Using AI to create a data analysis report

- It's getting exciting now. We're talking about the second most requested prompt, which as you can see from the title of this lesson, is writing a data analysis report. This is the kind of task that AI tools are continually getting better at. Their arithmetical skills are improving all the time in exactly the same way as mine aren't. However, I feel like I need to give a disclaimer here. If you're planning to use this information for business decisions, double check it just as you would do if you asked an intern to do it, right? Let me tell you about this prompt. You'll start by uploading a spreadsheet containing just the information that you want analyzed. I recommend creating a copy of your original spreadsheet and deleting anything that you won't be analyzing. That's going to reduce the chance of the AI getting confused, working with the wrong information, or losing the plot, because you've just overwhelmed it with too much information. It does happen. You should also check that there's no personally identifiable information or proprietary data in your spreadsheet. You may be breaking the law by uploading any of that. If you need to, you can easily anonymize that information by replacing it with sequential numbers or something else, or you can just delete it. Now add some information in the variables at the end of the prompt to say that the data is in the attached document. Give it some context about what the spreadsheet contains and what you'd like the analysis to focus on. Hit return and it will start as digital pondering. It'll begin by giving you an overview of what it understands from the data. This is how you know that it's doing what you want it to do and focusing on the right information. Next, it gives you a list of insights that it's gained from the data, follows up with a forecast if it applies to your data, and then recommended actions that you can take based on its findings. Then, where onto the added value parts of the prompt. It tells you what assumptions it made, while analyzing the data and gives you a list of additional information that might help it give a more robust analysis. If you have that information, you can give it as a follow up and ask it to run the analysis again. Now, remember that you can carry on asking questions to dig deeper into the data, because a good analysis should spark some questions in much the same way as my dress sense often does. In the next lesson, I'll put it through its paces to show you exactly what it's capable of.