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

Building dashboards

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

Building dashboards

“

- [Instructor] The primary goal of any dashboard is to be able to glance and get an idea of what's happening with the organization, or even within a process like sales, for example. The easiest way to think of a dashboard is to imagine the dashboard of a vehicle. The dashboard is a simple interface that gives us an idea of what's going on with the vehicle. There's one part of the dashboard we can all think about, and we use it quite regularly, that's the gas gauge. We know when it's full and we know when it's low, and we make decisions for when we stop for gas. There are lights that only light up when there is an issue with a car, like tire pressure, or even oil engine lights. This information tells us we should either stop driving, or drive for service to explore what's going on, and that's exactly how a dashboard should work. There are lots of things going on behind the scenes of a car that are supplying data to that dashboard. And even though we don't see it, we do see the end result. If you haven't ever designed a dashboard, and you're used to printed pages, well, building the first one will be the toughest, and then you'll discover, like everything else, it's a process. In this chapter, I'm going to walk you through the steps I've taken building a dashboard, no matter what tool I'm using. I'll gather and clean the data that's needed for the dashboard. I'll start building what I call dashboard artifacts. This allows me to explore the best ways to visualize the data to answer the questions. I'll build a sample dashboard, and I'll explore the user experience through the filters that are supplied, and then I'll refine and publish the dashboard as the final step. These are my steps, and they're pretty universal and pretty consistent, regardless of the tool that I use. There are other courses in the library that will have more suggestions for you to explore. And at the beginning, just having some form of process is better than no process. Before you ever put the first artifact on your dashboard, you'll have needed to spend time up front with your data. This means connecting to it and getting that data to the clean state. And for our examples here, we're using Power BI Desktop, and Power Query is the data prep tool that we'll use. Dashboards are visual expressions of the data, and they're meant to be interactive. That means they're not great for thousands of lines of data. Those will be more appropriate for paginated reports. When you're designing a dashboard for the first time, you may not know the best way to visualize the data. I get around this by creating what I call dashboard artifacts. And once I find the one that tells the data story in the clearest way, I'll pop it onto my sample dashboard. The world is full of paginated reports, and these are reports with lots of lines. You will find them PDF or even printed, and they're produced with tools like Power BI Report Builder, SQL Server Reporting Services, or even Crystal Reports. These are multi-page reports, and they're not always positive interactive experiences. And when laying things out on a dashboard, screen space is limited no matter how big your screen may be. So choosing what goes where is as much an art as it is a science. And you should explore the library for more information on the best ways to visualize your data, not just building dashboards. Because dashboard tools are meant for the user to experience instant gratification in exploring the data, most data tools have different layers of filters. Soft filters, as I like to call them, are filters that are applied by the user using slicers or even interactive visuals. In filters that we place on the page or on the actual visual, these are called hard filters, meaning that they're hard-coded into that visual or into that page. As I work through the filters, I'll save that soft filtering editing process as my next to last step. I'll set the final settings for those filters, and edit the way they interact with other visuals. If I had one piece of advice to follow about any layout, it's that you need to save the intricate sizing and formatting to the very end, because the objects will shift a lot, and you can spend a lot of time formatting and reformatting during the process. I save that intense attention to detail for when I know where everything is going to be on the dashboard. Trust me now and thank me later. Don't go crazy and build something unreadable. Just don't do all that daunting end work of sizing until you get to the end. When building a dashboard, never lose sight of the purpose of the dashboard. Your process may be a little different than others, but the goal is to have a process that helps you build effective dashboards. Your dashboard should provide insight to the decision-makers and the consumers of the information. Dashboards are not only valuable to your organization, but being able to build them is a very valuable skill.

Contents