Mastering Data Visualizations Using Tableau w/ Kate Strachnyi @StoryByData (Episode 21) #DataTalk

Listen to the podcast:

Every week, we talk about important data and analytics topics with data science leaders from around the world on Facebook Live.  You can subscribe to the DataTalk podcast on iTunes, Google PlayStitcherSoundCloud and Spotify.

This data science video and podcast series is part of Experian’s effort to help people understand how data-powered decisions can help organizations develop innovative solutions and drive more business.

In this week’s #DataTalk, we spoke with Kate Strachnyi about ways to improve your data visualizations using Tableau. Follow her on YouTube, LinkedIn, Twitter, and her blog.

Here’s a full transcript:

Michael: Hello. Welcome to our weekly #DataTalk, where we talk to data science leaders from around the world. Today’s topic is Mastering Data Visualizations Using Tableau, and we’re very excited to chat with Kate Strachnyi. She is the author of Journey to Data Scientist. She works in data visualizations and reporting. I’m super excited to have Kate as our guest today. How are you doing, Kate?

Kate: Hi, Michael. I’m doing excellent. Thank you for having me here.

Michael: Kate, you’re in New York, and I’m here in California. It’s exciting to chat with you from the coasts. Tell us a little bit about your journey to working in data science and the path you took.

Kate: Sure. I started out with working in risk management. Not much to do with data science, but in late 2014, I had my first child. And I don’t like the travel. I want to see my children, and they were able to find me an internal role. It’s not client-facing, but they gave me this large data set in Salesforce, and they said, “We need insights.” There was no direction. They just said, “Here’s Tableau.”

So, what’s Tableau? I started researching, and I just fell in love with data. Just being able to take data that people don’t even see because they enter it quickly into a CRM tool and then they don’t use it. But then years later being able to extract a wealth of information and seeing the trends or the challenges that are faced by people we were working with was really powerful. And that’s when I fell in love.

Michael: That is so cool. I found you through LinkedIn. You started a whole video series interviewing different data scientists. Can you share a little bit about that series?

Kate: It’s called Humans of Data Science, and I started it about six or eight weeks ago. The way it happened was I was sitting with my laptop watching TV after the kids went to bed, which is super rare because I’m usually passed out right after they’ve [crosstalk 00:02:22]. But I said to my husband, “Why don’t I start doing this video series where I just talk to data scientists?” And he said, “Yeah, go. Keep going.”

I posted something on LinkedIn where I asked for volunteers. I said, “Who would like to be interviewed for this Humans of Data Science, where it’s basically three to five minutes where we’ll be talking about data science and helping aspiring data scientists learn from your experience?” And the response for me was very positive. We have hundreds of people. I think if I post again, there’ll be thousands of people who are interested in participating. At this stage, it’s gotten a little bit difficult to manage all of those volunteers and participants.

But it’s been truly an interesting and amazing experience, because when you talk to someone on LinkedIn, even if you’re messaging them, you don’t really get to know them as we’re getting to know each other right now. We’re almost together. Thousands of miles apart, but I can talk to you and feel like you’re here. Just getting to know them. We do have a conversation online, and then we stay and chat a bit more after the recording goes off. So, we get to get even more personal.

Michael: That’s so cool. For those listening to the podcast or watching the video, where can they go to see the series?

Kate: It’s on Story by Data on YouTube. And I also have a Story by Data blog where they can also see … If they go to Humans of Data Science, there’s a link of all the people who have been interviewed. And there’s also a long list of people who are waiting to be interviewed.

Michael: Wonderful. And I’ll make sure that we get links to your YouTube channel and your LinkedIn profile on our Experian blog. And for those listening to the podcast, the short URL, which we can put up on the screen in just a moment, is Again, that’s, and that will be the blog post where we will have a full transcription of today’s episode along with links, so you guys can check out Kate’s awesome video series. Kate, before we get into the interview, can you also talk about your book, Journey to Data Scientists?

Kate: I started that project shortly after I fell in love with data, and I started Coursera data science online courses and just learning everything I can. I decided that the best way to learn about data science or if it’s something I even wanted to do was by talking to people who were already doing it. And I decided to go through different levels.

People who had just gotten into the role and people who have been into that role and people who are famous for it, being famous data scientists. And just asking them simple questions: What are the projects you’re working on? What’s your day-to-day look like? What did you have to do to get there? Do you even like it? What don’t you like?
[crosstalk 00:05:17] and then after a couple of those conversations, I thought other people might be interested in this concept as well. So, I decided to compile all those over 20 interviews and make it into a book.

Michael: And people can go to to get it?

Kate: You can get it on Amazon. I do have a link on Story by Data that takes you to Amazon.

Michael: Cool. I love that you did this — turning your own research into a book to help others. That’s awesome. And I also love that you have a very entrepreneurial spirit about you. Developing shows, doing a book, and all this is part of your own research to help you. But also, you’re helping tons of other people in the process. I’m curious about your first book and I … We’ll get into your second book in just a minute. But your first book, were there any surprises as you’re talking to different data scientists about challenges or personality types or anything like that?

Kate: One thing that really stood out for me was every single person I’ve spoken to, when I ask them what does a person need to become a successful data scientist, every single person said the person needs to be innately curious. Keep asking questions of the data. This came across, like I said, through every interview, which I thought was very interesting. I did expect that to some extent, but I didn’t expect every single person to say that. That was interesting.

Michael: That’s cool. Before we started today’s show, you mentioned you’re working on a second book. I know it’s not out yet, but can you share the premise?

Kate: Sure. This is a collective biography plus Q and A of some of the more well-known data scientists. And some data scientists you may not have heard of, but you should. They’re making interesting impacts, and I think people are going to like this one. It’s gonna be called something along the lines of The Disrupters, which highlights leaders of data science. I recently decided that it’s going to be published as a series. Book one should come out, hopefully, at the end of summer, if everyone’s schedules align.

Michael: That’s awesome. Where does all this creativity and entrepreneurship come from? Because you seem to be doing so many different things. You’re already working as a data scientist, but then you are producing books, videos. Where does this creative spirit come from?

Kate: I just like it. I think it’s fun. I usually tend to go with what I enjoy doing. And it changes over time. In the beginning of my career was risk management. So I had a risk blog. I had a risk book. It’s just something I do. I think it’s a way of learning for myself. When you don’t understand a concept, if you feel like you can explain it, then you feel like you can understand it. That’s how I got started, but then it’s also addicting just talking to people. I just like people. That might be it.

Michael: It’s so cool because you’re helping so many people in the data scientist community. Thank you for that. Tell us about the work you are doing now in data visualization. You mentioned how you stumbled into this role.

Kate: Yeah.

Michael: ’Cause you’re working at risk management, right? And then you’re like, “I work from home. I want to be with my kids.” And then they’re like, “OK.” And then, all of a sudden, you’re put into this data visualization role. And it shows you’re a self-starter, self-learner. Tell us about your process when you’re all of a sudden put into this data visualization role. What was your thought process on how to learn this and apply it?

Kate: When I first heard of, let’s say, Tableau, and then I got the data, my initial reaction was to start Googling tutorials, watching YouTube videos. That’s actually why I started my Story by Data channel. As I learned Tableau, I thought most people would benefit if they’re just starting out. If there’s just a simple video that they can watch in five minutes and go from never seeing the tool to actually building a dashboard. That’s why I started creating tutorials. It goes from very entry-level to some of the more complex and custom features of the tool.

But back to your question. The way I started was Googling, and Tableau software also provides some trainings that you can do online. And, aside from that, I also started looking at Tableau Public, where you have other people posting their visualizations that are free to download and see how they were built. It’s kind of like taking a car, pulling it apart and learning how it’s built. A lot of that is what I started with, but then also the feedback from leadership. Although they didn’t know much about data or the tool, they knew what they wanted. And what they wanted was usually something I didn’t know how to build, which helped me tremendously because then it would be, “OK, this needs to be done by tomorrow. I have to learn and scramble to figure out a way to make Tableau do something it’s not meant to do.” And I think that was my biggest learning process.

Michael: That’s fascinating. You’re working with data, trying to find insights, producing a chart, graphs, visuals to show the insight to leadership, and then sometimes they return back to you and say, “We want something else. We want a different visual.” Yeah?
Kate: They always want something else. Yes. You’re always trying to please several members of a committee or an audience, and some want high-level and some want the details. I think that most people can agree that that’s usually the case when you’re working with multiple audience members. And that’s why I love Tableau, because it allows you to show a high-level image and then allow people to click on the bar chart and get a table or another chart or a map view and click here and click there, drill down and hover and get more information. That’s helped meet the needs of the audience.

Michael: That’s cool. This is something that we don’t talk about a lot, but dealing with leadership … because, like you said, sometimes they’ll ask for something and you’re trying to please multiple people. Because sometimes going to your boss, sometimes going to a committee or your boss’s boss, you get all this feedback about changes. How do you deal with that? Because it can be complicated at times, right?

Kate: Yeah, it can. It’s knowing who the most important audience member is and then trying to appease everyone. But you never will make everybody happy. So, providing options. Sometimes what I do is I create different versions of the same data and just have two tabs of the workbook that some people can go here … As an example, one of the dashboards we use is to track expenses of different investments. And each investment has a target expense that they should be meeting or should not be exceeding, in this case. And some people want it as a table, so I have one very simple table where they have the numbers by period, how much we spent, how much we have. Then, most of the people wanted a chart where they can see a bar where it meets a threshold and it passes it, it changes color. Sometimes you create two different things and sometimes you don’t make everybody happy. That’s all.

Michael: Very cool. I like how your workflow is and developing different types of data visualizations for different audiences to help them. That’s very cool. I’m curious about when you were first starting out doing data visualizations, were there any mistakes that you made early on that you now are much more aware of and try not to make again?

Kate: Yes. Absolutely. I think we all make mistakes.

Michael: Yeah.

Kate: The most recent mistake that I can talk to is I built this dashboard. It was very well-received by leadership. Even the advisory leader said, “Wow. We need to use this everywhere. It’s great.” But the issue was the way I built it. It was slow. So, imagine sitting in a committee where the leader’s trying to show all the key metrics for investments and it’s just a white screen and it’s loading. And it’s loading. And it’s like [crosstalk 00:14:20].

Michael: Yeah.

Kate: [inaudible 00:14:24] only have a 30-minute meeting, and it’s not only embarrassing. It just doesn’t look good. It’s a waste of time. And what happened was that visualization wasn’t used as much because it just doesn’t load. People would contact my team to get the data, which ended up with us spending a lot of time. And the solution I came up with to that was I contacted Tableau Software. They gave me some tips on how to increase the performance and make it faster and optimize everything. And now it loads in like 10 seconds. So, everybody —

Michael: Oh, wow.

Kate: Yeah. It was just quick fixes that I had to do to allow it to go faster.

Michael: That’s cool. Yeah, I can imagine, you develop this great visualization and you’re in a presentation and all of a sudden the person’s sharing it and it’s loading, loading, loading.

Kate: And I had [inaudible 00:15:23] on my screen in that meeting. It was, I’m just … well [inaudible 00:15:27].

Michael: Oh, man.

Kate: Yeah.

Michael: That is rough. Especially when you work so hard on something and then the technology or the processing is not happening fast enough.

Kate: It’s like writing a book and you can’t open the covers.

Michael: You shared your experience learning data visualization on the fly. Googling, reading books, taking courses. Are there any data visualizations that have inspired you? Things that you’re looking at regularly to help your creative mind work on data.

Kate: Yeah. I do Google things if I need a dashboard. Google Images helps a lot because you can just scroll through thousands of pictures and get ideas, but I think what regularly motivates me is #MakeoverMonday. It’s this project that, I forget their names, these two Tableau awesome people started years ago where every Sunday night they post a visualization plus a data set that everybody can download. And they ask the public to recreate this visual in a better way. Seeing everyone’s visualizations is so inspiring, and I get a lot of good ideas with that.

Michael: That’s cool. Where is this available for people to check out?

Kate: I think it’s, but I can send you the exact link if you want to include that. I think that’s a good way to learn from others.

Michael: That is so cool. Seeing how different people are working with the data. What’s cool about you and what I find with a lot of people who work with data and are doing a good job with data science is that you have the two sides of being very creative, artistic, but then also very numbers-oriented. That’s cool that you have both that creative side to be able to show great data visualizations, but then also that very analytical side because you have to find the insight, right?

Kate: I didn’t start out that way. I started out more on the creative and visual side. Then one of my internships in college, I won’t say the name, but I was working in an underwriting role where we had to look at a lot of data. This was years ago, and one of the managers there said, “Why don’t you just do sales? You seem to really like people and you like to talk.” And I guess I wasn’t doing an amazing job with the data. I [crosstalk 00:18:11] mistakes. And he said, “You’re not detail-oriented.” And that to me was like, “Oh, really. Watch me. Watch me become detail-oriented.” That’s how it all started. I think that’s when I really started to pick up more of the details. It spoke to me. That negativity inspired me, I guess.

Michael: Yeah. No kidding. You were like, “I’m going to show you.”

Kate: I’m going to work with data.

Michael: Oh, man, that’d be so upsetting. You’re working hard and all of a sudden, “Kate, I think you’d be better in sales.” And you’re like, “Really?”

Kate: Yeah. They had two internships and they wanted me to continue, but they said, “Why don’t you do your next internship in the sales phase?” Well, they kept me in underwriting, but they just recommended. They asked if I’d like to switch.

Michael: Can you talk a little bit about, for those who don’t know Tableau, the tool and what sorts of skills are needed to use it properly?

Kate: Sure. For those who don’t know Tableau, it’s a data visualization tool. You can get access to that tool for free on Tableau Public. You don’t have to buy a license. You can start downloading data and creating your own visualizations today. And what you would need to know is some basics about data structure.

Tableau likes to have data structured in a specific way, so it wants rows and columns. It doesn’t like when you have an image in your Excel file or if you have a free column or free row before your data starts. It just has to be nice and neat. And then when you upload a file into the tool, it starts to read your column headers and determines if they are dimensions or measures. That’s how it treats the data. If it’s number of sales, then it would be considered as a measure and Tableau would understand it and use it differently in a chart versus something like a region name or a customer name, which it would think is a dimension.

Michael: OK. So, when you’re uploading a huge file and it can be a mixture of dimensions and measures, how does it handle the noisy data or the messy data that it doesn’t know what to do with?

Kate: It has specific defaults that it goes to. There’s a very long list. Tableau online help has that kind of go-to of … if your dates are not structured the right way, you can restructure them in Tableau. But my preference is to clean the data before it gets into the tool. I do use some of Tableau’s features, like the data reshaping tool or the pivot table tool to reshape some of the data. But, like I said, my personal preference is to get it clean before it goes in there.

Michael: And then once the data’s in there, can you talk about your process on how you begin to work with the data to find the insights?

Kate: Yes. Once it’s in there, the key is to always start with what question you are trying to answer and who the audience you’re trying to tell it to is. Starting with a question like, “Which category had the most sales?” Then you would know that you would need the dimension category and you would need the measure of sales in the view, so you can simply double-click it or drag it into where you want to see it.

And Tableau has this cool feature, it’s called “Show Me,” where it’s a little drop-down that has 16 or 18 different chart types that, based on the measures and dimensions you brought into the view, would tell you what it recommends you to use, a bar chart or a line chart, given the data types. If you are not happy with their recommendation, you can just click different ones that are available and it will automatically change your chart without any coding or any complications.

Michael: That’s cool. So, doing good data visualizations, it’s a creative process, analytical process, but also, if you don’t use the right data visualization, it can be very misleading. How do you avoid creating misleading data visualizations?

Kate: Yeah. A few examples of where you can be misleading is don’t use pie charts. Pie charts are notorious for misleading folks because humans don’t really have the capability to see whether one slice is bigger than the other, especially if there’s five different slices of data. The idea is either don’t use pie charts or just try not to use more than two slices. The axes can be misleading. If you picture a bar chart, we can talk to an example. Let’s say there are two categories that have sales. One has $50,000 and the other one has $100,000. If your axes start at zero, you can clearly see that one is $50,000 and the other … But if we start the axes at $50,000, it will look like category A is not only [inaudible 00:23:34] and you should stop investing money into that category. There are a lot of ways to be misleading. I think there’s even a book of how to lie with statistics that —

Michael: Yes, there is.

Kate: Yes.

Michael: Bill Gates loves that book.

Kate: Yes. I think a lot of people love that book. I have it. I haven’t read it yet. It’s in my pile of books that I’d like to read. One day. Someday. Soon. I mean, there’s just so many ways. There’s also use of color. If you use bright colors to accentuate something that might not be as important and you gray out something that is important. Humans just have immediate reactions to visualizations that they make assumptions in three to five seconds of what they think they’re seeing. It’s just very easy to mislead people, unfortunately.

Michael: Are there any favorite visualizations, your go-tos, that you like to look at the data in that specific visual — even though you may not use it, but just kind of a go-to?

Kate: Not really a go-to, but what’s happened in my role in the past few months is we care a lot about KPIs, Key Performance Indicators, and metrics and how we’re performing against [inaudible 00:24:45]. I tended to use a lot of line charts and bar charts, but they’ve been … They’re not my favorite because they’re not as fancy or cool, but they get the job done because people can clearly see if you went up or down or how you did over time. Those are the most used, I guess.

Michael: Are there any … What’s one of the most complicated visualizations you’ve had to make to tell your story?

Kate: I had to build a pie, not a pie chart. It’s a scatter plot that plots individuals of where their assessments fall out. But it wasn’t that easy because I had to use a background image. I had to use a negative axis and … But it was the most rewarding because it took me days to put together. And it’s used now for all the incoming interns and consultants hired there. They use that in the training, so it feels good to see it being used.

Michael: That must feel so rewarding.

Kate: It is.

Michael: The visualizations you’re creating are being used in meetings. They’re being used, maybe, in sales presentations. White papers. Are there any other places where your visualizations are being shown?

Kate: Yes. One thing you wouldn’t expect is we use it in client facilitation meetings. An example is we have new chief information officers who go through a program at work. Let’s say at a bank somewhere a new chief information officer is hired. And we bring them in for a day session that helps them plan what they should do in the first six to 12 months, who they should network with, what they should do with their talent. It’s an all-day session. And we have these huge panels that we call Smart Boards that are based on Tableau. Basically, it’s a tool that —

Michael: Wow.

Kate: It’s this huge blown-up TV screen. And a visual that they can interactively touch and play with and use drop-downs to see how other CIOs are doing or what they’re planning to do in the next six to 12 months. That was cool, seeing it used in a client setting like that.

Michael: That is so cool, with an interactive whiteboard.

Kate: Yeah. Exactly.

Michael: [crosstalk 00:27:21] the data on. So, for someone who’s playing with data, they obviously need to have a good understanding of Tableau.

Kate: Not necessarily. There are different tools available. There’s Power BI. There’s Domo. There’s lots of other data visualization and analytics tools. There’s R and Python that they can use to analyze and build visualizations. That’s just the one I fell in love with, and I think it’s growing in terms of popularity. And it’s really easy to use. I think that’s why I liked it to start with.

Michael: That’s cool. What would be a tip that you’d give to other people who are just getting started in data visualizations?

Kate: I think a tip would be to first download Tableau Public. Then, I do have some intro videos that they can watch on YouTube that in five to 10 minutes they can get up to speed on how it’s used. I even have sample data sets that they can use along with the video, that they can follow along and get hands-on. I wouldn’t recommend reading too many books. Although it will help you get an understanding of how Tableau works and maybe some of the tips. I think more hands-on and videos is what helped me. Get a data set that you feel passionate about or at least you can relate to. Something local, maybe the number of books that your library has checked out on a specific topic, and then try to visualize that data and see if you can help someone, help a local organization visualize their data. It’s kind of a win-win. You get to learn about data and visualization and they get some insights from their data, which they probably wouldn’t have before.

Michael: Kate, when you get a project, do your leaders ever tell you what type of chart they want from you? Or is it more, “Analyze the data, you tell us the story, and you figure out what chart.”?

Kate: I guess a mix of both.

Michael: OK.

Kate: In some cases [crosstalk 00:29:25] analyzing survey data, I would be told, “Last year we used these types of bar charts in the survey. I want them to look exactly the same.” In that case, I’m being told. My favorite projects are the ones that say, “I have this thing in mind. I’m not sure what it’s going to look like, but I want it to have these bits of information, and just go with it and make it your own.” And that’s what I love. That’s my sweet spot. I have a project now that I’m working on like that. So, I’m excited.

Michael: That’s cool, because it gets your creative juices going and the artistic side of you.

Kate: Yeah.

Michael: Where you get to play with data.

Kate: Yeah, you get that freedom of creativity. Exactly.

Michael: Tell me about your site, because I love Story by Data. I love the concept. Tell us about why it’s so important that you are involved in telling the story.

Kate: Because without the story, it’s just data. People can’t really just look at a spreadsheet of data and understand the insights. The reason behind starting the blog is to get my thoughts out there and try to help other people. It started out with just understanding the role, but now I want to help those who are aspiring to get into data science or data analytics or even just data visualization. Because on LinkedIn I get hundreds of messages a day about how to get started. And that’s why I keep creating these videos — so I can help more people. And that was the goal behind the blog and the channel.

Michael: I see you all over LinkedIn and you’re helping tons of people. [inaudible 00:31:12] all the incoming messages?

Kate: I try to respond as much as I can. If people just say hello or just say nice … I try to respond, but I usually tend to respond more to those who have an actual question. Especially one that I can answer. If not, I refer them to other people. I get a lot of messages about just how to get started and I try to sometimes … If it’s a question I think will benefit a lot of people, I will even create a post around it and tag that person and tag other people I think can help. It’s hard. Just making time for answering everyone’s emails or messages.

Michael: Yeah. That’s like a job in itself. Wow.

Kate: Yes. It is. But it’s fun.

Michael: For those who want to learn more about you and your books and your video series and other things that you’re working on, where should they go?

Kate: I would think LinkedIn is the best option. If you just Google or go on LinkedIn and look for Kate Strachnyi. There are not many Strachnyis out there, so you’ll probably find me.

Michael: If you’re on LinkedIn, look up Kate, K-a-t-e. And her last name is spelled S-t-r-a-c-h-n-y-i. Check her out. Follow her. Connect with her there. And if you can’t remember how to spell her name, you can always go to the Experian blog and I’ll have links going to her LinkedIn profile, her website and video series, and also the resources that she’s mentioned in today’s episode. And the URL, which we have on the screen here or for those listening to the podcast, is just That will bring you over to the blog post where we’ll have the video, the podcast episode and a full transcription.

Kate, thank you so much for your time. It was a blast chatting with you. Awesome hearing your story, and thank you for all the amazing work you’re doing. I love your entrepreneurial spirit. Constantly creating new things, trying to help people, writing books, doing videos, interviewing different people. It’s so cool to see what you’re doing because you have a full-time job, you have kids, and the fact that you’re doing all these different things is so awesome. Thank you.

Kate: Thank you so much. This was a lot of fun. Thank you for including me here.

Michael: Thank you so much, Kate. And hope that we can chat soon.

Kate: All right. Bye.

Michael: OK. Bye.

About Kate Strachnyi

Kate is the author of Journey to Data Scientist; which is essentially compilation of interviews that Kate herself conducted with over 20 amazing data scientists. — with backgrounds ranging from LinkedIn and Pinterest to Bloomberg and IBM. She is also the creator of Humans of Data Science (HoDS) – a project that works on showing the human side of data science (housed on her Story by Data YouTube channel).

Kate is a manager working in the data visualization & reporting space. She previously served as an insights strategy manager and research analyst, where she was responsible for enabling the exchange of information in an efficient and timely manner. Prior to working with data she focused on risk management, governance, and regulatory response solutions for financial services organizations.

Before joining the consulting world, she worked for the chief risk officer of a full-service commercial bank, where she was in charge of developing an ERM program, annual submission of ICAAP, and gap analysis of Basel II/III directives. Additionally, she worked as a business development associate at the Global Association of Risk Professionals (GARP).

Kate received a bachelor of business administration in finance and investments from Baruch College, Zicklin School of Business. Certifications include Project Management Professional (PMP) and Tableau Desktop 10 Qualified Associate. Follow her on YouTube, LinkedIn, Twitter, and her blog.

Check out our upcoming data science live video chats.

Never miss a blog post!

Subscribe to keep up with all things Experian.