Tableau Ambassador Spotlight with Jonathan Drummey

Version 1

    Last week, I had the chance to have an interview with Jonathan Drummey about his career path and journey with Tableau. Jonathan started using Tableau right around the time that I started working here, so he's one of the first customers that I met and got to know. He's always a pleasure to talk to and has had an interesting story (i.e. massage therapist before being a data analyst). Read more to learn more about Jonathan and get some of his insights on Tableau and the community.


    Jonathan with James Baker (on the left) at TC16.



    Tracy: Okay well let's get started. Let's start with how have you been using Tableau?

    Jonathan: Six years this year. I first downloaded 6.0, but I actually started using Tableau with as 6.1.


    Tracy: And how do you use Tableau, at work and in your personal life?

    Jonathan: In my work nowadays at DataBlick I do Tableau training, consulting, and support with our clients. Primarily my work is in healthcare, where there’s a huge variety of different data requirements.


    Tracy: And you started using Tableau when you were at Southern Maine Health Care, is that correct?

    Jonathan: That's right. I was the first data analyst in the quality management department. They only had Excel, and one person who had one Access database with two tables in it. I knew we needed data visualization software to get the data out to clinical staff in a usable form. They had a lot of data in different sources, but they couldn't get the data out. So I leaned on Tableau for that, and built out a number of different dashboards and visualizations. Primarily using Microsoft Access as the database, and then eventually we were able to directly connect to the SQL Server and MySQL databases running our electronic medical record systems. At the tail end of my time at Southern Maine, we brought in Alteryx. I’d known about Alteryx for a while, but I was one of those people who knew how to do SQL and could make the data dance and thought I could do it all myself. Then the lightbulb finally went on about what Alteryx could do, and that was a game changer in terms of being able to do data prep.


    Tracy: How long did it take you to learn Alteryx?

    Jonathan: Alteryx is my second favorite piece of software, after Tableau. It took me only a few hours to be productive. DataBlick is an Alteryx partner and we attended an Alteryx training, and in the training I came to the conclusion that Alteryx is visual programming for data. So instead of having variables that we’re manipulating with lines of code, we're manipulating rows and columns of data by visually connecting and configuring the Alteryx tools. I had this thought in the training: “You know, a Tableau workbook is ‘just’ a stream of XML data so how about I write something that will take apart a Tableau workbook and put it back together?” So after a couple of days of Alteryx training I built an Alteryx workflow to create a form of dynamic parameters for Tableau. The workflow read in the Tableau workbook XML, parsed it, found the parameter you wanted to update (from the comment for the parameter in the Tableau workbook), and used that information to grab data from another data source to rewrite the parameter list in the Tableau XML and tie the Tableau workbook all back together again. That was my first weekend with Alteryx. I must say that I did it with a little help from Joe Mako. Since people talk about dynamic parameters in different ways, this was the “create a refreshable parameter list” use case. (Check out how to do it here.)


    Over time at Southern Maine Health Care, I’d connected to more and more data sources and connected them to Tableau. Each data source had its own maintenance and overhead connected with it. I was connecting to so many data sources, dealing with changes to both data sources and changes to the views and dashboards that I'd built that all the maintenance was taking up most of my time, I didn't have much of the bandwidth to do any of the new stuff that was coming along. Having a good ETL/data prep tool like Alteryx enabled me getting some of that bandwidth back. It made that whole process so much easier and faster.


    Tracy: Another data question. Where do you get data that you do for personal projects?

    Jonathan: I don't really have personal projects nowadays.


    Tracy: Haha, what? You don't have a lot of time on your hands?

    Jonathan: Right! But I guess the biggest personal project that I have is in the healthcare space. Right now there is a ton of awesome public health data, but if you're working inside a healthcare provider and you want to see a Tableau viz and using a data set that looks like data you have, it often doesn’t exist. There's no Superstore equivalent for healthcare, so my personal project off and on for awhile now has been thinking about what that needs to look like...or what multiple sample healthcare data sets could look like. I'm currently building out a data visualization class (with Anya A'Hearn and the University of New England) for UNE’s health informatics program. As part of that I've been making up usable data sets for healthcare. I've got some blinded data sets and also some public data sets that I’m modifying to be more like something that could come out of an electronic medical record system.

    I'll be putting these data sets out to the public so people in healthcare can use them. My personal interest is helping Tableau users in healthcare be able get their questions answered. Currently they have a harder time because so much of their data is confidential or is protected health information (PHI) that legally can’t be shared. Going through the blinding process makes it that much more difficult for people in healthcare to share the cool things they're doing or get their questions answered on the forums.


    There are about 18 different factors that cover protected health information. I did a presentation for the Healthcare Tableau User Group that talks about the blinding process. The process to blind data and really do it right takes a lot of work. For example, if you have a set of patient data that has diagnoses in it, there are now about 75,000 distinct diagnoses codes, there’s even a code for a burn due to water skis on fire. If there’s a diagnosis like that that has only a few cases it could be possible to identify a particular patient, so it has to be removed from the data. 


    Tracy: Oh man - so many things I had never thought of! Have you always been in healthcare? Even before Southern Maine Health Care?

    Jonathan: Before I was at SMHC I was a part-time stay at home dad and a massage therapist.


    Tracy: Had you always been interested in healthcare?

    Jonathan: Before any of that, I was working at a software development company that made a billing and customer care system for telecommunications and internet companies. It was UNIX based system based on relational data bases, was highly configurable, and we were going up against companies that had mainframe based systems. We were growing 50-100% a year for years. We were started by a professor from MIT and some of his grad students and had an academic focus, for example we wouldn't implement features until we thought we really knew how to implement them. So somewhat similar to Tableau in that sense.


    I had a variety of positions over the years there and then I left that and took some time off to finish growing up in some ways. I lived and worked in a few different retreat centers and I experienced several different types of massage therapy. The personal transformation I went through led to me going back to school so I could learn how to help others. Part of what was so fascinating to me is that if you look at somebody's posture and how they move in the world there are all these patterns of behavior. Depending who's counting there are about 600-800 different muscles in the body, and so in one sense doing massage therapy is a giant multi-variate problem of “Where do I most efficiently do work with this particular client to help them relieve their issues?” If someone is coming in with the archetypal “tight shoulders and neck” problem, I’d do a little research and maybe find out they're sitting in front of a computer all day, and their shoulders are getting a little hunched because they're trying to mind meld with their computer to do their work. Their tight shoulders are there because they are off-balance from leaning towards the computer, and they're sitting all day, so in fact their hips are also tight. That's a connection to what is happening in their shoulders. If I did some work on the muscles and tissue around hips, then their shoulders start letting go. And that helps them have more awareness in the future when they're leaning forward or not, and gives them more choices about how they move in the world.


    So for me there was a part of doing massage therapy for my clients that was about data analysis. It was just a very different data set, in a very direct physical manipulation kind of way. And I was horrible at marketing myself, so as independent massage therapist, that was challenging. As a family we could make it work for a while, but my daughter was getting older, and then the quality management data analyst position became available in the local healthcare system. They wanted someone who could talk to the IT side, to have the technical background to be able to do that, but also have the human piece as well. So translating from the IT world to the nurses and doctors who don't have that background. I was somebody who obviously had both.


    Tracy: Yeah - you were the unique candidate they needed!

    Jonathan: Yup. So they hired me, and it worked out really, really well.


    Tracy: I've never thought of massage therapy like that - from a data analyst perspective.

    Jonathan: Here’s a story about that: When I was first out of massage school I was working at a clinic where I was seeing a couple of clients a day, several multiple times a week, I was doing a lot of massages on a regular basis. Massage therapy is physical labor. The average career length of a massage therapist is between 2-3 years. One part is the physical labor piece. We are human beings and we resonate with each other, so you have to have really good personal boundaries to not take on the client's stuff, and not empathize with them so much that you are getting the same pains that they are. That certainly happened to me sometimes. The other part is that a lot of massage therapists are independent contractors, like I was, and there's a whole skill set there that is necessary to be successful, and if you don’t have it life can be more difficult.


    Tracy: I can imagine. Essentially, you're a small business owner.

    Jonathan: Yup, exactly. Going back to the clinic story: I was doing all of these massages, and I'd be tired at the end of the day. I got to noticing that the days that I was more tired, I was less grounded and connected with myself. But at end of the days, where I was more connected with myself, I would feel more satisfied, and though I'd be tired, I knew I had done good work. But I also noticed that the days I was more grounded, I got bigger tips - like hugely bigger tips.


    That helped me really listen to my massage clients, to really try to understand what was really going on for them, and that is a skill set that I have carried forward with me. Whether I'm answering questions on the Tableau forums, or working with a DataBlick client in trying to answer their data question, there's this place for me about trying to get out of my own head, get really grounded, and really listen to them, listen to what their needs are. When I do that, I ask deeper questions to really identify what it is they are trying to do.


    For example, so often on the forums, the presenting question is someone is trying to build something – for example a grouped bar chart, and they’ve tried some different things and have gone a ways down the path of trying to build that chart and now they are asking for help. And it can be a pain to build grouped bars out in Tableau. I can start asking “What you are really trying to do here?” and then they say, “Well, I'm trying to show one bar for the data and one bar for a target,” or something like that, and we do a deeper dive around what their real goal is. And maybe instead of the grouped bars a bullet chart or a bar with reference line suddenly makes more sense. And that bullet chart or bar with reference line is way easier to build in Tableau, better at answering the underlying business question, and ultimately better at meeting the person’s needs.


    There's been a lot of different paths that I've taken in my life that seem to come together now, in working with data and working with people.


    Tracy: It really has. It's a unique story! In a somewhat similar vein - how did you learn to use Tableau?

    Jonathan: Trial and error. Lots and lots of error! I devoured all of the online training videos. I have read the online help manual twice. I've read it for version 6, I've read it for version 8, and we're now version 10, so it's about time to read it again. (My first job out of college was as a technical writer so I have a soft spot for manuals.) A lot of looking online and web searching. There weren't so many Tableau blogs at the time. I read everything Joe Mako, Richard Leeke, James Baker, Alex Kerin, Shawn Wallwork, Andy Kriebel, Andy Cotgreave, and other people posted on the forums and elsewhere to learn from them. That was a whole chunk of my learning.


    Another thing is that I like to explore the edges - so what can this product do and what can't it do? I also have to go over the boundary of something, and fail. Or is there a way that I can force and shove it to make it happen? That's where Joe Mako blows my mind. One of the key things I've learned from him is that Tableau is a data driven tool for visualization so the rough boundary of what we can do in Tableau is defined by the structure of our data. If we change the structure of our data, we can change what we can do in Tableau. Whether we are pivoting or pre-aggregating or cross joining in different ways or do things like data densification, we can make magic happen in Tableau. That was a whole area of interest for me.


    Some of the context in which I did that learning was through answering questions on the Tableau forums. I really wanted to answer the person's question and get it right. At the time, there were people like Richard and Joe there who would write these long, detailed responses of what they'd done and they were demonstrating a place of going through the process of trying to understand what someone was really trying to do, maybe providing multiple solutions to their question, and explaining how each solution would work, and where it might run into problems. That was something that naturally fit for me, as well. My work in the hospital at the time only exposed me to a limited number of data sets so being on the Tableau forums exposed me to different data sets and different kinds of problems. By trying to answer people's questions, I'd learn things and then carry them back to my own work.


    Annnnd, in some ways, I probably spent too much time on the forums after and during work. Answering people's questions can be a little addictive!


    Tracy: I agree with you on that one! It sounded like you discovered the Tableau Community by just doing a couple of Google searches and exploring everywhere.

    Jonathan: Yup. I would search for something in Tableau, and the answer would come up - and back then it would be one of Joe Mako's answers on the forums. Something else for me was that I was the only one at my organization using Tableau, and there weren't a lot of people in my area either. I went down to the Boston Tableau User Group when I could, but there were no local people who I could bounce ideas off of on a regular basis so having the community on the forums and Twitter was just fantastic.


    Tracy: So what would you say is your favorite part about the Tableau Community?

    Jonathan: The attitude. It was James Baker, I think, who started the community, and then Dustin Smith and you came on. I appreciate you for your passion for Tableau and Tableau (the company) for supporting us users and giving us a place to connect. Another part is that the first users who got involved really wanted to help people get their questions answered. Joe has talked about getting into the shoes of the person asking the question, and understanding what the world is like from their point of view, and from that understanding answering their question. So having that as part of the culture in the Tableau community, really resonates with me, just out of my own set of values, and the work I've done professionally over the years. That culture where when someone is asking a question we're not telling them it's a stupid question. I don't think that's ever been written on the forums, and hopefully, never would be. The person asking the question has a need to be met, and we're a bunch of users who love Tableau, and want to help them meet their needs. That's what really stands out in the forums for me. Some of the people that I've met there have become friends or co-workers - so it's just an awesome community. It makes things like user groups and Tableau conferences sort of like "old home" week in the sense of meeting a bunch of old (and new) friends that I've never met in person.


    I think there's also something to be said in that we're using a tool for data analysis, so we're coming at things with a learning mindset. When somebody comes in and helps somebody else - we get excited that people are helping each other learn more - even the person who is answering the question is often learning. It's part of the culture of the forums, where we're all helping each other learn the product better, learn data analysis better, learn how to do data visualization better, and improving that. Outside the forums, things like the MakeoverMonday project that Andy Kriebel and Andy Cotgreave started (and Andy Kriebel & Eva Murray are continuing) are fantastic. We're skill building, and it's a community that supports skill building.


    Tracy: I've never really thought about it like that before. But you're right, and you need to be open-minded if you want to learn, and because of that, people are open minded when they come to the community, and it makes it a more enjoyable place to be.

    What advice do you have for new users to Tableau and/or new members of the Tableau Community?

    Jonathan: I've actually done a couple of blog posts on this! Here are links: Getting Good at Tableau and So You Want to Be a (Tableau) Zen Master?

    The variety of visualizations that Tableau can build and the number of different things we can do in the product itself, the combination of those things is so huge - we have to focus. Pick something and work on that and learn about that. Don't try to completely learn and understand everything about visual design, table calculations, mapping, and Tableau Server all at the same time - it's too much. Narrow your focus, and pick things to study and learn. So my take was starting with calculations and table calculations, and really understanding those and how they worked. That's what got me going.


    Find new ways to use Tableau if you want to improve your skills at Tableau. Whether it's in your work - if you're used to using Excel to build something force yourself to build it in Tableau - even though it's something you can do in Excel in a couple of minutes, and could 30-60 minutes to figure it out in Tableau, try it out, and you'll learn about the product. Something I’ve come to believe is that learning Tableau is like learning three different languages: There’s a "language of Tableau" that is how to use Tableau-the-tool. Then, there's the "language of data visualization". How do we effectively help people see and understand their data? And then oftentimes Tableau enables organizations to work faster with their data so ultimately they can be working with more or different kinds of data. So there's the "what kind of data is that??" question. Learning languages takes time, so having some focus about what it is we want to learn and do helps.


    This year in Tableau, I am working more on my design skills because I have this co-worker, Anya A'hearn, who does amazing design, and I'd like to learn from her. She critiques my designs - she's lovely, and she can be incredibly harsh.


    Tracy: I don't know what you're talking about - she's terrible at design

    Jonathan: I get to learn from Anya, and from elsewhere too. She's been doing design for a very long time so things that are very easy for her, are very new for me!

    The other part of your question was about getting involved in the Tableau community. The way I approached it was more in a step by step fashion. I didn't try to answer questions right away. What I would do is go through other people's answers and go through the answers they had built out, and go through the workbook and steps until I understood what was they had done. The next step, was to take a problem that had been answered by somebody else, and try to answer it myself without looking at the answer. Then I would have the workbook that had been built out to be sort of an answer key. I remember one workbook that Shawn Wallwork did that had this really neat layout for the axes and grid lines on a scatter plot, and I was trying to figure out how he did it. Turned out it was a background image.


    Tracy: Ha! You were probably beating your head against the wall trying to figure it out.

    Jonathan: Yeah, exactly! But it was a part of Tableau that I didn't know about so doing that was part of a whole learning process. Then, eventually, I graduated into answering people's questions for myself. So really, having this focused model of wanting to learn Tableau - the areas I picked questions in were around that focus of calculations and table calculations. I wasn't trying to answer all the mapping questions out there because I didn't know about all the different map projections, among other things. So that was my whole way of getting involved.


    Writing and blogging can be another way to participate. With my background as a writer I like to write down what I’m learning. When I was first learning I ended up with 60-70 pages of notes on Tableau in a Word document, and I turned it into a wiki on my blog and that has been a whole resource for me. Sometimes I'll do a web search on a Tableau question and my own site will come up! I’ve heard other Tableau bloggers say that as well about their own posts.


    If you’re interested in blogging one thing to do is find a topic that you know about that isn’t so common. For example, I know some about healthcare so I might blog about that. Or, another example, I haven’t seen a lot of information about Tableau for retail. If that's something you know about, you can start blogging about it and people will point to it. You'll be learning and sharing with the community. You'll teach yourself something you didn't know, and people will ask you questions, and you'll learn and make more connections with people.


    If you have an interest in building your life around doing more Tableau work then being in the community - whether it’s the community forums or on social media - is a fantastic way to build up your resume because employers can see what you did, and learn more about you, and you can demonstrate your skills.


    Tracy: Totally, it becomes a portfolio for you.

    Jonathan: I can say that's why I was able to leave where I was and be at DataBlick now. I have this body of work that is public, and I've been able to help other people, and we get clients out of it.

    Certainly it helps that Tableau is a growing company, but I think part of why Tableau is growing is because of Tableau's fundamental mission of helping people see and understand their data. It is this mission of sharing, and learning from one another. And the community holds that as well. If I figure out some technique or new way of doing things, hoarding that for myself doesn't really win much, whereas if I share it in the community, all of us win.


    Tracy: Yeah - it's a pay it forward mentality.

    Jonathan: Yeah, or in a way - also a pay it backwards, as well. At one point Joe had close to 3,500 responses in the forums in a year or something like that. One way I honor the effort that he put into that, that helped me learn so much, is by sharing what I've learned as well. And hopefully, people who see what I write will be inspired to do the same thing.


    Tracy: I know for a fact that you've inspired many people out there.

    Jonathan: Thank you! Last year there were a bunch of people like Ville Tyrväinen, Robert Rouse, and Rody Zakovich doing all this really cool stuff with Filter Actions in Tableau.  I enjoyed seeing them advancing, learning from each other, and showing techniques that I didn’t know were possible in Tableau.


    Tracy: It is very inspiring when you think you know something so well, and someone brings something new to your attention.

    Jonathan: Yes! Someone else is Steve Wexler, he’s built out this wealth of knowledge of how to work with survey data in Tableau, and gives it all away, check out


    Tracy: To your point before, there's not a lot out there on survey data.

    Jonathan: True, and I think part of the reason for that is that when we search for Tableau and survey data, nothing but Steve’s blog comes up, because he's done such a good job at it!


    Tracy: Is there anything else that you want the community to know about you?

    Jonathan: Maybe a quick little plug for DataBlick. We're a Tableau & Alteryx & Mapbox consultancy. We do consulting projects and training for our clients, and something else we offer is our HelpMeDataBlick service. It’s our version of Tableau Doctor, and it’s a way for companies to get short term consulting services. For example, you have a dashboard that you know needs some help and you're going to be showing it to your CEO, and you'd like to spend a couple of hours tuning it up. We can do that, and it’s very fun work.


    I created the @helpmedatablick Twitter account where I tweet out a Tableau tip of the day. If you’re reading this and have tips for things that helped you, or time savers, please let me know! I’m always looking for them.

    Finally, one more thing about getting involved in the community - don't just do Tableau and data viz online! There are many, many community groups and non-profits, city/county government that data visualization can help. Whatever your interest is you can use data to help serve people.


    Jonathan's Community Profile: Jonathan Drummey

    Follow Jonathan on Twitter: @jonathandrummey

    Jonathan's Blog: Drawing with Numbers | Thoughts on data visualization and Tableau