Interesting topic, I think the methods of data analyses all depend on what are the purposes. To set up a goal first and make it target clear is probably the most important step to the data analyses successful. The next step is to the find method to deal with data cleaning, data preprocessing, data manipulation, modeling building, modeling testing and report writing. The final step is to present the data/analyses to the targeted audience, make a good story. From my personal experience, all the three major steps are essential, it is just what the role of you are playing in your team, first one, second one or final one?