I had the honor of spending an hour chatting with Simon Runc about how he discovered Tableau, the Community, and his love for data. If you ever want recommendations on some fascinating books to read, Simon has some! Oh yeah, and did I mention, he's a cat guy? You probably will recognize him by his avatar which is a picture of his cat, Mitchy
Simon: I've been using Tableau for about two and a half years now. I first heard about it from my current employer, and managed to convince my Employer at the time (The Body Shop/ L’Oreal) to get me a copy.
And from that second on, I was like, yeah, this is the tool I've been waiting for my whole (working) life. I wasn’t really given the time required (at the Body Shop) to use it properly, so it wasn’t until I moved to Atheon that I got the time to really see what she could do.
I think maybe two years ago to the day that I moved to Atheon…and never looked back!
Tracy: Oh, congratulations! So what does Atheon do?
Simon: Thank you. They’re a Visual Analytics company that have been going for about ten years, concentrating on the grocery retail sector. I actually met them about nine years ago (when I worked for one of the big 4, as they are known in the UK) and they were brought in to do a project. I kept in touch with them just because I had an interest in visualization. I always had an interest in data-driven thinking and they showed me that visualization is the easiest way to do that.
We kept in touch, sharing cool things…etc. and eventually, after all my badgering, they offered me a job!
Tracy: What were you doing at the Body Shop? Were you in an analyst role or something else?
Simon: Yeah, I've always worked in analyst roles. So I was the international commercial analyst. The Body Shop was in 65 countries. Each of the countries had their own analyst teams (some had bigger teams than others, but they had generally at least one analyst), and I was looking at the whole of the global picture.
I'd have to pull together key findings from 65 countries and try and make a single, coherent picture.
Tracy: Sounds like a lot of work.
Simon: It was a lot of work, especially since we were mostly using Excel.
Tracy: Did you use anything besides Excel?
Simon: A bit of Access, as I progressed through. I really just could not fit the amount of data I needed to fit into Excel. Without realizing it formally, I was starting to understand the idea of analytical flow….I’d find an interesting story, which like all good analytics would raise another question…and I'll have to go back to the database, re-pull a report…etc. So I wanted to try and get as much data as I could in one place so I could continue my journey. So just shear data size led me to using Access and I learned a little bit of SQL along the way. Thinking that I now have models with 100s Millions of rows, it wasn’t really that big after all!!
Tracy: Interesting. So, then, you went over to Atheon Analytics?
Simon: That's right. And they had been doing data viz for 10+ years (when the phrase Data Viz wasn’t really well known). As I said, I first met them about 9 years ago when I worked at Sainsbury’s and then at The Body shop, they came in to do a project with another visual tool called -- I don't know if you've ever heard of it, called Omniscope
Tracy: Omniscope. No, I haven't. Is it still around?
Simon: Yeah it's still going. They used Omniscope for about seven or eight years until they found Tableau, but still had some stuff built in Omniscope when I joined….Omniscope was quite a good piece of software for its time (I still have a soft spot for her!), and while at Sainsbury’s I’d been trained up to use it. One of my first assignments was to go a do some Omniscope training. I hadn’t used it for a couple of years, so was interested to see how much it had changed/advanced. This is where I got a real appreciation of the genius and consistency of the underlying principles of Tableau….Over the years, they had added in every single request somebody had, so if somebody wanted to do a line in a bar chart, there was a button for that, want to do a network chart, we’ve made a new chart type for that…etc. But gradually, it became almost impossible to train because every single viz type was like its own bit of software.
So while they attempted to try and create a one-click solution for thousands of business problems in a very simple-to-use tool, they'd created the most complicated tool you could ever imagine using.
Whereas Tableau has only really got like three rules that guide everything. And whether you're in a bar or a line or whatever, it's the same, it's just a different mark type.
Tracy: Yeah. That's crazy!
Simon: So it actually ended up being very restrictive, although its attempt was to try and be free and easy. So, yeah, it's a difficult path and it's one of those things -- careful what you wish for.
Tracy: Interesting. Trying to please too many people at once sounds like. You've got to look at the bigger picture.
Simon: Completely. You don't want to end up having to spend 50 minutes trying to create a bar chart.
We do visual analytics, and they were ahead of the game in terms of that. And so I joined them as a Tableau developer. As I said I'd only been using it for about two months or so, but I was given the time, and that included getting involved with the community. My boss is a massive supporter of me spending time in the community because I learn so much.
Tracy: Really? That’s amazing!
Simon: Yes…now I'm spending, you know, more like 8-10 hours a day in Tableau every single day. And it becomes like an extension of yourself after a while.
Tracy: That's fantastic. So obviously, you use Tableau at work. Do you have clients that you're working with?
Simon: We have three general streams; one is that we are a Tableau reseller (and offer training with that as well), although that's quite a minor part of our business. It's not something we really focus on very much. It's more that if we do have a client project and they think, oh, that was really cool, how do you do that? We'll come in and show them, and actually we show them that they can buy three or four licenses (a bit of training) and get themselves up and running. It’s not something that we spend a lot of time pushing and doing. It's normally off the back of projects that we do that.
And then the other part of the business is consultancy. So (generally) clients come to us through various channels and it’s companies who’ve got all this data, and have no idea what to do with it…we normally have them give us their data, we'll come back in 5-10 days and we'll show them some interesting stuff (visually, of course!). Then (if we’ve done a good job) larger projects feed off the back of that. The third part of our service is we host a service that we call SKUtrak, which is built in Tableau… In the UK retail market, if you're a supplier to the supermarket, you get access to a portal in which you can download the data for your products, sales data, depot stock, delivery, missed deliveries, availability, all of those kinds of things.
And about three years ago, we did a consultancy project with a client, and started getting data from these systems and putting it together into (a then Omniscope) model/discovery tool. They kind of realized, actually, we could just plug this into a different suppliers’ data set and SKUtrak was born (although it was called SKUview at the time). So that's what we do now. We offer a kind of a "freemium" model whereby you sign up with us and we will go off and get the data every night (well some code does!) and build up a nice database over time of all of your data and you get a free high-level dashboard, giving you some key things to check and some top-line stats.And then if you then sign up for the paid service, then you get access to the Tableau model. Which is a really deep visual/interactive tool that lets you drill right down into exactly every single part of your business with that retailer….From delivery-into-depot all the way to the end-shopper. So you can see your deliveries to depots, depots to store where those deliveries are being missed, and then you have access to availability, promotion analysis, price analysis, distribution, all that kind of good stuff.
So those are the three things that we do. We do a sort of a mixture of consultancy, software, service that we provide and a bit of resale and a bit of training as well.
Tracy: And do you fall in all three of those or do you focus on a particular part of the business?
Simon: Yes, I do a bit of all three. Including the building of SKUtrak models we use.
Tracy: Oh, that's awesome. I can imagine that’s a very useful tool for retail companies.
Simon: Yes I’d have loved a tool like that when I worked directly for a retailer….although to qualify when I say I built the SKUtrak model, my part is the Tableau side, which although is the visible bit, it is really the easy part of it (you know how fast/easy it is to develop in Tableau!). We are quite a perfectly formed company where each person has an important part in the tool (building web-scrapers, re-shaping/sorting the data, putting it in database (and building me the views), Tableau server stuff, backend web portal stuff…and lots lots more!). I also do consultancy and I do any Tableau training (which I really, really enjoy actually).
Tracy: So how many people are at the company?
Simon: We're actually expanding. So we're up to 17 people now. There were 8 when I started.
Tracy: 8? So, yeah, you've almost doubled since you've been there.
Simon: Yeah, exactly. We’re definitely getting a lot more traction. People think that the big grocery retailers are such big companies that people assume (like Amazon) they have all latest tech and the latest thinking. But actually they are pretty traditional (with an insane amount of Excel usage!). This is (gradually) changing, as more and more graduates from the last five, ten years are coming through, they expect to look at data in a data-driven way. So we're getting more and more buy in. And we're finding more and more CEOs/Senior managers are thinking differently…as they say, without data, you're just a person with an opinion!!
Tracy: Do you do a lot of personal projects with Tableau or with data in general?
Simon: Yeah. I pretty much use Tableau for everything. It actually is a bit of advice that Bill Lyons gave me. And he just said that one of the things he'd advise anybody to do is to do everything in Tableau. My presentations, everything. So I pretty much do everything. So if I'm reading a news story and I find it interesting but feel the article has a bias, if it's got a table, I can copy and paste it into Tableau, look at it myself and see if I agree with what they're saying or not. (One of the overlooked features I think in Tableau is the paste data feature).
On personal projects…I'm a bit of a starter, and I'm not a very good finisher. So I probably have like nine or ten half-finished dashboards I've started. …Like I've got a Brexit tracker, I've got stuff on all kinds of demographics around the world. I've got stuff on land use (I’m a vegan, so I'm quite interested in the land use side of how inefficient meat rearing is compared to feeding people with vegetables), etc. So I've got all of these things which I found insights from, but before I release it to the world, I like, you know, I get a bit over-polishy, OCD about the whole thing. And that can lead me to sort of never finish anything. But I learn a lot doing them, which I then bring into other things. Yeah, so I use Tableau for absolutely everything now.
Tracy: Where do you usually find that data? I know you said news articles, but where else?
Simon: Ideally, I find people who have done kind of academic work on it and have released the data. So they've already sorted it out for me.
I might do a bit of scraping sometimes if need be. The World Bank IMF has a lot of good data as well (or just use GapMinder who have sorted it pretty well). And in the UK, the Office of National Statistics have a lot of free open data as well that's usually quite well sorted. So usually those are my main sort of avenues. Generally, I type a search into Google, but the World Bank, IMF, WHO websites usually also have the best source of curated data.
Tracy: Yeah. Without having to do too much work to the data itself.
Simon: That's the biggest thing as well. I've have so many ideas for other vizzes and other things I'm really interested in looking at. My boss is very supportive, but not supportive to the point of I'm going to go off and spend a month reshaping a load of land-use data to prove the virtues of veganism!!
Tracy: Right. Right. That makes sense. Besides the community, how have you learned Tableau?
Simon: So, yeah, initially when I learned to use Tableau, I watched videos. I watched the introductory videos for everything to get myself going on. In my training now, and what I try and get across to people is that you need to understand Tableau, and not (parrot) learn it. The most important concepts to grasp (IMHO) are VizLoD and to understand the difference between the green things and the blue things. Once you get those concepts down, you are pretty much there! On a 2 day training, I'll probably spend at least a third to a half a day just on Blue and Green pills.
So I find the easiest way, and I much prefer to understand than parrot learn, because if I understand what a blue pill does and a green pill does and how it works, I don't have to remember where to put them and if I want to create something, I can create anything I want to create just by knowing that.
Once I kind of got that, then the rest of it became much easier. So if you need to create something, if you think in those terms, you shouldn't have to remember how to create a panel chart or how to create a waterfall…you can work out -- you can kind of reverse engineer your own brain!! In general, I try and stick to keeping it as simple as possible.
Tracy: Yeah. I like that. All right. What is your favorite Tableau project that you've ever worked on?
Simon: So I would say personal project would definitely be my India viz. Just because I don't get a chance to do, oddly, a lot of mapping stuff. And I absolutely love maps. So it was a chance to do a bit more on the mapping side.
On work projects, one I'm working at the moment that I’m really enjoying, is to create an iPad-specific dashboard for an exec at one of our clients. I've never designed for iPad before (so that’s been fun). And I built a visual tool to help a retailer identify regional preferences in product choice. I’ve loved this as it’s meant quite a bit of mapping, and the tool just does one thing! Generally a lot of the stuff we build needs to be a bit more ‘general purpose’ which means lots of compromises. Building a viz that just does one thing, but it does that one thing really well, is something I’ve really began to appreciate (and will be taking that thinking into future projects)
Tracy: Are you using Tableau 10 or are you using 9, for the iPad project?
Simon: I had to build on 9 so we can still host it on our 9 server. But it's the restriction on screen space and thinking about the actions that still don't work on iPad that made it a really fascinating project and a real massive learning curve.
That's the other thing that I absolutely love about where I work is that I get to do -- although we concentrate mainly on the grocery industry, we have about 10 to 20 percent my boss allows us to go off and do other clients. So we do things like with a school's charity, we've done stuff with like local councils and looking at population demographics.
So, yeah, I am really interested in a variety of things. I'm just interested in -- I just really want to know how the world (really) works. That's all I want to know.
Yeah, it's that very simple goal. How does every single thing work? That's what I want to do.
Tracy: Haha, yes, it's a very simple goal.
Simon: It’s super interesting…and I’m also really interested in the way people think (and make decisions)…all that thinking bias stuff…how people can make different wagers on the same odds and depending on how you phrase the wager, stuff like that. So, yeah, I just love anything like that…the other day I was looking at the statistical proof of the “friend zone”, some analysis looking at what people bought when going out for a drink with a friend, whether it was a date or not a date, whether it’s a soft drink or an alcoholic drink whether they're a friend or a friend-of-a-friend, how much that increases the propensity to think it's a date, or not a date. It’s fascinating, but at the end it's all about human decisions.
If you want a really interesting book on all of that, I don't know if you've read a book called Dataclysm by a guy name Christian Rudder who runs OK Cupid website. He founded it. Absolutely brilliant because he's -- because he's founded and run the company, he has the keys to the kingdom, as it were. So he spent five years writing this book. And it's about what we do when no one's looking. And so it's, obviously, about how people really make decisions about who they date….and most interesting the differences between males and females (men don’t fair that well!!)…but then you bring in stuff on how our brains evolved (and where and why) and it’s all gets even more interesting!
Tracy: That's super interesting. Okay. What have we got next? Oh, what's your favorite feature of Tableau?
Simon: I think my favorite thing about Tableau is its consistency and it's determination on the part of the Tableau developers, not to break those rules no matter whatever happens and keep that core set of rules the whole way through. That's my favorite thing about Tableau. My favorite feature, I've got to say, are action filters.
Tracy: I love action filters, too. I remember the first time I learned how to use them when I first started at Tableau, and I was, like, oh, my gosh, this is fantastic.
Simon: And they're so cleverly designed. I think people underestimate how flexible they are. And, again, with a simple set of rules -- I've been building on my iPad dashboard I've been building I've got like four tiles that look at last week, last four weeks, last 12 weeks, last 52 weeks.
And by clicking on each of them, there's a chart below showing you the trends, and that chart is basically a sheet swap. So if you click on the one week, you get the one-week chart. You click on the four week, you get the four-week chart. And the way you can send actions and exclude and include, the way you can play those against one another, gives you total flexibility with one or two very simple sets of rules. So action filters are probably my favorite Tableau feature I'd say.
Tracy: Yeah. That is a good one. And it's one of those ones that's been around for a while. Okay, what about on the flip side of that, what would you say your biggest feature request would be?
Simon: I am caught between two here. One would be a nicer transition of vizzes. You know?
Tracy: Nicer transition?
Simon: Yeah, so when you hit an action (or when the dashboard initially loads) the way that the viz rebuilds itself. I look at the stuff our d3 programmer creates (the free SKUtrak dashboard is built of d3), and I see the beautiful way the vizzes build from the axis (for example)
But I wouldn't want to sacrifice performance (as it’s a nice to have) but it does give that real polished feel, especially for clients who have seen d3 (or similar) content…however what they don’t know is what he can build in a week, takes me 10 mins in Tableau! So that would be one. Dynamic parameters, I think. Like everybody.
Tracy: Oh, yeah. Which they're working on. They're taking it down step by step.
Simon: And I fully get that, and trust the final solution will be awesome…take LOD calculations; it would have been very easy to create an interim (hacky) LOD feature earlier than they did. But to take it away, and investigate what the actual problem is and come back with such a well thought out solution. The depth is almost infinite. And it's just such a clever solution.
Tracy: Right. You said it. Okay. How did you discover the Tableau community?
Simon: I had a problem which I didn't know how to solve initially, which was how to shade weekends. I remember quite clearly now. Yeah, it was a shades weekend problem. I asked the community, and Joe Mako, came back to me and offered a screen share with me, which was absolutely fantastic because he was, you know, Joe Mako. In the end, the information was so good, it didn't require it, but the offer was there. It was just brilliant.
Because he had been so helpful and, obviously, there were others, whenever I typed in a question I had on anything Tableau, the links that came from Google were always the community. So there were a couple of easy questions that I sort of answered and just kind of got going that way really and found out that I was learning a lot -- a lot of things were questions that I answered, I found a technique for doing something I didn't know before. And as I started getting up the leader board, that kind of incentivized me and my boss was really encouraging . So it just sort of built from there. He gave me the time to do it as well. He said, you know, you can spend half a day or whatever a week to do this because I can see the benefit of it.
Tracy: Yeah, it's great. You sound like you have an awesome boss.
Tracy: That's really cool. What's your favorite part about the Tableau community?
Simon: I think it's the really supportive nature. There is virtually no vitriol or any of that at all. It's such a friendly place and everyone is so supportive.
Even if an attempted answer is wrong…I like being wrong because it means that once I know that I'm wrong, I've learned something I didn't know before (so I’ve learned something). A quote I like is “I like to be wrong for as short a time as possible” and even if someone has written something incorrectly, the discussion around why it's wrong and the learning is a really useful part of that whole discussion. This is done in a supportive way (other forums you don’t even need to be wrong to get a barrage of abuse!). So in short…everyone helps each other learn. They don't berate one another for not knowing something. I think those are my sort of my key things of why I love the community.
Tracy: Awesome. What advice do you have for either new members of Tableau and/or of the Tableau community?
Simon: The new members -- so new users of Tableau, I would absolutely start with the videos. You need to have a base, you know, you need to know where the “connect to data” button is.
But then it is -- it is to not try and learn it in a parrot fashion “how I do jump plots”, “how do I jitter scatter plots”…but actually get the understanding of it. You’ll learn more, it will be more fun, and you’ll come up with your own tricks!
My advice…if you think you know the answer to a question, have a go at it because no one will berate you for not having the correct answer. And, often, like all these things, there are multiple correct answers. You know, if it works, it's the correct answer. And so quite often there are threads where people will offer three or four different alternatives. They're all correct. One might be more efficient than the other. So my advice would be don't be scared and get involved.
Tracy: Great. I love it. Is there anything else that you want to share about yourself or that you want to make sure the community knows?
Simon: I'm vegan and I love cats, I've got four rescue cats.
Tracy: You have four cats!? Oh, my gosh.
Simon: They're living in my two-bedroom flat in London, so yeah it’s a pretty furry place to live! My avatar is a picture of one of them with me.
Simon's Profile: Simon Runc
Follow him on Twitter: @Runski
See his Tableau Public Profile: Tableau Public