The setting
This guide does not apply to academic researchers, governmental research organizations, or think tanks. You are assumed to work for a private institution whose sole interest in a data analysis is to enhanced decisions. The leaders of this institution care little for the research itself, how it is written, or how it would be viewed by others. If it helps them make better decisions, it has value. Otherwise, it is a waste of time.
In a more limited way it might be relevant to a non-profit or governmental institution with clear, practical objectives.
It is assumed the results will only be read for your organization, and is not for the general public. That is, you are not writing, say, a public relations piece describing to the public your company's strategy for protecting the environment. Instead, you are writing to your immediate bosses and so do not have to worry about politics.
People are busy, and you are supposed to save them time
The recipient of your research likely wants to know a few answers and nothing more. While they do want confidence in your research skills and integrity, they do not want to know how long your worked on the project, all the steps your analysis involved, or the obstacles you faced. They want the most important results in the most succinct form possible. If you waste their time showing off your skills or make them ask you to summarize your results better, they have already lost some confidence in your because you revealed an inability to understand their needs.
Remember, you were hired to save them time, and the more time you save them the higher they value you.
Brevity is essential
If you write a long report, it is absolutely essential that you summarize the essential details, and for that summary to be clear and polished.
Expect your reported to be whittled down as it proceeds up the chain of command. I once visited one of the largest private businesses in the world to observe how they make financial decisions. They start with an extensive analysis, resembling a dissertation. If the financial prospects of the project are auspicious, shorter and shorter summaries will be written as it goes to higher and higher positions. By the time it reaches the president—which may be weeks or years after the research began—the report may be only five lines. The president discusses the project with her most trusted executives, and they come to a discussion quickly. Be prepared for the vast majority of your work to be discarded—but that is a good sign, because it indicates interest at the highest levels!
You are only as good as your best graph
People want and expect graphs. We don't like numbers, but bar, line, and pie charts express much information clearly. While your words will be trimmed down repeatedly, if your graphs are clear and effective they will go untouched. That is why you must make sure your graphs represent your work ethic, skills, and character. People will infer more about you from a graph than your first impression.
While a good manager uses data to help them alter strategies, there are some managers who want data for the sole reason of justifying their strategy. These are not the most effective managers, but they do survive, so expect some managers to care nothing about what the data actually say, and will want you to frame the results to cast them in the best light. What you do depends on you and the situation, but know there is nothing wrong with emphasizing some results over others if that is what your customer wants, but never perform a fraudulent analysis, despite the pressure placed on you. This isn't because the fraud will harm you, though it might, but bad people do often thrive in this world. No, you refuse to make data lie because you value your honor most.
It's all email
Forget everything about how you are taught to write research reports in school. Most communication occurs via email, and while you may often wish to attach a large research report to an email, this email should contain a few other items to aid the person to busy to read a large report.
First, make your emails almost as polished as you would a research report. Typos are easily forgiven in emails, but too many causes others to lose respect for you.
Second, make sure you summarize the report in one or two paragraphs within the email, so the reader doesn't have to open the report if they do not wish.
Third, when possible, always include one to four graphs as attachments to your email. Readers appreciate the opportunity to hear the results of your research in a clear and attractive graph.
Fourth: never, never, never forward emails from other people. Not only is it rude towards the other person, but is illegal.
Fifth, in everything, from your email to your research report to every single graph attached to the email, including your name and contact number as the creator. If a graph elicits much attention and gets passed around, you want everyone to know you are the proud creator, and gives other people the opportunity to contact you for more information.
Defending yourself with a technical appendix
Sixth, although you will only rarely have to defend your research methods—after all, they trust you to perform the analysis correctly, and are too busy to supervise you—always keep a document outlining all your research methods and sources. You may call this a "technical appendix", thereby making it clear who will and will not be interested in the document. For example, if you are charged with reporting how oil prices vary in Texas, Oklahoma, and Arkansas in the last ten years, make sure you document in writing exactly where your price data was acquired. Have the data prepared in a file so that you can email the data immediately. If you alter the data in any way for a graph, make sure you document how the alteration occurred. For instance, if you report average oil prices each month using an Olympic average instead of a conventional average, record this assumption and your reasons.
The idea here is that someone may question your methods, or even question your abilities. If this occurs, you do not want to fumble around for an answer, or say you can't remember, or say you lost the data. Your reputation will be enhanced if you can respond to all accusations by saying, "I have a complete description of my data sources and research methodology in a single document, and will email it to you immediately.
The rules for writing a technical appendix is completely different than the materials you send to your bosses. You want to be detailed to the extreme. Longer is better, as you want to exhaust someone who seeks to discredit you. Because this is a technical appendix, you do not have to worry about the reader understanding it, for you may assume that the reader possesses your knowledge and skills.
