get to it in detail.
In short: don't hide information from your audience. There's plenty of other, more effective ways to make a scientific presentation interesting.