At the end of a typical research phase, your brain will be very full. Your hard drive might be, too, with dozens of pages of notes, dozens or even hundreds of photos, and possibly many hours of video or audio recordings. All that information can be overwhelming and hard to use, so you have to condense and massage it into some kind of structure that makes sense. Models are excellent tools for doing this.
A model is a description that helps people understand and communicate about observed behavior. Bohr's model of the atom and Freud's ego, superego, and id, for example, give us frameworks for understanding the complex ideas they stand for. Similarly, modeling the results of your research will help you condense and visualize information to understand human behavior patterns, workflows, and trends.
The primary objective of modeling (and the subsequent phase, requirements definition) is to enable informed action. You're not only trying to crystallize your own understanding; you're also trying to help the entire product team build a shared view of the problems, opportunities, and potential next steps.
Like design research, analysis should be rigorous yet efficient, focusing on the aspects of the data that will facilitate design and business decisions. The time you spend on analyzing and modeling should be commensurate with the size of your data ...