From the course: Data Visualization: A Lesson and Listen Series

Listen with Stephanie Evergreen

From the course: Data Visualization: A Lesson and Listen Series

Listen with Stephanie Evergreen

(bright music) - Am really happy to have Stephanie Evergreen with me today. She's the CEO of Evergreen Data, across industry data visualization and design firm, grounded in research-based best practices. Stephanie has a PhD in interdisciplinary research, and she did her dissertation on how to communicate about information and research. So, her expertise is exactly what we want for this conversation. Oh and she spent plenty of time creating visualizations in Excel and writing a ton of blog posts, sharing the tips and tricks that she's developed. Stephanie, thank you very much for being here with me today to talk about this fun subject. - I'm so glad to be here Bill. - All right, cool, so we're going to start right at the top, all right. Excel is our subject for the day, and so I'm just going to ask you, you know, when you think about Excel's, visualization tool specifically. What do you love about it? What frustrates you about it? - Yeah, so at the time that we are taping this interview, there's a story on the news, you probably saw it Bill about something like 16,000 coronavirus cases, that didn't get reported because they were stored in Excel inappropriately. And so, the headlines are all like, Oh, only fools would be using Excel. And every once in a while, we see these stories about PowerPoint to Excel getting kind of kicked around in the news. But for me, it's never the tool, it's always about how we use it and that's kind of why I love working with Excel. Everybody has it, I mean, especially those scrappy folks who are working in public health or nonprofits, they're we're relying on the tools that they have to do the work. And Excel can do great data visualization, if you just know what buttons you need to push. So, it's learnable, it's doable, everybody's got it. And you can make great charts, charts that aren't necessarily defaults if you just know how to hack into it a little bit and I guess that's part of why I love it. I don't know if they're doing that on purpose or not, but you can sort of break Excel and make it create awesome database. But I guess the flip side of that, is that it sets the part that needs to be improved, right? We shouldn't have to break into the software to get it, to make great stuff. The defaults should just be better than they are, especially when you look at the growth of data base over the last 10 years. There's just no reason that our defaults aren't stronger. But that said, I mean, I have that beef about every one of my graphing software. - Yeah, that was a great points. And, you know, yeah, it's the tool that everyone has. And so, we should learn how to use it. And the problems that it has, as you said, they're solvable, you can make weird, different, unique bespoke type of charts by hacking into it. You know, the issue about the part of defaults is interesting too, because the defaults used to be really bad. And they have improved them, they've gotten slightly better, and so hopefully down the road, they'll be even better. Which brings me actually to my next question, so, you know, if you could get the ear of the Excel team at Microsoft and say, "Hey guys, come on, can you fix these three things?" You know, what would they be? What is your sort of top priorities, that you wish that they would improve upon? - Sure, so I actually have a channel on my Instagram account. That is my love-hate letter, to all my various graphing software. So, if Microsoft is listening, I already have several of these complaints sort of detailed out, but one really, you know to just... Our previous point about the defaults, they should just be better. There's no reason to have 3D graphs even available to people anymore. I mean, and we could replace those with more high-impact graphs so they currently have to hack to get, so just better default would be great. I would also love easier exporting of high-res images, I mean, so often our graphs are going into some beautiful displays, some kind of report or something where they need to be high quality. And right now, I have to use some kind of plug-in or some screenshot tool to get high-quality images out. It seems like it ought to be a little bit easier for us. I'd also say, easier setting, interactive graphs online. You can do some of it now it's just kind of rough around the edges and some of the interactivity features don't carry over. So, I mean, that's just the reality of our reporting needs these days, they should be making it easier for us. And if I can add one more, I would ask for just a more consistency between the Mac and PC versions. My theory is that it's something about, you know, cooperate level conflicts between Microsoft and Apple. But the Mac version is always just a little bit behind the PC version. And I do a lot of teaching workshops, teaching people how to make great databases and the people who run Macs, they're always just a little bit more frustrated than they should be. Because the menus are called different things, the menu options are in totally different places. It's just, it's not fair. - Yeah, so you just taught me something. I didn't know that you could do embedding from Excel, that's great to know. And you know, to do high-quality images, my default is I copy and paste into illustrator. So, I have to take it into illustrator which is nice, but boy, that's a pain in the butt. And then yeah, the Mac PC thing, I mean so the Mac version of Excel for those of you don't know, doesn't have power query, which is one of the most amazing features of Excel that most people who use Excel don't even know exists. It allows you to do data transformations, which by the way, would have solved that whole coronavirus error 'cause you could transform the data instead of having it... That was a column limitation problem, instead, you could use the million rows instead of the limited 16,000 columns, you could sort of set up a transformation process and it would automatically pull in the data the right way I could go on and on. But yeah, so good stuff and those are four good asks and we didn't tell you, you could have four, but I'm going to give it to you, because why not? And hey you know, Microsoft owns LinkedIn, so you never know they could be listening. (Stephanie chuckles) - Listen to us. - So yeah, seriously maybe there's a way we can sort of do the back channels to get to them. So, yeah, like any tool as you said, Excel does have limitations. But yeah, it is the tool that we all use regardless of role, industry, experience level, every single human in fact, I'd love to know the numbers. Like what percentage of humans with computers have Excel use Excel, it's going to be ridiculously high. So, on the one hand, yes that's the tool that we're going to use and we should strive to do it better for data visualization, which has challenges, but yeah, we can do it. On the other hand, you could say, well, maybe we're asking too much of Excel. Maybe we shouldn't try it be, the tool that does all things for all people. So, I guess the question here is, should we just encourage people to use different tools to do this work and you know, why or why not? - Yeah, that's such a great point. You know, I think Excel could be called a pretty decent Jack of all trades, but specialist at none, you know, it really can do a little bit of everything. And I think that's actually why it's so appealing to so many people, especially those who can't afford to have a tech stack. I know, most of my clients are in that situation where they might like a better tool for data cleaning or data analysis, but their budgets are small. So, they're working with what they've got, or they might want to have something like R, that can do it all for free, but they just can't slow down long enough to learn how to code. So, they're working with the tools that they have and that's always my recommendation too, is that we just mastered the tools that we've got and once you have achieved some mastery, that will be the point when you know, whether you need a different kind of tool to do that specific job better. So, I think it's going to be around, I think we just have to get used to using it. One of the reasons that I appreciate it so much is that it can create some independence between me and my clients. So, because they all have it and because they more or less know how to use it, it means that, you know, I might make a chart for them. And then when the data update next year, they can just change the data in the spreadsheet instead of having to come back to me and ask me to write up a new contract with them and give me more money, so that I can just update a chart. They have that ability and that power to do it on their own. - Yeah, that's a great point. You know, we use this tool, as you said, clients are using it. It is the tool we go to, it's probably not going to change much, you know, it is a giant corporation, it's going to change slowly over time. So yeah, we need to use it and everybody does, and so, the question is how can we make our visualizations more impactful? How can we sort of start down that road of doing it? And I gave a very brief flash lesson on that subject, which, you know, plays on this course. And I pointed out a few things, but it was incomplete as of course it would be in a five minute lesson. And maybe not even a well prioritized list, so, I wanted to ask you, what's your top advice, for getting more from Excel in terms of visualization? The one thing you would just say to our audience, which includes, by the way, the people you're talking about their HR people, IT people, yeah, some data analysts and some, you know, people from all walks of life, all skill levels, all industries. - Well, I would say that if you're using it internally, anything goes, right? If you're the only one who's going to be looking at something, you're fine to just use it however you wish. But anytime you're going to start communicating that data out to someone else, we've got to change up those defaults. My golden rule here really is, that no one should be able to tell what software you use to make your visuals. Then that goes for any software that's out there, no one should be able to look at it and be able to detect the software that was used. So that is going to mean in Excels case, you're always going to want to change the fonts and colors. Even if you take an extremely simple, basic bar chart, but you change it away from calibri, and a way from that Excel default blue, it's going to take you so far. Beyond that, I mean there are all kinds of little formatting tweaks that you can make to take a default chart and make it even better. Of course, there's even more, you could do to hack your charts to make even cooler ones that tell better stories. And that is all learnable, I mean I think it's really only a couple of hours of training can get people to a place where they're taking those Excel defaults and doing something amazing with it. - Yeah, so you've been doing this for a long time and you've been changing these defaults and transforming Excel charts into bespoke better, more wonderful things. What are you most proud of? Something you've actually accomplished? You know, maybe particularly esoteric chart that you've been able to hack, as you said, or maybe some really highly nuanced annotations you're able to just make something really come to life or some other elegant solution that you come up with. What's the best example of artistry or some other accomplishment that you can point to that might help us, inspire us to lift our own Excel game? - Sure, I love a good make over. So, I'll point people to a make over that I have on my website that was done in Excel. And I think that I'm particularly proud of this one because it tackles important issues. So, I started with a graph I found on the Susan G. Komen website. It was about breast cancer, incidents and mortality among white women and black women. And on their website, they had just shown this data, incidence and mortality in the same graph as lines and align charts. They're showing this as trends over maybe 20 years or so. And that line chart, you know, it isn't terrible, it just could be so much better to tell the story that they're trying to tell what the kind of urgency that is really deserves. So, I did a couple of things, first of all, I broke apart the incidence and mortality, and I broke apart the data into each race and I reformatted it a couple of different ways. So first I put them into overlapping bars, great graph type that shows subsets one thing being a subset of something else, some mortality being a subset of incidents, only a portion of the people who actually get breast cancer are going to die from it. And when you look at the data that way, one set of overlapping bars for black women to one set for white women, you see that the mortality bars are so much higher for black women, it's such an important story to tell. We can always reshuffle the data too and put incidents in one chart and mortality in another. Because if you think about it, those two things really don't belong on the same scale. When you put them on the same scale, we're kind of like squishing the difference between whites and blacks for incidents and mortality, because the scales have to be so large to contain both of those metrics, when really they shouldn't be in the same scale at all. So, we pull them apart, and then in that case, I turned them into dumbbell dot plots. It's an example of how we can really draw attention to that kind of disparity that's there, where in that case, it's very evident that white women, are more likely to get breast cancer. They're higher rates in white women, but black women are the ones most likely to die from it. So, the disparity switches places in the two charts. And I think it's just such a stronger story, and the data deserves that kind of storytelling. I want to say one more thing, so, I'm really proud of the make overs that I do, but I think what I'm most proud of is what my clients do, the people who come to my workshops and learn how to make great data as an Excel. Seeing them go and run with it and do amazing things, many of my clients work in like state-level departments of health. So, to see them using the things that they've learned from me in their communications about COVID, for example is probably the work that I'm most proud of. - Yeah, and you know the point that you just made is really interesting about the breast cancer, sort of the changes that you made and why. You know, big part of this is about data literacy, right? It's about understanding data, understanding what am I trying to communicate about this data? What is the nature showing trends over time versus a proportional share versus, yeah, incidents versus mortality rates and sort of really taking a step back from, Oh, I got to crank out a chart right now and really thinking about it more, what am I really trying to communicate? And that's the part that a lot of people either forget or maybe they just sort of haven't developed the skills to think about it that way. And that's a really important part of this. So, one question is when you, it's time to thorough sort of think beyond Excel 'cause the Excel can't do every single thing. You can't accomplish some tasks at all in Excel, no matter how much you hack it, possibly. How do you make decision, for you personally and what you recommend to other people when to go beyond Excel? So, for instance, mapping really is not up to par in Excel, and to your point earlier about PC versus Mac, it's not even in the Mac version, yet. At least not in my Mac version. Other than maps, when do you recommend people consider going beyond Excel for visualization, and why, and maybe what other tools would you recommend at that point? - Yeah, so, I'm going to wholeheartedly agree with you on the mapping thing, Excel's mapping, I believe is using Bing Maps as the background, and it's one of the... Actually a lot of mapping software does this, where they use like Google Maps or Bing Maps where, the map base is so full of noise you know? Like every river and every park in town is included in the background there, and it just feels really cluttery when that happens. So, if I'm mapping, I go to Tableau, that's the software that has the least obnoxious-based mapping to work with. And, it's also very easy to map inside Tableau. The other thing I would go to Tableau for would be anything that I want to embed on a website that requires interactivity. And I don't think that's necessarily a lot, interactivity has a time and a place, it's not something that we should put out there for everything that we produce. I think oftentimes well done static graphs are more than sufficient and that's what I use Excel for. But if it's going to be some kind of like interactive dashboard that people want to have living in their website, that's when I would use Tableau. And really those are the two tools I use, I know there are so many out there and I'm not saying anything bad about any of them. Just use what you have, right? Master the tools that you have. - Yeah, yeah so Tableau, you know, is in that category of tools, that's, you know, dashboarding, BI tools. and of course, Excel's sister product Power BI exists also from Microsoft. And of course, Microsoft is making a big play with their licensing structure. It really trying to grow Power BI's market share. Do you think that down the road one day, Power BI and Excel may be merged together, you know? The example I brought up earlier, power query, which is this incredible data transformation tool is available on both those tools. It feels like they're starting to sort of maybe match together. You know, do you think that, that happens? What, do you think is the future of Excel in the marketplace, as some of these other BI tools maybe start to gain in popularity? - Yeah, I think we saw Microsoft start to scramble to make Power BI more useful after Tableau really came onto the scene in a big way. I think they're trying to position Power BI as the Tableau competitor, but I'll tell you what work with Power BI makes my heart hurt. I avoid it as much as humanly possible. (both chuckle) Because it doesn't have that flexibility that Excel has where you can hack it to make things that it doesn't normally do. And so, if I'm going to be limited in the chart types I can work with, it's just not really a viable platform for me. They do have a little like plug-ins store, where third parties make some plug-ins, that you could carry some of these more, interesting high impact graph types. However, then you're relying on a plug-in and plug-ins get updated, plug-ins break. I'll give you an example, I made a dashboard in Power BI that I embedded into a website and it used a plug-in for one of the charts. And it was out there like living in the internet, people looking at it and stuff, I didn't know it, but I went back to go visit it may be like a year later. And the plug-in programming had changed and my chart was like bright pink and the font was all like aggressively, very large. It didn't look at all - Yes. - Like it belonged in the rest of the dashboard and you just can't have that kind of instability out there you know? So, I feel like Power BI really has a long way to go. I'd love to see it works together. Microsoft also like launched and then quickly stopped talking about Charticulator. I don't know if that was ever- - Yes. - Yeah, whatever happened with that? It looked like it was going to be really cool and then they just like dropped discussion of it altogether. So, I'd love to see those things all merged into one, at some point that would make life so much easier for people. I think that will require Microsoft to be a bit more de siloed than it is right now. And I'm just saying this from an outsider who doesn't really know what it's like in there, but in my experience, working with large corporations, the bigger the corporation, the harder it is to turn the boat. And I think that might be the case there too. But getting to the bigger question about the future of Accela, the future of data vis, you know, I think that where we're heading with all of this is that we will be able to take these foundational pieces of knowledge we have about data visualization, best practices, whatever that software might be that we're using, and we're going to be using it more and more to tackle these bigger problems, these more pressing global issues that we all have. And I think we're starting to see it already, I mean, we're seeing graphs show up where president Trump and a journalist are arguing over a bar chart in an interview, I mean 10 years ago, I would've never even imagined that would be a thing, and now here we are, so. I mean, I think that's just pointing to even more using data vis to help us tackle the more pressing issues that are out there. - Yeah, I mean, you know, reversed curve or flatten the curve, you know, this year was that, you know, that it is entered this like data visualization, no question. By the way, so Stephanie brought up Charticulator,. so yeah, Power BI there was this tool from Microsoft research called Charticulator, which was essentially a click and drag whizzy wig, tool to allow you to make custom charts for Power BI, which would have solved one of the problems that she was talking about, which was incredible. And then, yeah, I haven't heard much about it. It's been at least over a year, I think. But the one... In defensive Power BI, one of the cool things about it is that you can do other things inside of it. You can put ArcGIS things into Power BI and bind it to the data, you can do R, you can do D3. So, it's interesting, like, you know, Power BI has a huge flaws and huge limitations in its built-in charts, and the plug-in library is definitely has some interesting quirks. I hadn't heard about it plug-ins being updated in ruining your dashboard. (Bill chuckles) But then it also has other things that Tableau can't do. And so, it will be very interesting, I think, to see where those sort of evolve and certainly, you know, Tableau does amazing custom visualizations, and Power BI does other things perhaps in some cases a little bit better. So, you know, I think for a data analyst type folks what I always recommended is if you can manage it to learn both 'cause I think there are pros and cons of both those tools. And certainly, Tableau is the leader in the industry right now. And Power BI is definitely playing catch up. Any final thoughts, words of wisdom for our audience on data visualization, best practices within Excel. And we're running out of time, we're pretty short. So, if you can keep that brief, 'cause I do want to also give you time to talk a little bit about some of the stuff that you do and make sure you have time to mention some of your programs and make sure people can find you to find more words of wisdom down the road. - Well, I always say that if you can imagine it, we can probably make it in Excel. If you said out there somewhere in some cool ad, we can probably make it in Excel. It's just a matter of learning how to push the buttons and that can look like it's going to be monumental. It can look like it's going to take a year as to figure out how to get there. It's just that the make over is so drastic that it can feel like it's climbing a mountain, but in the actual amount of time it takes to learn, how to do it, we're talking great charts in under three minutes. So, it's doable work and we can make it happen. - That's great. Yeah, and I have to say, you know, and I think I mentioned this in the lesson portion, you know, like anybody I'm using Excel, I'm trying to do something nice and fancy and fun and wonderful. And I Googled it, of course and I ended up on your website all the time. I can't remember, I think it might've been a slope graph. I was like, how do you make a slope graph? I think I may have found that from you or maybe it was a dot plot, but you know, it happens and you do a great job explaining it. So, we're running out of time but I do want to just give you a chance to make sure to share with everybody, what you do, I know that you have various programs that you run to help people sort of get up to speed on their Excel work and visualization work generally. So maybe you can just tell us a little bit about some of those programs you offer. - Yeah, we have the data visualization Academy where we have about 70 tutorials, new ones come out every month on how to do graph, making in Excel and Tableau and in R. So, we're really proud of that enrollment for that opens up twice a year. So, you can just keep an eye out for it. We also have a more advanced program called Graph Guides, where you go through 50 of those tutorials in one year with a coach who really holds your hand and helps you along the way, and for some people like that's what you need in order to make yourself like actually sit down and learn something. So, we're really proud of those. My books are also great resources, where you're going to be obviously learning on your own but it has those step-by-step instructions. And the blog is a great place to go. If you have zero budget and you just need a quick point in the right direction, as you mentioned, Bill, we've got some of the more basic chart types, that and how to make them, listed there for anybody to access anytime. - Great good stuff. And I can vouch for it as I said, I've ended up there many times myself and gotten real help from the blog. So, thank you, Stephanie, so much for being here today. I really appreciate it. I know that our audience, in particular, you know, we've done 20 or so of these episodes so far and they're all great and amazing and wonderful. And I think this is one of the ones that's going to probably have in some ways to have the biggest impact, because as we keep saying, this is the tool everyone uses. And, I think everyone will really appreciate that what you brought forward today. So, thank you again. - Thank you so much for having me, Bill.

Contents