When I approach a Data Sufficiency question, my goal is to squeeze as much information out of the question stem as I possibly can. Ideally, I'll be able to identify the pieces of information that the statements must provide in order to be sufficient.
For example, if a question stem sets up a percent change problem, I know that there are pieces of information involved: starting point, ending point, and amount of change. I need two of these in order to find the percent change.
By figuring out what the statements must provide before I even look at them, I've made the evaluation process a lot easier.
Give it a shot and let me know what you think!