Choose your topic and focus. The scope of your topic can be broad, but a specific element must be the focus. For example, your topic can be the ethical practices of big oil companies, but your focus needs to be something involving a specific issue, such as the 2004 Royal Dutch Shell Oil shareholder scandal in which the company deliberately misled shareholders about the amount of oil reserves, thus compromising the future value of shares.
Define your problem statement. Here is where you identify exactly what you want to investigate about your topic and focus. Your problem statement must be extremely specific and written in measurable terms. It can't always be written as a question. Consider the following: If your focus is the Shell Oil shareholder scandal, then your problem statement could be something along the lines of, "Did the Shell Oil shareholders scandal have a negative lasting effect on the general public's impression of the company?"
Write a brief history of your topic, focusing on the specific area of your topic. For example, if you write about the ethical practices of oil companies, include some examples of ethical and non-ethical behavior from major companies. Then write about the company on which you focus, including the company's history and information on its ethical practices.
Design your method of data collection. This includes deciding how exactly you will gather your data. For case studies, it is likely you will use mixed methods--a combination of quantitative and qualitative data. If you measure the public's impression of the oil company, you might collect opinions of individuals (qualitative data) as well as examine share prices and retail sales of the company before and after the scandal (quantitative data).
Analyze your data. Examine the data you collected, and organize it in a meaningful way. That is, find meaning in your data by applying some kind of coding system to your qualitative data. For example, use a scoring system to measure negative and positive responses from your interviews. For your quantitative data, you can use specialized software to run statistical analysis.
Explain your findings, drawing conclusions about your experiment. Revisit your problem statement and answer the question if applicable. If you investigated the general public's impression of Shell Oil after the 2004 shareholder scandal, then address that explicitly and describe how your findings apply to the problem statement.
Combine your introductory history, description of your data collection, analysis and conclusions into one cohesive document. Ensure it has some kind of storytelling aesthetic; it should leave the reader with a full understanding of the broad topic as well as the specific nature of the research conducted.