While the structure of a talk reflects a lot the style of the
presenter, there are nevertheless some useful guidlines that can be
followed. One way of presenting a paper is to follow the structure of
the paper itself. As most papers follow the structure
1) Introduction
2) The model
3) Equilibrium
4) Results
4.1) First results
4.2) More results
5) Conclusion
the talk could also follow this structure.
Clearly, a talk should (can) also include some entertaining elements
and should focus more on general understanding and less on detailed
proofs (though this might also be interesting for certain types of
papers). The main point of the paper should be stressed in the
introduction, in the main part and in the conclusion.
An overview of the contents of the talk at the beginning might also be useful. A discussion of the related literature is useful in most cases. What does the paper borrow from others, where does it differ?