Speaker 1: It's easy to create a data visualization by just throwing data into a graph and publishing it. But how do we make professional-looking visuals that are attention-grabbing, easy to read, and clearly convey the data we want them to? There are several design concepts we can apply to our dashboards and reports to make them stand out from the average visual. Let's take a look at how we can make our data look great. One of the basic design concepts to consider is balance. Often we cram as many visuals into whatever screen space we have, which results in unreadable chaos. One use of balance can be alignment, to simplify where the reader has to look. Repetition, which can help the eyes glance across the data easily. Contrast to highlight or clearly display differences. Hierarchy to show importance. Symmetry, which is similar to alignment in reducing the feeling of chaos or clutter. And intentional imbalance, to show tension or contrast from one set of data to another. Along with balance, most visuals need much more white space to reduce the clutter and chaos. If several visuals are conveying the same information for the same KPI, reduce it to just the one that most clearly represents that KPI. Try to limit each screen to focus on a specific topic, and use interactions to drill into deeper or adjacent data. Our brains love patterns. We always look for them and then use them to speed up our understanding of what we're looking at. Use this to your advantage. Try and use the same color patterns to represent the same data across all your visuals. The most common is green for good, red for bad. Also common is using a company's brand color to always represent that company anytime it has data. Also make sure you're using the same type of visual every time you want to display the same information. If you're using a scorecard to show defect percentage, don't switch to a bar graph the next slide it comes up, and then a pie chart the time after that. Make sure your data is ordered consistently, such as your bar graphs being ordered in the same pattern when possible. Choosing colors can always be one of the hardest parts of design. Some basic rules are to use contrasting colors in data so it's very clear when one ends and another begins. There's a lot of color palette tools out there, but colors.co is a good one to start with. You can explore popular palettes and try some out with your visuals, but be sure not to overdo it. Try not to use five or so colors in a single visual. If you have a lot of bars or lines in a chart, use shading rather than a different color for each. Consider reserving color differences to highlight the extremes of data. So what about text? How much should we put in our visuals? The goal is just enough to guide the user through the data, but not so much it complicates it or demands more attention than the data. When it comes to fonts, don't use too many. Try to stick to two at the most and make sure they are easy to read. Be sure to always label your axes, including units if there are any. Beyond that, add clear and concise text where a user might need more information. Be aware that our eyes will focus on the text first, so don't put text where it will draw the focal point away from key information. When possible, put labels directly on the lines, bars, or data instead of off to the side. And last is making sure our visuals are clear themselves. Be sure you're using the right type. Let's say we have 10 products we want to show sales numbers for. Most commonly people will use a bar or pie chart. We're much better at processing the difference in size between bars than we are the proportions in a pie. Let's look at some example visuals. At a quick glance, what information do you take away? Could you tell which products are performing well and which aren't? What about with the pie chart? Could you take in the specifics on performance over time? With the sparkline, you quickly see the trend, but not much else. Does the fancy 3D help or hurt the processing of information? Does this common graphical display get the information across any better than the bar graph? It doesn't hurt to use this quick test with a visual to make sure the right information is getting across. With these easy to implement design concepts, our visualizations can be more appealing and more accurately convey the data to the users. Thanks for watching. If you enjoyed this video or learned something, a thumbs up would be really appreciated. Stick around for more data content by subscribing to the channel or clicking a video on screen. See you in the next one.