From the course: Learning Data Analytics Part 2: Extending and Applying Core Knowledge

Breaking down the ask

From the course: Learning Data Analytics Part 2: Extending and Applying Core Knowledge

Breaking down the ask

“

- [Instructor] Soft skills are invaluable to an effective data analyst. Being an active listener and a critical thinker is not just important, but it's mandatory. You have to be able to break down what's being asked of you into workable requirements for you and the data. In every data request or ask, I have some universal questions that I know that I will go through. I'll confirm where the data is located. Is it a system or Excel spreadsheet? I'll ask if there's multiple places that this data exists, and I'll usually try to understand why that's the case. I'll try to determine the key terms and definitions that the business uses. For example, do you call them customers or clients? And I'll try to determine what format they need to receive the information. Do they need a spreadsheet, the visualization, a dashboard, or is it just a talking point? And who all do I need to include in the results? The other questions may vary by request, and they're meant to confirm my understanding and help me build my requirements. You've probably experienced giving information to someone only to have them say, "Oh, I forgot to tell you," or, "What I really meant was." Just remember, people have more details in their minds and they often don't communicate those details to you effectively up front. That's why it's important to ask the questions. It's our job as a data analyst to pull out the details as much upfront as we can so that our first presentation of the data is as close to what they were originally thinking as possible. Let's go through a few examples of an ask and how we'll break it down. So let's say we're working with a 10 year old company and its data, and the employer asks you to provide a list of customers. This seems really simple, right? But even one request brings up all those questions. Are all the customers stored in the same system for all 10 years? And depending on the answer to this I will document where to get the customers. Also, how do you define customer? Assuming that I don't already know this from properly defined business rules, I'll confirm what they mean by customers. It could be someone who's placed an order or someone who's paid their invoice. And then of course, again, I'll ask, how do you want to see this data: in Excel, in a chart? And then I'll document that response. The answers to all the questions, again, make up my list of requirements. Now let's go through another tricky one. Can we get a list of top performing products? This request seems simple because the question is simple, but it's actually tricky. Now this question might be one that I ask out loud or I just answer to myself based on what I know already, but where are all the products? And two, what do you consider top performing? Sometimes you would think it's just been the ones that have been ordered the most, right, the ones with the highest earning revenue. If we really dig in deep, it could just be the ones with the highest margin, which is not only the list price, but the actual hard cost of the product, and then how much it's been ordered. Of course, I'm going to ask, where are all the sales? Again, this can get complicated based on the age of the company and how they've handled sales over time. I'm going to also ask, how do I get the cost data to determine margin. They may have some different type of calculation for margin. I just need to confirm that. Although the questions will differ based on the request, the who, what, where, when, and why type of questions make up most of the questions I ask the clients at the beginning. Asking questions gives me a better sense of what they're looking for, so I can tailor my analysis later. Remember, data analysts not only analyze the data, but they must also analyze the ask.

Contents