Last week, I had the opportunity to jump on a call with Tableau Ambassador, Yuriy Fal, all the way from Moscow! It was an absolute pleasure chatting with him, and I learned so much about him! For starters, he's been using Tableau longer than any customer I know about, he's helped tons of people, and he's got a thing for jazz.
Yuri: I think it’s been about 8 years now. As I remember, I started using it in 2008. I worked in a distribution company that sold computer accessories and printer supplies. It’s a high volume business. My friend and former colleague, started this company some years ago, and he invited me to join. So I worked for them for 4 years. I was a product manager, doing business buying stuff from Taiwan, Hong Kong, and Europe. I had enough spare time to do things other than buy containers of DVD-Rs and trucks of supplies. That’s why my boss told me to pay attention to support my colleagues in buying and selling decisions by creating some analytics for them. And that is how I started working in an analytics role.
When Tableau 10 was released, there was a question in the community (actually ) about our Tableau stories. I recall my : I was asked by my boss to calculate a cash conversion cycle (CCC). Cash is very important in the distribution industry. Being short of cash on a high turnover means literally blowing up a business in just a moment. Every day we should see the whole picture and understand how we are doing. Are we moving quickly enough (and hopefully in the right direction)? Or are we doing something that increases the risk of cash shortage – up to the point that is out of our control?
Answering those questions wasn’t easy, you know. For analysis, we used to use Excel -- sipping data from Oracle reports (we had a Data Warehouse then ). It was very time consuming and error prone. I tried to find a way to escape Excel for the presentation of the data. At that time, Tableau was starting to be more active in marketing and making their product visible to others. It coincided with the time I was looking for a solution. It was lucky for me. That's how I got Tableau and started to using it. I think it took a couple of months to see the power of the instrument (it was version 3 at that time).
Tracy: Did I hear you correctly, did you say that you started on version 3?!
Yuri: Yup. Frankly, it was version 3.5 (as I could remember).
Tracy: Oh my goodness! How did you even find Tableau then?
Yuri: I started with Googling two key words, "data visualization." Why I looked up those two words back then? I don't know exactly. Luckily, I found Stephen Few and other players in the field, and I felt that it was innate to me. I felt I could build up to this and be serious about data visualization.
Back to the business. I did the cash cycle analysis and made some thought-provoking slides for my boss and other top managers. At the time, we were having some arguments negotiating with our vendors. All that lead to tough decisions with big consequences for us (including reorganization & layoffs). It was at that moment that I began to understand that visuals worked (albeit in unexpected ways).
The next year I joined a service company (a subdivision of a major retail electronics chain, one of the top players in Russia). I worked with them for 3.5 years making analytics to support the services business. They have a lot of service requests, a lot of things to repair, and they have the network of service centers all around Russia -- 50 or so. Thousands of people were involved -- always busy, having a lot of work, and plenty of data. So, I tried to support the field stuff, service engineers and team leaders to have their operations effective, day by day. My work mainly was about their everyday business questions. I tried to give some quick answers (and usually I missed by a hefty margin ). That's how I came to understand that the major KPIs are about timing and quality of the service, not about money, not about volumes, not about vendor’s compliance. Instead the internal focus was on the customer - as in, we need to do our job as quickly, but as effectively as possible. With customer satisfaction and service quality KPIs we would have teams compete, and we'd have rankings of service centers and such. It helped to keep it like a game. Weekly meetings usually began with a ranking chart followed by the “Team of the Week” celebration.
One moment that flashed for me was when we had a simple bar chart that could tell that we had some open service requests that took too long - days (or even months). But this insight was buried when we had aggregated data. If we do lots of service requests quickly, in a day or less, then these 10's or 100's of requests make the averages look like nothing is wrong. So we needed to disaggregate the data, and we found these metrics, and we were able to target the long overdue requests. These requests were floating up – popping-up -- before the eyes of the managers and the eyes of team leaders. We had a process which was using this dashboard as a main tool. Every day people were targeting this overdue request, and trying to do everything they could to close them sufficiently to make customers happy. This simple instrument worked. Every morning we saw that managers and team leaders were starting to see how things were going, starting from this simple view on their computer screens. I felt that we targeted them correctly by showing a simple and effective visual. From that point on, I have preferred to do simple things which have an impact. I can see people quickly understand it and have discussions about the figures. I do all of this without the help of Excel. I do it right in Tableau. We’ve used live connection to SQL Server as the backend – but nothing more, nothing in between. It was version 5 by that time. And it was great. They were rewarding times. We were a big company - several thousand people working - but it was felt like a small team working together.
Tracy: I love your point about simplicity. I think people will sometimes add more to make it look prettier, but it's really about getting the message across. Like you said with the bar chart.
Yuri: Yeah. In terms of prettiness, not one single manager demanded it from me, as I can remember. Everybody just wanted the tables from me, but I think making simple charts instead of tables is better. They usually agreed.
Tracy: It's easier to see what you're looking at on a bar chart than a table. Ok, my next question for you is, do you use Tableau at all in your personal life?
Yuri: No, not really.
Tracy: Not enough time?
Yuri: That is another explanation. I think I have a little personal life that I share with my wife and my daughter. My daughter graduates from the university this year. She is thinking about a career in science. I also graduated from Moscow State University with a degree in biology, as well as my wife. That's how we met each other. Our daughter studies soil sciences, her diploma is on the bio-diversity of yeasts – it’s close to what we studied. She really enjoys it, and doesn't want to be done with university
Life in Russia can be hard sometimes, depending on the circumstances. I'm too old to move, so I'll stay here, but that's why I'm thankful for Tableau and being able to be a part of a global community. I have fellows and friends all around the world because of the community. Thank you to you, Tracy, and to your team and colleagues for giving me the opportunity.
Tracy: Well, thank you! It goes both ways, you've helped us just as much (if not more) as we've helped you.
Yuri: We could be giving each other thanks for hours! Let's move on to the next question.
Tracy: How did you learn to use Tableau?
Yuri: Mainly, I used Tableau "on my skin" - on my own. A lot of the learning resources were not available when I started. I try not to make big mistakes, but I am not afraid of making mistakes. I think I’ve made them all! So many times, I have deleted the wrong workbooks, and made data sources unresponsive when doing a certain query (there were no extracts then ). With time, with experience, with work, I overcame all of these. Now, I feel that Tableau has enough stuff to overcome this learning curve. The first hours, the first weeks of working with Tableau are just a breath of fresh air - doing complex stuff and getting answers quickly. At times, I found that the answers I have found in Tableau are just the beginning, just the tip of the iceberg. When doing real analysis, you should think hard, think a lot, about how to prepare data. I try to learn more about how analytical databases work. I found some bright people in this area -- Daniel Abadi (invented C-Store which eventually became Vertica), Peter Boncz (MonetDB & Vectorwise) and Thomas Neumann from the Hyper team. They are absolute heroes of our field. Their work is foundational to us. They’re doing the Database Science.
To answer your question about learning, the people are the answer. Finding the people who are doing things right is of great help. I'm very thankful to Joe Mako, Jonathan Drummey, and Tamas Foldi. Simon is very bright and energizing, and his curiosity is beyond the limits (cheers pal ). I learn from them and I hope that they've learned something from me (the cross-pollination works).
Tracy: What’s your favorite moment from the community?
Yuri: The greatest moment with the community that I remember was two years ago on the 13th of February, when I was following a thread about moving distinct counts. It had been a thread that was open for 2 years up to that moment. Tableau 9 was in Beta. We were all feeling that the LOD could give us power to make a lot of things that weren’t impossible, but very hard - and make it possible. When we were discussing moving distinct counts, the problem was clear to me. You need all this detail, such as customers and days (or even smaller time slots) to get the distinct counts together for the view - say to see the weekly moving distinct count (7-day window). Joe Mako proposed a solution based on table calculations which was very simple - if you get it, but it needs the full details, all the customers, all the days on the view just to get the picture. With millions of customers and number of days, it pushed the limits of the Tableau engine. I felt we needed to have the database do the work - have some stuff pre-aggregated and then lookup from one or another partition.
It was a late Friday night, and I was a little bit tired (and I think some good things come when you are tired, it's a rule of life), then something flashed in my brain. 7 days is a full week, and if we could aggregate the week up to one day, such as one day should be up to Monday and then next to Tuesday. If you could aggregate up to one day, then distinct counts are counted for one day, easily. So we needed just 7 partitions -- pre-aggregated up for each week day, and it should go back to the previous (say Friday). With that idea, I looked at the Level of Details once more, with the aggregation that we should be able to do this without table calculations at all. I made all this andlate at night, and went to bed. When I woke up the next morning, I saw the reply from Jonathan. It was very heart-warming for me to have this recognition from him.
People like Jonathan Drummey are doing so much for the community, learning and sharing. I couldn't stop working for the community. It's very rewarding. My wings are just starting to stretch. It’s a very good feeling, like a narcotic of some kind. That's why I've been answering more questions, and have been getting my hands into the tough questions. Crow's Nest is a great place to start with new questions. I am not a Speedy Gonzales, I am not the guy who answers quick and dirty. I prefer to think about stuff, and have time to think. That's why I choose questions on the community that no one else is trying to answer. Usually, it's questions that are unclear - you need to ask the topic starter to provide more context. Or the questions that don't have straight forward answers, and you need to be quite creative - sometimes you need to work out of the data structure - reshape it or change the structure. These types of questions are very helpful for the topic starters, and finding the answers are important for me. I feel better and confident in using Tableau when I get to answer questions that were hard to solve at the first sight. And looking at the other very smart, very talented, power users of Tableau - it feels like people are doing great stuff here.
I am very jealous of the people who have their hands on Tableau Server, such as Mark Jackson or Jeff Strauss. They have the experience which I don't have access to. From time to time I'm able to do some Server stuff with my customers who have Tableau Server environments, but they (the environments) are not that sophisticated (but the customers are). I think I need some real Tableau Server training and hands on experience. Looking at the people who are doing this as their jobs, I feel I need to get more involved with the Tableau Server stuff.
I am thrilled about the HyPer engine and the performance. Hope it would really be magnitudes better in the “tough” conditions (I mean extract-heavy dashboards). Also, my customers are waiting eagerly for the Linux Server. It's very exciting.
Tracy: What is your biggest feature request?
Yuri: The main ingredient that is not here right now is to have a DML/DDL access to Data Engine. Not just from within Tableau. Not via Extract API, but being able to make a SQL request to the data engine. You cannot make SQL statements to the engine from outside clients or make changes to the data already in extract. It is my hope for HyPer as the successor for Data Engine (and possibly the engine to run Repository in Tableau Server). If HyPer would be exposed as the full-fledge SQL database engine – it would be great.
Tracy: How did you discover the Tableau Community?
Yuri: I think I discovered it little by little. When I had registered, I think it was in 2013 - when I started working for Analytika Plus, a partner of Tableau and a consulting company. I started working with them, first as an independent contractor, and then as a member of the team. One of my responsibilities was to find the answers to the questions they and their customers asked at the moment. Of course, I went to the forums, as we mainly do with other vendors. What was surprising to me was how vivid the forums were, and there were lots of bright people, bright personalities in there, and I found Shawn Wallwork, Joe Mako, Richard Leeke and Matt Lutton. Everybody was awesome. I was very excited about finding the topics they were discussing. They were not boring. Typically, if the answer wasn't already there, someone was asking what are you actually trying to do? Maybe this is not what you are asking about - for example, maybe you're asking how to create a table, but really the question is about finding the answer that you can share with your colleagues and it should be in some kind of graph or visual to look at instead. The discussions are very vivid and interesting. They are not just about the technical stuff, they are also about the people problem stuff.
I feel the people in this community are friends. Some are sharing what they have, freely to all.
If we have a dozen of energetic people in the community who are energizing others, it's great. It really matters to me. I am proud to be part of them (though I am far from being an Energizer).
I am very excited of what Shinichiro Murakami has been doing in the community forums. His productivity and energy is ski-high.
Tracy: Is there anything else you want the community to know about you?
Yuri: I don't have much spare time for hobbies, but if I had time for one, I would say it is jazz. I am very fond of jazz. Typically when I'm on the run, I have my headphones on and I am listening to jazz. There are communities of jazz musicians who are playing with each other, jamming with each other and making great stuff together. I think the jazz musician communities are very powerful. They ship a lot of music to us. I have jazz music I listen to when I need to work and be in the flow.
Tracy: Who are some of your favorite jazz musicians?
Yuri: I think the one who is my hero is Jaco Pastorius. He is a bass player genius.
Tracy: I'll listen to him today while I'm working away.
Yuri: Yes! You need to listen to his song "Portrait of Tracy." I've made my homework for you to go do
Follow Yuri on the Community - https://community.tableau.com/people/yuriy.fal
Follow Yuri on Twitter - @yurifal