SDS 091: Lessons from a successful career in Data Visualization

(background music plays)

Welcome to the SuperDataScience podcast. My name is Kirill Eremenko, data science coach and lifestyle entrepreneur. And each week we bring you inspiring people and ideas to help you build your successful career in data science. Thanks for being here today and now let’s make the complex simple.

(background music plays)

Hey guys, welcome back to the SuperDataScience podcast. Super excited about today’s episode. We’ve got the legendary Andy Kriebel on the show. You may have heard of Andy from his blog, vizwiz.com, where he talks about Tableau visualisations and dashboards. He’s also got a blog called datavizdoneright, so you may have heard of him from there. He’s got a Twitter where he shares amazing and stunning visualisations and other analytics tips and hacks. And finally, Andy is of course the Head Coach at the Data School, which is part of The Information Lab.

So in today’s session, we’re going to find out all about Andy’s journey, why and how he got to where he is, how he gives back to the community, what he’s passionate about, and lots and lots of tips on how you can structure your career and how you can progress through this field of data science. Personally, I really enjoyed this episode. Andy’s got some amazing stories to tell us, some amazing stories to share, and at some point, I even caught myself just imagining what he was explaining about his journey through Coca Cola, through Facebook, and now through The Information Lab. It was as if I was watching a movie.

So sit back and relax, and prepare for a very inspiring adventure. And without further ado, I bring to you Andy Kriebel.

(background music plays)

Welcome everybody to the SuperDataScience podcast. I’ve got a very exciting and inspiring guest today on the show, Andy Kriebel. Andy, how are you today?

Andy: I am absolutely fantastic. How are you?

Kirill: Beautiful. Beautiful as well. I was going to make a joke about it being rainy in London, but as you said, it’s sunny today, surprisingly.

Andy: Very rare.

Kirill: Yeah, one of those five days this year that it’s going to be sunny.

Andy: 3 days in a row, too! That’s even more rare!

Kirill: Oh wow, that’s crazy, man. It’s like one of those movies where tomorrow you’re going to wake up, and you’ll realise it was all a dream, the past 3 days, you know, it’s so improbable?

Andy: I think so, yeah.

Kirill: Ok, so you’re obviously not from London though. Where are you originally from?

Andy: I was actually born in Philadelphia, Pennsylvania. I was there through high school, and then I went to college in Florida, at Eckerd College, it’s a small school on the Gulf Coast near Tempe, Florida. It’s actually in St Petersburg. And I studied mathematics there.

Kirill: Interesting.

Andy: Yeah. And then I kind of moved around a bit. Lived in Atlanta for a long time, and then California, before moving to London.

Kirill: Ok. And you have four children, right?

Andy: I do. Yes.

Kirill: That’s so many!

Andy: Keep us busy!

Kirill: That’s so cool. So how are they taking the move?

Andy: Pretty good, I guess. The schools are quite different, and they’re the only Americans at their school, so they tend to get picked on a lot, mostly because of Trump getting elected, but you know, hopefully that situation resolves itself soon enough.

Kirill: Ok. Gotcha. Alright, so Andy, you are a really inspirational guy. You have the website vizwiz.com, and you work at the Data School, so you’re Head Coach at The Information Lab and Data School. Tell us a bit more about all these things that you’re doing. So maybe I’ve missed something there.

Andy: Well, we moved to London about 2.5 years ago. I was working for Facebook in California before this, running their Tableau practice and data visualisation there. And I would travel the world a lot and do training and things like that. And my wife came to a conference with me in London, and it was her first time. And on the way home, she said, “Let’s move to London!” Which caught me off guard a bit, because she didn’t really want to move to California, because her family was all from Atlanta, and moving away from family and stuff. But maybe that was a good intermediate step.

So I actually just started reaching out to my network and spoke with Craig Bloodworth, who I’ve known for a long time. I’ve known Craig and Tom Brown, who’s the Founder of The Information Lab, since about 2010. And Craig and I started chatting, and were like, “Hey, why don’t you just come work for us?” and were like, “Ok, well that was an easy interview.”

Kirill: Yeah! Fantastic.

Andy: So from that point, it was a matter of talking about what I would do for them. At the time, Tom was hearing from a lot of customers that they wanted long-term consultants, and that’s not really the way that our core consulting team is set up. They’re set up for short-term engagements, but we’re seeing a greater and greater need for analytics talent and people for long-term consulting engagements. So, that’s kind of where the idea for the Data School came about. Basically, the way that the Data School works is we hire eight people three times a year, so every four months, and they go through a two-year training program. And the first four months they are training with me at the Data School. My job is basically to make them as good of a data analyst as I can in four months. And people may scoff at that and say, “How can you teach somebody in four months?” and I say, “Talk to any of them and I dare you to say they’re not a great data analyst when they leave.”

Kirill: Yeah. And we actually had one of them on the podcast. Rachel Phang was on the podcast just a couple of weeks ago.

Andy: Yeah, she was in DS5, so DS6 — DS is our short term because it’s too hard to say Data School all the time. But she was in the fifth cohort and the sixth is in training now. Yeah, Rachel is great. And I think one of the really fun things for me is actually the hiring process. It’s actually very unique. I don’t know anybody else in the world that does it the way that we do. And it’s really, really fun.

Kirill: Tell us more.

Andy: Basically we don’t ask for CVs or resumes. I think those are actually probably the worst way to identify talent. And that’s the way recruiters do it. The problem that I see with recruiters is recruiters generally don’t know what it takes to do the job. They are good at LinkedIn, they’re good at talking to people, but they really can’t identify talent, so why bother with recruiters? And on top of that, we’re going to teach them how to be a data analyst, so how do we then take that part of the equation out and identify just people that are passionate about data analytics? That’s essentially what we’re hiring for.

Kirill: Passion.

Andy: We’re hiring for passion, right. That’s something you can’t teach. You can’t teach somebody to be passionate about data analysis. They either have it or they don’t.

Kirill: And it’s very hard to see from a resume.

Andy: It’s impossible to see from a resume. People lie on them anyway.

Kirill: Yeah. (Laughs) Yeah, gotcha.

Andy: So, basically, what they have to do is they have to create something in Tableau, they publish it up to Tableau Public, send us a link to it, and that’s their application.

Kirill: That’s so cool.

Andy: And what’s really interesting about it is this is a great filter. Think about all the resumes and CVs we would get for this kind of role if that’s what we asked for. So, it’s an incredible immediate filter because anybody that takes the time—most people that apply have never used Tableau before, I would say 80%-90% have never even seen Tableau before. So, if they have to go and learn a tool to create an application and they’re still interested afterwards, you’ve already identified people that have that passion you’re looking for. Yeah, it’s really, really fun. And it’s also great just to see them develop.

Another neat thing that it does also is that a recruiter that’s looking for data analysts is going to look for very specific educational backgrounds and past experience in data analysis and things like that, and we don’t look for that. Just because you worked in data science or data analysis or have a math degree, it doesn’t mean you’re going to be a good data analyst. And on the flipside, let’s say you’re a journalist or you have a degree in English or something like that. That doesn’t mean you can’t be a data analyst. You know, we’re looking for people that have passion and we’re going to teach them how to be an analytical thinker.

Kirill: That’s really cool. So, you said eight people per cohort is how many you hire?

Andy: That’s correct. Yeah.

Kirill: So, right now you have DS6 going. So DS5 has finished, you know, 5 times 8 is 40—

Andy: They’ve finished their training, right.

Kirill: So out of 40 people that have gone through, what is your success rate? What would you say you’ve got that you identified the talent correctly and these people are indeed passionate and they’ve learned and what you’ve taught them is going to help them in their careers?

Andy: Yeah, that’s a really tough question. So, the curriculum has actually changed quite a bit since we started. I would say that everybody that’s made it through the training program is successful. They’re all successful in their own ways and they come out with different strengths and weaknesses. We work a lot on the things that they’re not good at, you know, so trying to make sure they’re very well-rounded. But they come out with different sets of skills. For example, Rachel, who was on your podcast before, is great at storytelling and communicating with an audience. Other people are great at maybe doing some predictive analytics, or are particularly great at design, or things like that.

So, what’s really fun over the course of the four months is to kind of see how people developed in these different areas and sort of find their own niche as well. So they all get kind of the broad experience of both the technical and the non-technical skills, the soft skills they need to learn, but they all kind of find something they’re particularly passionate about.

Kirill: That’s really cool. Yeah, and then they can build their own careers in that direction.

Andy: Exactly, yeah. And one of the things we stress a lot is kind of building their own brand while they’re in the Data School as well. I’m a big proponent of people building their own brand because nobody else is going to build it for them. So, one of the things that they’re required to do is a lot of blogging, creating lots of visualizations on Tableau Public. We record most of their presentations. As long as it’s not a presentation for a client, we’ll record it and put it up on YouTube.

You know, they’re kind of building this portfolio along the way that kind of becomes their CV. It’s better to actually have practical things that you’ve done and you could point people to “Look, here’s my 50 visualizations up on Tableau Public. And here’s the 70 or 20 blog posts I’ve written,” whatever it may be. “And here is a series of recordings of presentations that I’ve given.” You know, they are required to present every week, so the way their curriculum is structured is every week is a different project. About half of those we do for clients for free, so they get practical experience trying to solve real problems for customers, and then the other half are weeks that I make up the projects.

Some weeks are harder than others, and it’s all about basically trying to develop the really well-rounded, kind of holistic experience. But at the end of each week, every Friday at 3:00, they have to present. Sometimes they’re presenting directly to a client in the room, and other times we’re just presenting to each other. We tend to have a lot of people come in on Fridays to visit us, so there’s always people sitting in the back of the room. You never know who the audience is, so that’s a unique experience as well. They become really good presenters. When they have to do it 20-30 times over the course of four months, you can’t help but become a good presenter.

Kirill: Yeah. And Rachel mentioned that to us. She said that’s a great way that you guys have developed for people learning these skills to get into this mode of presenting, that they feel in a safe environment because they’re still at the school but the clients come to them. You know, the client comes, delivers the challenge that they need to solve and then they have to present it at the end of the week or in a few weeks or something like that. That’s really powerful because it gets you, as you say, presenting already in that experience, but at the same time, it’s like the first step towards that, it’s not as scary as going to the client’s office and being in an unknown environment. You’re kind of presenting in the same place.

Andy: Yeah.

Kirill: And also, it’s really cool—is it correct that the team is time-pressured? Like, when they get the challenge, how much time do they have to solve it and then come up with the presentation for the client?

Andy: So, for a client project, the client comes in on a Monday morning and kind of gives a brief about the project, introduces them to the data, and that’s really where the team’s learning how to be a good business analyst, you know, how to ask the right questions. The whole purpose of that project kick-off meeting is to really understand exactly what the client is looking for. And that’s really hard for them to do at the beginning. And part of it is I let them do it themselves because I want them to learn from their mistakes as well. I could do the business analysis and kind of ask the questions for them, but then they don’t learn.

So a lot of their successes are actually built upon their failures. Or they realize, “Okay, these are the things we didn’t ask,” those sorts of things. And then for the client projects, they rotate who the project lead is. Let’s say there’s eight people in the Data School and there’s eight client projects. Each person will be the project lead for one of those client projects. So they get a bit of project management experience, you know, we follow the Scrum development methodology so we do a daily stand-up, we actually introduced something new where a project lead has to tell a really corny joke to start the day, which is really interesting. Some of them are horrendously bad, but that’s okay.

Kirill: It breaks the ice.

Andy: Yeah, and they’re the primary interface with the client, I’m not, so basically they run the project for the week. Now, during that week, they also have training. So, they only probably get about 15-20 hours to actually work on the project. So they’re very, very time-crunched for these really complicated questions that the client brings to them. And then usually in the middle of the week, they’ll do a check-in with the client and show them what they’ve progressed on and make sure that they’re actually approaching the questions the right way and that sort of thing. And then Friday the client comes in, we’ve got a beer fridge, you know, try to make it a comfortable environment for both the client and for us, and they present back and then we turn it all over to the customer. It’s their work for free.

Kirill: Okay. That’s so cool. That’s a really cool setup. And then after several months like that, four months, then they go and work on the clients, you attach them to clients for some time periods.

Andy: Yeah, they do three 6-month consulting engagements for us. The idea there is to give them a pretty rounded experience. They hopefully work in three very different industries and three very different types of clients. You know, I try to stress to them all, “Think of those as three 6-month interviews.” It’s their chance to go in and really impress the client and hopefully get a job offer at the end. Ideally, they get three job offers from their clients at the end of the two years. We will make offers to one to two people per class potentially, it would be pretty rare to do more than two. So, they need to look at it as “These are chances for me to really impress.” And I kind of stress to them, “You know, if you finish your 6-month placement and the client is perfectly okay with you leaving, or they don’t ask us to keep you for under 6 months, I would probably feel like you didn’t do something right. Maybe you didn’t fulfil all of the potential you could have with the client.”

It’s also a way for them to interview the client for 6 months. On the flipside, they need to know if it’s an environment they like working in. Do they like the challenges? Do they like the people they work with? Do they like the corporate environment? You know, those sorts of things. So the idea is that at the end of two years, we help them find a job, we kind of become—I don’t want to use the word ‘recruiters,’ but we help place them—we don’t really become recruiters, we become placers, I guess. I don’t know what the word is because we’re not really recruiting them, we’re placing them.

Kirill: Yeah, agency kind of thing.

Andy: Yeah, agents, maybe that’s a good way to say it. I’m very involved with them when they’re trying to make their decisions on what they want to do if they don’t get an offer to stay with The Information Lab. They all approach me and want to talk to me about what their options are and things like that, and I kind of become their coach again. I help coach them through their career decisions.

Kirill: That’s so cool. And if some of them go to The Information Lab, do they then go on to those short-term consulting projects or—

Andy: Right, exactly.

Kirill: Okay. Interesting, so interesting. So, they work at the Data Science Lab. And do you get any feedback from that transitional period, from being at the Data School for four months and then they move on to their first client, do you get any feedback on that transition, do they feel like cultural shock or a corporate shock?

Andy: We have a couple of people, they’re on the core consulting team, but part of their job is to check in with the Data School-ers while they’re on their placements and just kind of see how it’s going and maybe discover any issues they’re not bringing up or things like that, and then if I need to get involved, they’ll let me know if I need to meet them for a coffee and talk to them, or whatever it might be.

We use a tool called Convo, which is kind of like a combination of Slack and Facebook for Work. It’s an amazing collaboration tool because our company is primarily remote, everybody is remote, so it’s a great way for people to post their issues and talk about what they’re working on and ask for help and things like that. So, we stay in touch with them pretty often. While they’re on their placements, they have lots of avenues where they can ask for support. A lot of times, the hard thing is for them to learn that it’s okay to ask for support, there’s no such thing as a dumb question.

But in between their placements, they have a week back at the Data School and they’ll all present about what they did with their client and we come up with a list of things that they want refreshers on or things like that. So there’s a period in between their placements where they come back and we kind of run a couple day workshop for them.

Kirill: Okay, that’s fantastic. Thank you so much for the overview. I’m sure this has gotten a lot of people excited about the Data School and The Information Lab. And in parallel to all of this, you also run vizwiz.com. Tell us about that. Why did you start the website and what’s it all about?

Andy: It’s been going a long time, over—gosh, I don’t even know how long I’ve been doing it now. Actually, Andy Cotgreave from Tableau credited me with creating with the first ever Tableau blog, which I think is pretty cool. I actually started it really as a way to remember how I was doing things, as a way to kind of document my learning along the way and things like that. When I look back at my first few blog posts, I’m like, “Oh, my gosh, that was terrible.” But I actually think it’s great because it shows progression as well. And that’s one of the things that I try to encourage people that work with me, is “Don’t ever delete anything from your profile because it shows a progression.”

So I started it primarily as a way to document what I was learning along the way and to kind of share with the Tableau community. I think the Tableau community is very unique in software in the way that they freely give to each other and I hope that I’m a good example for people for that to kind of give freely and don’t expect anything in return. I think that’s the biggest thing, you know, it’s very important for people to share and not necessarily expect it to be reciprocated. That’s not why you should be sharing in the first place.

So I write a lot, at least three days a week, sometimes every day. I’m very regimented with my time, so that helps. But basically on Mondays, I run a project called Makeover Monday with Eva Murray, who works at EXASOL. This is the second year of the project. Last year I did it with Andy Cotgreave, but this year it’s actually really, really kicked on. So that’s basically the way that project works, is every Sunday we post a dataset—even though it’s called Makeover Monday, a lot of people do it on Sunday. So, we post a link to an article and to a data visualization and do a dataset and say, “Here is the original. How can you make it better?” Some of those charts are really good, most of them are bad, and the idea is, how can you tell a better data story?

And it’s really fun because you get 52 different datasets to work with throughout the year that aren’t things that you do at work. It’s really fun. A lot of companies have actually taken it on now. You know, they’ll run their own Makeover Monday sessions where they just give their people time to be creative.

So that’s Mondays. Tuesdays, I try every week to record a video with some kind of Tableau tip; I call it Tableau Tip Tuesday. You know, I used to write out all of the steps that I would go through, but that took a really long time, so I think maybe two years ago now I started doing YouTube videos instead, and it’s much easier for me and it’s much faster because I can just record it once. I make lots of mistakes, so I’ve actually gotten feedback from people that they like that I don’t heavily edit my videos, because it shows how I mess up and things like that. And they actually pick up things that I don’t realize they pick up on.

So I try to do those every Tuesday and then this year on Wednesdays, I started a project with Emma Whyte, who runs the Centre of Excellence at The Information Lab. We do a project called Workout Wednesday. Last year we called it the Data School Gym because they are basically challenges that the Data School-ers would give to each other. And then we decided, “Well, why don’t we just open it up to the community?” I would say Makeover Monday is for any audience, where Workout Wednesday is Tableau-specific, and it’s for people that are really looking to kind of move on to that next level. So, it’s not really for beginners, it’s for people that want challenges. These are typically very difficult, they take you hours to do, but it’s all about developing and learning.

And then Thursdays I have another website called datavizdoneright.com where I kind of highlight—I follow lots of blogs using Feedly, and on Thursdays I highlight something that I think is a great example of data visualization and I write about—you know, it’s really simple, it’s just a picture and then it’s what works and what could they improve, which is kind of the same format that I use for Makeover Monday. But it’s a way for me—again, I essentially created it for myself as a way to have a place where I could go back and see visualizations that I really liked and maybe get some inspiration.

Kirill: That’s fantastic. I’m actually looking at your Makeover Monday, one of the most recent ones, “how has access to toilets changed for girls in India?”

Andy: Oh, yeah.

Kirill: So there’s an initial visualization which wasn’t very telling, which was kind of hard to read, and then you changed it into a Tableau interactive—

Andy: Yeah, I did a slope graph on that one, yeah.

Kirill: That’s so cool. It sounds like you’re very busy. How do you combine that with four children?

Andy: I’m very good with my time. I’m also training for a marathon at the moment, too.

Kirill: How? What does that mean? You mentioned your regimented on time. Can you give us some tips, some insights?

Andy: Sure. I typically wake up between 5:30 and 6:00. On my running days, I get up and I take the train about halfway in, and then I run the rest of the way or sometimes I run all the way from home, it kind of depends on what my training is for the day. So, I train for marathons Tuesday, Thursday, Saturday and Sunday, and then I recently bought a bike, so now I’m biking to work Monday, Wednesday and Friday, which ironically saves me 30 minutes a day each way. Maybe you’d think it wouldn’t, but it does.

Kirill: London, yeah?

Andy: Yeah. Pretty much, yeah. So, you know, I do all that stuff and I’m usually done by 7:30 in the morning. So I’m done before most people are even awake with my exercise. And doing that early on the weekends doesn’t interrupt my family time either. And then I do most of my blogging between 8:00 and 9:00 in the morning, and then the Data School starts at 9:00. So I do that through the day and then at night, I head back home and that’s my family time.

Kirill: Oh, fantastic. So you get all of those extracurricular activities out of the way before you even start work?

Andy: Pretty much. Most of the time, yeah.

Kirill: That’s really cool. All right. And in addition to all of that, of course you are a Tableau Zen Master. Can you tell our listeners what a Tableau Zen Master is?

Andy: Well, I think that’s part of the mystery, is it’s kind of undefined. First off, there’s no checklist for somebody to become a Zen Master. You won’t find a list anywhere. You know, Microsoft has their MVP program. Well, that’s a checklist you can follow and you get it automatically. I think it’s a designation. I think that’s the best way to describe it. It’s people that are experts at Tableau that freely give back to the community. I like to think of it as kind of a select group of people that go above and beyond what anybody else does in the community.

Again, a lot of it is undefined and that’s intentional. So every year, I always wonder if I’m going to get it again. Like, next year I’ll try to do more than I did this year so that I can continue to show my value to the community. That’s kind of how I see it. It’s really, really neat. Tableau does things for us throughout the year. They give us special access at the conference, they have special areas set up, designated for us where people can come meet us and things like that. That part is fun, but it’s a bit weird as well, because it can be kind of like a celebrity thing, which is really weird for a piece of software that you just like using to play with data, but it’s kind of cool too, because then you get to meet people that are leaving comments on your blog and things like that, so it’s really fun to me to put a face to the names and things like that.

Kirill: And speaking of Tableau, are you happy with the way the software has been developing? Like, it’s introduced lots of new features and it’s very exciting. Do you see a bright future for Tableau in the coming years?

Andy: Oh, of course. If I didn’t, I wouldn’t use it. What do they say? Imitation is the sincerest form of flattery? Every other data visualization project is trying to copy what Tableau does, which means Tableau is doing something right. And what those companies don’t realize is they actually build it in a way that isn’t nearly as user-friendly as what Tableau does. And they’re in a constant game of catch-up because Tableau is innovating so quickly that how can somebody ever keep up with the innovations that Tableau is doing? So, for me, Tableau is fun. It makes data analysis fun and I never thought I would say something like that.

Kirill: Yeah, totally. So, before The Information Lab, you were at Facebook. Tell us a bit about your time there. You were already working on data visualization. How did that all go?

Andy: Maybe I should step back before that.

Kirill: Yeah, sure.

Andy: I worked at Coca-Cola before Facebook, and I worked in revenue management. And one day the director of our group—we were in the annual planning process, and the way that works is you sort of set budgets for the sales team and then they have to plan to those budgets. And basically, we were using Microsoft cubes in Excel, which was fine, cubes are great if you know the answer to every question you’ll ever ask. But we couldn’t really create dashboards. And we wanted dashboards so people could know how they were tracking toward their goals. You know, they have their goals and they have these plans they were creating, so how do the plans compare to the goals sort of thing. Very simple, right? In theory, it’s very simple.

And we didn’t have dashboard software, so I just happened to google “dashboard software” and Tableau was the first thing that came up and I read the little Google description that comes up when you search, and I was like, “Oh, that sounds kind of like what I’m looking for.” So, I go to the website and I download it, I watch a couple of videos and I had the entire project done in 30 minutes. And at that point, I was absolutely hooked.

You know, Coca-Cola went through a reorg because they do that every 18 months, and my boss actually encouraged me to move into a different role where I could get to do this kind of work more. She knew that I love doing data analysis, and I love presenting, and I love using Tableau. So she helped me get into a role where I was doing price promotion optimization, which is a mouthful, but basically what that means is “how effective are promotions?”

I worked with the sales teams for CVS and Walgreens and Rite Aid, who are three massive drug store chains and also convenience stores as well, huge Coca-Cola customers, and basically I got to work with the sales teams and go on sales calls and present promotional analysis I was doing for them. It would be things like—okay, let’s say it’s a 12-pack promotion, a 12-pack of cans, and they might sell 3 for 9 versus 4 for 12. While the price point for those is the same, the consumers behave very differently to those.

And what was really fun—all those places have kind of like shopper cards or loyalty cards. When you would start looking at the loyalty card information and what are people buying when they buy Coca-Cola versus when they buy Pepsi? And how much is the value of the rest of that basket—that’s a pretty common term in CPG for consumer product companies, is the “rest of the basket,” basically which one drives people to buy more stuff in the basket, Coke or Pepsi. Well, no surprise it’s Coca-Cola. So, it was really fun because then I would get to go take this analysis and sit directly with customers and let them ask me questions along the way.

You know, we give a presentation all done in Tableau, no PowerPoint, and one of the things that really blew them away was that I could answer their questions without saying, “Let me come back to you in three weeks,” and help them really drive their business and be kind of an unbiased, holistic adviser. So, we would tell them, “These are the weeks you should run Pepsi. These are the weeks you should run Coke. These are the promotions you should be running.” And then to see their business take off was really, really gratifying.

So I presented that case study with the kind of promotion analysis work I was doing at the Tableau Conference in 2011, and then I got a phone call shortly after from Facebook. Actually, I used to have kind of like a chat widget on my blog that broke, and I used to get messages from people all the time just asking questions or e-mails asking me for help and things like that. I got one from this guy named Namit Raisurana at Facebook and it was just like, “Hey, Andy, we’re looking to implement Tableau. We just have some questions for you.” I was like, “Okay, when do you want to chat?” and he was like, “Can we talk this afternoon?”

So I get on the phone and there’s three or four other people on the phone and about halfway through, I realized that it was actually an interview, which is probably the best way to be interviewed, not knowing that you’re being interviewed. And then they made me an offer to move to California and kind of run the Tableau and data visualization practice at Facebook. And that was a really incredible experience, a lot of fun. Yeah, so I did that for a couple of years and then, you know, I kind of went through the story of how I ended at The Information Lab earlier.

Kirill: That’s so cool. I just caught myself that I was listening to your story as if I was watching a movie. That is so interesting, the experiences you went through.

Andy: I kind of pinch myself. It’s like, if I had never done that first Google search, what my life would be like now? I know I wouldn’t be as satisfied as a person, so I owe a lot to Tableau for that.

Kirill: Yeah. It’s kind of like—when you’re ready, you get these opportunities presented to you and you just need to grab them by the horns and go with it.

Andy: Yeah.

Kirill: That’s so cool. On all of that, I have an interesting question for you. You’re an inspiration to many. There’s thousands and thousands of people who read your blogs and follow you on Twitter and people that you’ve helped in Data School. Who has been an inspiration to you along your journey?

Andy: There’s been a few key people, I think. Alpa Sutaria, who was my boss at Coca-Cola before, when I worked in revenue management, the encouragement she gave me to pursue my passions was massive. I read a book at the time that I believe she told me to read called “48 Days to the Work You Love” by Dan Miller, and it has a bit of a religious slant to it, so I kind of ignored that part of it, but it helped me identify what I was passionate about. I started reading that book at the same time that I started using Tableau. So I don’t think that was a coincidence. But I owe a lot to her for pushing me to do the work that I love to do.

And then the boss I had after that, James, was just incredible. You know, I loved doing the price optimization work. It was so much fun going and visiting customers. It was all stuff that was all new to me and I really enjoyed it. And then when I got the job offer from Facebook, I went to him and I told him, “Hey, James, I’ve got this offer. What should I do?” And he said, “You are an absolute idiot if you don’t do this.”

Kirill: Wow! That was your boss?

Andy: Yeah, exactly. And then from the Tableau community perspective, there’s tons of people to mention, but Dan Murray, who I met in Atlanta and I met at the very first Tableau conference in 2008, he’s always been kind of a mentor to me, you know, encouraging me along the way, pushing me, always a sounding board. He was the first person I called when I got the offer from Facebook, and he was telling me that I’m an idiot as well if I don’t do it and that sort of thing. Those three people in particular have been huge influences on the early part of my career.

And I think here at The Information Lab, Tom Brown, the owner of the company, he’s absolutely incredible. I think I get a lot of things done, and then I see what he does and I don’t know how he does it. But part of it for me is that he lets me do the work that I love to do. You know, he doesn’t muddle with the curriculum at the Data School, he just lets me do it. He trusts me. You know, having trust from your boss is really, really gratifying.

Kirill: Yeah. It sounds like you’ve been very fortunate with bosses and mentors in your life.

Andy: Yeah, I’ve had my fair share of bad ones as well, like everybody has, but I think that helps you recognize when you have a good one, and it also helps shape the way that you manage people. You know, I try to treat people the way that I want to be treated.

Kirill: I totally agree with that. So once you find a person who’s good for you, who might be a mentor, a boss, maybe an employee or a friend, how do you keep that relationship going and how do you find some way to maximize the lessons that you learn from that relationship?

Andy: Oh, my gosh. There’s probably a book in that somewhere. (Laughs) I think part of it is not being afraid to ask for help. When somebody offers their help to you, take advantage of that. You know, I’m still in contact with them. Either they’ll reach out to me or I’ll reach out to them and say, “Hey, how are you doing?” I think it’s about keeping those doors of communication open and letting them know that you’re doing well and you appreciate them and that sort of thing, because it’s good for them to hear that as well.

I think it could be very difficult to recognize when somebody is a good mentor for you. A lot of times you don’t know until after they’re no longer your mentor. I think that’s how you know they’re a great boss. If you don’t realize they’re a great boss, then that probably makes them a great boss. I don’t know if that makes sense or not, but in a weird sort of way, it’s almost like you don’t realize it until it’s gone sort of thing. I look for people that encourage me and inspire me in the work that they do and that’s what I look to do for people that are around me. I try to give back to them as much as they give to me.

Kirill: Fantastic. I want to move on to a bit of a different question, but also relating to your career. What has been your biggest challenge in this world of data?

Andy: Oh, boy… Probably access to data. Companies make it way more difficult than they need to to let their employees find insights in their companies. Let me give you a couple of examples on polar opposites for those things. When I was at Coca-Cola, they used Teradata for their back-end storage, but you’re not allowed to access the data. So how are you supposed to actually do your job as a data analyst if you can’t get the data? So, we would have to actually go through a lot of hoops. We used MicroStrategy because that was the only thing that was allowed to connect directly to Teradata. We would basically run tabular reports in there that we would then export into CSVs and import into a SQL server that IT didn’t know we had.

And the thing is you hear those stories all the time because people are just trying to get stuff done. And IT is too often a roadblock. So, that’s how we worked around it at Coca-Cola. And what was great is, for example, when I started using Tableau and we wanted to buy some licences, IT wouldn’t buy it for us because it wasn’t on their official list of things that pay them back. I know there’s a lot of side deals done in these things, I’m not ignorant. They didn’t want to support it so we just bought it on a credit card and expensed it. But, you know, when it came to renewal and the CFO is using your reports every day to run the business, guess how easy it is to get approval then for that expense. IT still didn’t support it, but the CFO did so it got paid for.

So that’s kind of how we worked around it at Coca-Cola. Actually, one of the biggest concerns Facebook had about hiring me was would I be able to handle going from a very corporate culture at Coca-Cola, a very tight restrictive type of environment to a very open culture at Facebook. At Facebook, all the employees have access to all of the data. That’s the first thing. You’re trained in the beginning what data you are allowed to access and what data you are not. And if you access data that you’re not allowed to access, you’re fired. It’s very, very simple and I’ve been there and I’ve seen security people walk up and walk people out the door. That also frees up people to get the data they need for the right project. So that was very, very liberating.

Those are kind of the polar opposites of the data access part of it. I swear, every data analysis project that I’ve ever done is 80%-90% data prep and 10%-20% of data analysis and data visualization. I think that’s probably pretty common. You know, right now I use Alteryx now that I’m at The Information Lab, and I wish that I had that a long time ago. It would have my life so much easier.

Kirill: Yeah, I totally agree with you. And speaking of getting datasets, if you can share this, where do you get your datasets for your Makeover Monday and Tableau Tip Tuesday, Workout Wednesday, for all these amazing projects that you’re doing? You need a lot of datasets.

Andy: For Makeover Monday, it’s actually pretty easy. There’s a lot of crappy visuals out there. A lot of people send them to us. I also get some of them from the blogs that I follow or the websites that I follow. For example, there might be a particularly poor visualization that I think would be a great topic, or it’s timely, or something, but the data is just too hard to get together, then I’ll just ignore that one because I want it to be as minimal data prep as I need to. But even I have a very, very extensive backlog of topics. You know, we’ve probably got at least another year’s worth in our backlog right now.

The Tableau Tip Tuesday and Workout Wednesday, those are typically done with kind of canned data sources that everybody gets when they install Tableau, so that way they can follow along, but I do publish those as well so that if they want to use the same dataset that I’m using, they can. So, Tableau Tip Tuesday and Workout Wednesday, those are done with kind of standard datasets that everybody has. But Makeover Monday can be challenging, but I timebox myself on the data prep and I say, “You know what? If it’s going to take me more than an hour to get this data ready, I’m just going to use something else.”

Kirill: Okay. That’s a pretty good way to approach it because I sometimes spend days on data prep – not hours, days.

Andy: Yeah. I think Makeover Monday in particular, anybody can go out there now—we’re in Week 33 this year, so there’s now 85 unique datasets out there for people to practice data analysis with.

Kirill: Yeah. That’s really cool. So, guys listening to this podcast, check it out. There’s a whole—yeah.

Andy: It’s very similar to Data Is Plural and data.world. They’re curating datasets now as well and it’s great.

Kirill: Yeah, those are some good websites. Okay, that was an interesting answer to that question. I’m sure a lot of people have that similar challenge. And on the flipside, what’s been a recent win that you can share with us in your role or in your career?

Andy: Oh, boy, that’s a good question. A recent win? Gosh, every day feels like a win.

Kirill: (Laughs) That’s good! That’s a good life to live!

Andy: You know, it’s hard because I try to live my life where if I don’t learn something every day, I feel like I’ve failed. I’m constantly trying to learn. For me, every four months I start with a new group of Data School-ers, and they know very little when they start. So I get so much satisfaction out of watching them grow every day. You know, I get little wins like that every day. I love helping people. And Eva and I have started—we now have our own BrightTALK channel so we can host webinars whenever we want.

Maybe recently, one thing that’s been particularly fun is we do this thing called Makeover Monday Viz Review, where people can submit things that they’ve created for Makeover Monday onto Twitter. We use the hashtag #MM for Makeover Monday, and #VizReview, and we will review it live on the webinar. We don’t look at them ahead of time and we just give our initial impressions. Those have been really fun, and we’re getting lots of really good feedback from the community on that. It’s been great working with Eva on the project because she’s got all these amazing ideas and so much energy and it’s been really great to have a partner in the project that really pushes you and makes you better.

Kirill: That’s fantastic. And I can attest to that, that having a partner on the project or a business partner is really good. It helps you see things from a different perspective. Sometimes you get stuck in your own mentality, and then when you speak to somebody else, you get different opinions and it frees up your mind to approach things differently.

Andy: Yeah. The funniest thing working with her is she’ll send me a Whatsapp message that says, “I have an idea” with the emoji that’s like the straight face showing the teeth, you know, the—(Laughs) That’s when I know I’m in for more work. But I also know it’s going to be a great idea, so it’s funny.

Kirill: That’s really cool. That’s amazing. I’ve got a philosophical question for you here.

Andy: Oh, boy. Okay.

Kirill: Yeah, if you thought that what we had so far isn’t enough, here we go. Where do you think the field of data science is going? What should our listeners prepare for if they want to be ready for the future that’s coming?

Andy: First off, I think data science is a bit of an overused term because everybody is kind of a data scientist now, or everybody should be a data scientist now. I think the field is actually just starting. It’s immature as far as how long it’s been around, I guess, in the modern sense of data science. You know, data is everywhere around us now, and whether it’s people that want to understand their fitness, or they want to manage their diet or whatever, they can do all that kind of stuff through data and the Internet of Things and wearing a Fitbit or whatever it might be. I think people are starting to see how data science and data analysis and data visualization can impact people’s lives. You’re seeing a lot of really good social causes people are doing now because these charities don’t really have the capabilities in-house or the technical expertise, so there’s plenty of chances for people to freely donate their own time and things like that. I mean there’s so much opportunity and there’s so much need as well. The number of companies that need data scientists is far greater than the number of data scientists that are available. And with the way data is exploding, it’s only going to become a bigger gap. So it’s super encouraging to see all of these different avenues people can take, whether they want to learn Tableau and Alteryx, or they want to learn D3 and R, or whatever it might be. Whatever you’re passionate about, pursue it. You’ll find a job.

Kirill: Totally. I love that advice. Andy, we’re coming to the end. Thank you so much for coming on the show. Where can our listeners contact you, find you, follow you so that they can see how your career develops from here, or get some more inspiration and insights from things that you share around?

Andy: Sure. I’ll just give you a quick rundown on some of the websites. Well, I guess I’ll start with Twitter. On Twitter, I’m @VizWizBI. @VizWiz was taken and they wouldn’t give it to me even though they don’t use it. So, I’m @VizWizBI for Business Intelligence, it’s the best thing I could come up with at the time. And my website is vizwiz.com, and then I also have datavizdoneright.com, and the Data School is thedataschool.co.uk, and Makeover Monday is makeovermonday.co.uk.

Kirill: Okay, gotcha. Are you active on LinkedIn at all?

Andy: Yeah, people can connect with me on LinkedIn.

Kirill: But probably Twitter is better, yeah?

Andy: Either one is fine. You know, LinkedIn is good. If somebody wants to add me, make sure they tell me how they heard about me, otherwise I’ll think they’re just some person trying to just add me to their network. I’m pretty picky about people I add. So, if I just get a request and I don’t know who they are, I’m going to decline it. I think it’s always good on LinkedIn when people add a little personal note when they try to connect with people, so that would be some advice as well.

Kirill: Fantastic. And one last question for you: What is a book that you can recommend for our listeners to help them become better at visualization and data science?

Andy: Well, I think there isn’t one book for everybody. I actually wrote a blog post back in May of 2016 called “12 Books Every Great Data Analyst Should Read,” so if you just go on my blog and you search probably just ‘books,’ it’ll probably come up. In that, I go through data visualization books by Stephen Few and Cole Nussbaumer, you know, by the “Wall Street Journal.” I think Alberto Cairo is someone everybody should read, “The Functional Art” in particular is an absolutely fantastic book.

But I also think people should be pretty well-rounded, so read books about how the brain works. “Brain Rules” by John Medina is one I would highly recommend. And there’s also a really neat book by Austin Kleon called “Steal Like an Artist,” the subtitle is “10 Things Nobody Told You About Being Creative.” So, I think it’s about reading a broad set of books as well.

Kirill: Yeah. Fantastic. Guys, definitely check this out. There’s even pictures of the books here, so you can just click on them and get the one that you want. We’ll include the link in the show notes.

Andy: And I don’t get money for them. These are all books I’ve read.

Kirill: Fantastic. Once again, thank you so much, Andy, for coming on the show and sharing all these insights. I’m sure you’ve inspired so many people to at least check out Tableau if they haven’t yet, but more importantly, to get stuck into data visualization and start building their own personal brand.

Andy: Yeah. Thank you very much for having me. I appreciate the invitation.

Kirill: So there you have it. That was Andy Kriebel, the Head Coach at the Data School at The Information Lab, and I really hope you enjoyed it. It was a very inspiring podcast and I personally really enjoyed the way Andy told the story of his career and how he’s progressed through different companies and done different things and how he gives back to the community. Personally, my favourite part was the way that Andy—actually, two favourite parts: The way that Andy constantly learns, that he doesn’t stop learning, and that for him, success in a day is if he learned something new. And I think we all can learn from that. I think we all can adopt that. That’s such a great mentality to have in life. On the one hand you’re learning new things, and on the other you always feel fulfilled, feel successful from the days as they’re passing by.

And at the same time, the other thing that I found very inspiring was the way that Andy gives back to the community. So, he’s not doing it for revenue, he’s not doing it for fame, he’s not doing it for himself, he’s doing it for the people, the people he’s helping and it brings him joy to see how the people he’s been able to help, how they’re progressing and how their lives are changing, the transformations that they’re going through. That’s very, very inspiring and I think we can all also learn from that and be more like Andy in that way.

So, I hope you enjoyed this podcast. Make sure to connect with Andy. The best place to find him is of course on Twitter, his Twitter handle is @VizWizBI, so check it out. Of course, you can connect with him on LinkedIn, make sure to follow his blogs, which are vizwiz.com and datavizdoneright.com. And if you’re in the U.K., maybe check out the Data School at The Information Lab. All of these links will be available in the show notes at www.www.superdatascience.com/91.

And if you did have fun in this episode, make sure to share it with someone you know who’s into data visualization, or who could benefit from getting into the space of data science. Like Andy said, it’s a very, very young industry, very young space, and it’s going to grow and there’s going to be lots of opportunities, lots of jobs. So if you know somebody who might benefit from this, who might be inspired by Andy’s story, make sure to send them this podcast. You might change their life. And I look forward to seeing you here next time. Until then, happy analyzing.