Data visualization and dashboard design enthusiast with a strong focus on the business end-user. #VizItPhilly

Twitter: @josh_tapley
Website: https://vizification.com

Josh Tapley is a native Rhode Islander and, alongside his family of four, moved down to Philadelphia about 4 years ago when he had the chance to join Comcast.  Josh’s wife Melissa is working on her dream to become a librarian (splitting her time between school and working part-time at a local library) and they have two girls, Autumn (15) and Willow (12).

He is well known in the Tableau community, and I had a chance to have a conversation about his journey in Data and Tableau.

Toan: Tell me, how did you get into Data Visualisation and Tableau?  How did it all begin?

Josh: I was first exposed to data visualization around 2009 when I attended Edward Tufte’s One-Day course. At the time it was eye-opening and really changed the way I approached everything. When I noticed that management consultants at work were already employing these techniques, I quickly worked to build out those capabilities in-house and started building my first full analytics team.  From there we migrated from Excel and PowerPoint to interactive dashboards with Business Objects and Xcelsius, both now owned by SAP.

Before long, I had caught the visualization bug and was spending my nights and weekends building dashboards and entering competitions. Then in December 2011, I left my career in healthcare to join Cleartelligence (a Business Intelligence consulting firm in Boston) as their Data Visualization expert and first full-time employee. Since then, they have grown their data visualization practice to include others from the community like Dustin Cabral and William Strouse.

It was during my time at Cleartelligence that we determined the power and flexibility behind Tableau. From there I re-entered the healthcare field to roll out Tableau and, finally, I was recruited by Comcast in Philadelphia to build out a cutting-edge Data Visualization team develop their Tableau Centre of Excellence (CoE).

Toan: Good to see another person from a Business Objects and Xcelcius background. A lot of people have heard the term Centre of Excellence; can you tell us about the process of building one?  What was your thinking?  What did you try to achieve? What was the overall experience like and the challenges you faced?

Josh: When I thought about a Centre of Excellence (COE), I was really focused on enablement, training and providing resources. Ultimately, I found out that at this point that CoEs often have a negative connotation because they are often strict governing bodies that force rules upon the way that you will develop and deploy things. Since I am an analyst at heart, that is not what I meant to accomplish.

A Community of Practice is generally a more favourable thing, where users band together for something that they are passionate about. So we basically combined the two concepts to have a more positive experience in supporting our users. Challenges that I faced were around the support and funding for such a CoE; since we are not charging back to the business for the services that we provide, we often had to do it in addition to other work. Ultimately that meant that we spent most of our free time helping and enabling others, but over time we gained support from our leadership to dedicate more people full-time to that effort. It certainly does not hurt to have several hundred departments praising our team.

I think that our last virtual training session had over 400 attendees.

Toan: Tableau projects come in different sizes, I am curious, what has been your most rewarding work in the Tableau space?

Josh: For me, the most rewarding part has been teaching and mentoring others. I really approach this in 3 ways, as a manager I recruit and develop talented individuals, as a co-leader of the Philadelphia Tableau User Group (PhillyTUG) I help coach and engage the local community, and finally, as an adjunct professor of Data Visualization for Temple University’s graduate program I am able to inspire and mentor students who may have never realized how awesome Tableau is. 

Knowing I helped influence amazing folks like:

This by far is the most rewarding part.

Toan: According to the last Partner Kick Off I was at, Tableau currently has 500+ user groups globally, this is truly phenomenal and shows how active the community is. For those who are not lucky enough to have a TUG, what would be your three main considerations before starting one up? What resources and support is available?

Josh: For starting a Tableau User Group, there are always people active in the online community (forums & Twitter) that would be happy to help you out. Tableau’s Community Support and local sales representatives are also a great resource, and there is reasonable reimbursement provided for each meeting. However, since we are in a large city, the reimbursement does not help much since it usually covers less than half of a meeting space. If you meet after hours for example, most of the surrounding companies need to pay for security and room rentals as well as work with in-house food and beverage vendors.

This is where it is critical to network with Universities, a wide variety of local business and Tableau partners to make this a success. The other major factor is to not go it alone. It is certainly achievable, but if you do not have support (and diverse ideas), you will burn out quickly and may not offer the breadth of content that speaks to a wide audience. Before I started a user group, I would also test the waters to see if there was enough demand. Maybe an informal lunch meetup would be a great and affordable way to gauge interest and meet other folks who are dedicated enough to take time out of their day.

Toan: You seem to have done some interesting projects, what was the most challenging Tableau project you have been a part of, and why?

Josh: My most challenging project with Tableau was in collaboration with former Zen Master, Pooja Gandhi. Together we worked on near-real-time (NRT) reporting on some very large streaming data sets. 

We were leveraging Kinetica as a GPU database with a failover on MemSQL, but they did not support dates the same way, so every change required two development tracks. Ultimately with well-planned aggregations and passing URL parameters into the custom SQL, we were able to generate incredibly fast reports. Many people that we work with push millions of records to Tableau and have challenges with both Extract times and performance.

Toan: Has the introduction of Hyper helped the extract times and performance?  What are the things people should look at if they are experiencing the same challenges?  What was your thought process for solving this issue?

Josh: Our team works closely with our data engineering counterparts to optimize the aggregations, and here Hyper has certainly displayed better performance. On the CoE side, most of our challenges are around slow database response times (Hadoop) or trying to run massive queries or highly complex processes (nested volatile tables in the initial SQL) during peak hours on the database. The Hyper load, therefore, is only as fast as the underlying connection. From a performance perspective, we are seeing a tendency to bring in more data with the additional capacity.

Toan: What would you do if you could go back in time?

Josh: If I could go back in time (do not all Deloreans time travel?) I would bring some of the tips and tricks videos from the past few Tableau conferences. Those would have made my life so much easier as I was getting started with the program.  Four years ago, the advice I received was to ‘click it out.’

Toan: Let us jump into the DeLorean and go back in time, we are at the Tableau Conference with all readers in attendance, what are your top 5 Tips and tricks?

Josh: In no particular order,

  1. Understanding the Tableau Order of Operations.
  2. Clicking on the colored box next to a data item when editing colors to sample a custom color.
  3. Saving images as image@2x.png to get a higher resolution on better displays.
  4. Keyboard Shortcuts: Like right-click + drag (Windows) or hold the option key (Mac) to automatically show the “right-click menu”.
  5. Using a text box as a dividing line.

Toan: Tableau is amazing, where do you see Tableau going in the next few years.  What features would you love to see, or functionality improved upon?

Josh: I think that growth in the enterprise space would really benefit by making Tableau the BI tool of choice. I know that they are working on this, but there are still some gaps between Tableau and other tools that cause us to support multiple platforms. If we could go ahead and improve some things like tables (thank you for the Data Tables extension!) directly in the tool, we could move away from other legacy tools and focus on standardizing with Tableau.

Toan: Your welcome for the DataTables extension, it was a fun thing to build. I think that Tableau was initially the data visualisation tool that sat beside the enterprise reporting solution but now we can export to PowerPoint, and soon will be able to subscribe and receive PDFs. What advice would you give people who are looking to adopt Tableau as their sole data analytics solution?

Josh: I think that most of the other roadblocks are down. Things that I would still like to see would be better drill-in-place, better dashboard controls (like we are starting to see in 2019.2) and the ability to govern resources by site (backgrounder processing time and the ability to restrict tabcmd access) so that some people cannot monopolize the system.

Toan: Tableau has an amazing community for learning and there are loads of places to go.  What is your top tip for really excelling within Tableau?

Josh: My tip for excelling with Tableau is to find a topic that you are interested in exploring (sports, comic books, movies) and try to make something awesome. You will encounter challenges along the way, but it will be a labor of love and you will forget how hard you are working to learn new techniques.  It is like watching TV while you exercise.

As for learning from the community, engage with individuals that inspire you. You will motivate them to stay active and they will likely challenge you in return. In the end, you will likely become friends, and that whole experience is more powerful than any software tool.

Toan: Social media is a wonderful tool and does connect a great many of us. However, what would you recommend to someone who is not social media savvy?  Are there any resources, books, courses, websites etc that you can recommend to those that are perhaps somewhat introverted?

Josh: I would recommend Matt Francis’s course on http://lynda.com and watching key videos from the recent conferences, like Ann Jackson and Lorna Eden’s “Tableau speed tipping.” For books, I highly recommend Storytelling with Data, #MakeoverMonday and the Big Book of Dashboards.

Finally, I would stress that blog posts and the free Tableau help videos posted on their site is how I got started with Tableau. Use those resources and practice often.

Toan: Lastly, can you please write a paragraph or two about yourself?  Not the Tableau expert, but the person.

Josh: We also have a dog (Bilbo) and 3 cats (Ramona Flowers, Zelda, and Vincent Vega). Our pet names should give you some insight into how nerdy and pop-culture obsessed our family is. When I am not consumed in some sort of visualization activity (work, teaching, user groups, vizzing for fun), which is rare, I enjoy gaming. Board games, video games (e.g. DOTA) and my favourite would be pen and paper RPGs.

Since I do not have a lot of free time, I try to make it to events like PAX Unplugged a few times a year. I am not sure one can really viz without music. At this point in my life, I am clearly into alternative rock, but I still find myself listening to a lot of punk and Ska on Spotify lately.

Data visualization and dashboard design enthusiast with a strong focus on the business end-user. #VizItPhilly

Twitter: @josh_tapley
Website: https://vizification.com

A Conversation with Season 1:

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.