Tools from the Pros #2: Joe Mako on Tableau

Ok guys, here we are with a new interview of Tools from the Pros, the series in which I interview data visualization professionals about their favorite tools.  This time we have Joe Mako talking about his experience with Tableau.Before I start telling anything about Joe, let me tell you how I ended up  interviewing him. I was looking for an expert to interview with proven experience in designing advanced visualizations with Tableau, so I decided to ask to some twitter friends. Result? Lots of names but only one always there: Joe Mako. If this is not enough give a look to the impressive list of video tutorialshe has in his blog.Joe is employed at S2 Statistical Solutions where he does data integration and visualization. This is what Joe wrote when I asked him to send me a short bio:

I have used Tableau extensively since 2008, creating interactive viewpoints of data to enable people to get answers to their complex questions easily. Currently, I specialize in integrating complex databases from health insurance companies, hospital networks, and the government to enable better evidence-based decision making. I am active on the Tableau user forum, solving a variety of situations for many Tableau users ranging in skill from beginning to advanced.

I really enjoyed reading his interview. He provides lot of interesting references and links. If you are thinking about using Tableau I am sure his tips will help you a lot with your final decision.

How did you start using Tableau?

About three years ago in 2008, I had been reading FlowingData for a few months when I noticed Tableau was a sponsor and decided to check out their software to see if it could help with some projects I was working on. I felt like I was decent with formulas and VBA in Excel, but always had trouble making a decent chart. When I first saw Tableau in action, I knew it would make my job of making sense out of numbers easier because a good chart was easy to make. The first big project I used Tableau on was reporting on data quality and monitoring the cleanup of the records. With the guided analytics Tableau enables, I was able to make interactive dashboards allowing a view to see what records were wrong, why we knew they were wrong, how much revenue was lost because of the error, and then tracking what records got fixed and the increase in revenue. The project was a success, and I knew creating visualizations in Tableau was my passion. In the past three years, I’ve rarely gone a day without using their software, being a part of the Tableau user community has become a big part of my life. The many great people I have meet, and the friendships I have gained by participating in the community are most valuable.

What’s the best and worst aspect of Tableau?

The Tableau Data Engine is the single most valuable feature I would miss the most if Tableau was removed. I don’t know if there is a specific term that can fully describe it, because it is unique, and it has a long list of benefits: bulk text loading, super fast aggregations, incremental appends, and just all around seamless experience. Suffice to say, if I am working with data, and I don’t need a real-time feed, I’m loading it into the Tableau Data Engine (TDE) every time. It has been called an “in-memory” database, but that may not be the most accurate term for it because it is not like other “in-memory” databases. Instead of loading the entire data set into RAM, the TDE intelligently selects what data to load into RAM, so that we can work with data sets larger than our available RAM. So I am not sure if there is a good way to compare it to other data storage systems other then knowing that the TDE was created specifically to work with Tableau, and it is a beautiful thing.

Tableau is a focused and opinionated piece of software, meaning it is not a complete solution, but for what it does enable, it does great job. The number one thing I believe is lacking from Tableau is easy to use and fast statistical functions. Currently, with custom table calculations, and data preparation, I have found that Tableau can compute nearly any calculation, but it is too much of a work-around to force the software to do something it was not designed for, because it adds unnecessary complexity, and commonly makes the interaction slow. There is already a built-in delay with published workbooks (waiting for the Server generated images to download), and the additional delay of waiting for the computations to be evaluated becomes a major drawback.

Ok, I am a beginner and I want to learn Tableau, where do I start?

Tableau provides phenomenal training resources for free. Their On-Demand training and Live Online would be the first place to check out. There is no shortage of interesting workbooks that you can download and inspect or try to re-create from places like the Visual Gallery, their Blog, and there are live workbooks embedded throughout their website (I don’t think I’ve found them all yet). Their Knowledge Base with over 300 step-by-step guides on how to accomplish useful tasks. Then with the Q&A Forum there is no shortage of interesting situations and people like myself eager to help you accomplish what you want in Tableau.

How is the learning curve vs. return-on-investment of Tableau?

I remember learning Tableau was real change in my approach to data, and I still feel like I learn something new about Tableau every day. My experience in learning Tableau has been like playing a game, the first things are easy, and some really amazing analysis can be created with just the use of the mouse. I think of it like “The Princess Rescuing Application“, specifically slide 16, where it is a series of short learning leaps, and each one brings joy with accomplishment.

While Tableau on the surface has a clean interface, many complex operations are just under the surface, a click away, and once you know how to do something in Tableau, it becomes simple and fast to perform. The main exception is custom table calculations where there are a multitude of non-obvious factors effecting their evaluation. I believe an understanding of SQL would make Tableau more understandable and less mysterious. If you need to make sense of numbers, the return-on-investment is easy to see, things that take hours, or require programming, take minutes and drag-and-drop inside of Tableau. I consider it having a conversation with my data when I use Tableau, because as quickly as I or the person next to me can ask the question, Tableau enables me to provide the answer.

What other tools would you recommend other than Tableau?

Once you understand Tableau’s approach to data, I am sure it will be clear that Tableau does not stand alone as a complete data solution. While Tableau is fantastic at the human-centric tasks, it does not perform tech-centric tasks (see “BI Has Hit the Wall” by Stephen Few), and you will need software to help you prepare your data for Tableau. Every few months I am changing my tech-centric applications as my needs change and I try new things, but I think Pentaho Data Integration (Kettle) is wonderful for ETL. There are many ETL applications out there, and I recommend trying them all to find the ones that fit your style and needs best.

4 thoughts on “Tools from the Pros #2: Joe Mako on Tableau

  1. Mike Nealey

    There will be many, I am sure, but I’d like to thank Joe for all he does to help others. I am sure the founders of the internet had people like Joe in mind.

  2. Dimitri B

    Well said…
    My “path to Tableau” is strangely similar to Joe’s. If I am permitted to add to Tableau’s pros and cons, I would add clean, uncluttered, friendly and easy interface as a pro, and the fact that one needs a PhD in Rocket Science to master Tableau’s custom table calculations as a con (Joe somehow mastered them without one, however).

  3. Hari

    Nice one Joe. Well described.. I hope tableau doesnt have 64 bit machine. isnt so? so how it handles big data ?


Leave a Reply

Your email address will not be published. Required fields are marked *