You probably recognize Mark Fraser by the yellow Mr. Happy face on his avatar and his willingness to always at least try to get some sort of response to everyone's answers. Well, I had the opportunity to have a conversation with Mark last week, and honestly, I have no idea how he has enough time to even participate on the community, let alone contribute as frequently as he does! Mark is in an industry that I don't know much about, and has been doing some amazing stuff. Take a look into his story, where he gets inspiration, and how he uses Tableau. (And yes, I finally got him to send a real picture of himself!)
Tracy: Okay. So you're in the film industry.
Mark: I guess so, our company makes software for film distributors.
Tracy: So what is your role specifically, then?
Mark: I've got quite a mixed role. I started at MACCS and was very much the BI guy. When I was brought in, they were originally looking at SAS -- visual analytics. And they'd actually done some preliminary meetings before I started. Actually, before I even started, I went to SAS for a day to actually review it on their behalf, which was quite strange.
And that’s when I said, “okay, you've come so far with SAS, but have you checked out, the rest of the market?” I have to admit I wasn't particularly impressed with the market research they'd done in the BI space.
So I wanted to go back and do it properly, look at it in terms of our clients’ needs, but also in terms of hardware & software integration, we need to consider the whole stack because we are a software supplier ourselves. How are we going to integrate this with our existing products?
So long story short, I embarked on a period of software research. We looked at Power BI, the Office 365 suite, SAS Visual Analytics, I had previous experience with SAP Business Objects and Tableau. We were looking across the range - low cost, high cost, something in the middle, market leader, et cetera. Trying to tick all those boxes.
And I have to say, we went for Tableau, obviously, but we never looked back. I had a colleague with me this afternoon over from our Los Angeles sales office. It was the first time he'd seen any of our BI stuff. And I think I heard the word "awesome" about a hundred times in the last hour. I didn't realize Americans said "awesome" quite that much… but then coming from London I have my own quirky language!
Tracy: Haha, that's fantastic.
Mark: But, you know, he's working with our clients, people like Warner Bros., for example, and he's starting to understand what we can offer them in terms of understanding their market better. Particularly as you've got digital cinema coming in, you know, all the various things with piracy and DVDs and Netflix and it's an increasingly difficult space for the distributors and the cinema.
So with what they've got, they need to work smarter. My hope is that we can give them the intelligence they need in order to, effectively, work smarter within their space.
For example, say there are less admissions year over year. It's about maximizing admissions by better targeting consumers by offering titles their interested in. One of the things I'm doing with Tableau is, on a local level, to help them understand how films are performing. At the moment we're quite European focused, but as we move forward, also the U.S. It's really quite interesting stuff.
One thing is when you look at, say, titles, it's amazing, particularly in America, something like a teen comedy or something of that type will not play out well in the Bible Belt areas. The way a film plays across the different states is quite interesting in terms of the way the genres perform. And it's for me to help them better understand and help them market the film in terms of deciding where to even play some of these films. And we're using, Tableau to help decide where a film should be played, ultimately.
Tracy: So are you looking at, historical data or a little bit of everything in terms of how to make that decision?
Mark: It's a mix. It starts off at the beginning looking at historical data and you look for a comparative title. For example, say you're releasing something like Star Wars. They're called "tent pole" movies. So you want some equivalent in the sci-fi space, but also something of an equivalent marketing and production budget.
You’re initially looking at how those perform across areas, looking at demographics, for example, how they play against men, women, ages, et cetera. So the historic data will determine the initial play run, and where the film is initially released.
What then happens (films work in a weekly way) is they'll release it in America on a Friday, and what they'll do is they'll play over that initial weekend. They'll then take a first cut of the data on a Monday morning, and that they'll use that for planning for the following week.
Then, they'll decide how a film is doing. Does it need to be pushed out to more cinemas? Does it need to be restricted? Do we want to change what screens we're showing it on? Where we are showing it?
They'll use an initial first look at the data to make some initial sort of decisions. And then throughout the duration of the film, that information is ingested each day, each week, and the picture, obviously, changes. And they'll change the scope and duration of when it's playing accordingly.
And as I say, this is happening across multiple titles all at the same time. This isn't just a one-off exercise. This is happening continuously looking both backwards and forward. The more data they have, the better the decisions they can make, particularly around demographics, genres, ticket information, things like that.
We're also seeing it in terms of the cinemas running things like more male-orientated or female-orientated evenings. In the near future, they're running Bridget Jones Nights, for example. So they'll just pack all the screens, showing the same film and it will just be female attendants or equivalent on the male side.
Tracy: Oh, yeah. So then does it also help determine at what point the movie or the film is released, say, to Amazon or Netflix or something like that?
Mark: It will depend, again, on the title. So big tent-pole titles, will normally have a fairly pre-defined route to Netflix and DVD. So these will be like Star Wars and things. Whereas, the smaller ones, not only will they bring it onto the Netflix earlier, they may, as I say, change the scope of the play run entirely. They may pull from theatres completely and actually put it six months later, for example, that's not uncommon where a film is expected to perform a lot better and doesn't, it could just get pulled entirely and shifted. So it's a real moving picture, but it's the data that's determining these decisions.
Tracy: Yeah. That's so cool. So this what you do every day?
Mark: A little bit, yes. We are a software company, so we're providing solutions to our clients. So we provide the framework for them to do these things. So we're not actually doing the analysis.
That's a distinction. There's a certain crossover. I cross most of the Ts and -- dot the Is, et cetera. But what I can't do is the analysis for them, they have to do that bit for themselves. They have to make those decisions. But what I can do is set up a reporting framework for them to hopefully make the right decisions.
One big benefit of using Tableau is for us to be able to take information from completely different data sources and bring those together to create a more complete picture. By that I mean, we may be taking rating information from IMDB or Rotten Tomatoes, we can take reviews from Twitter, using scrapers, for example, but also the box office information both domestic & international from places like Mojo and BoxOffice.com, et cetera.
What Tableau allows us to do is to be able to take these seemingly distinct pieces of information, but actually to glue them together and to use them to create a much clearer picture rather than being a bunch of separate items.
Tracy: This is so interesting! So do you consider yourself a bit of movie buff?
Mark: No, no. Embarrassingly, I haven't actually been to a cinema in the time I've worked for the company!
Tracy: You haven't gone to the movies once?! How long have you been working there now?
Mark: Three years. The last one was Skyfall. My girlfriend used to work for Transport for London (TFL), they manage the Underground in London and Skyfall had permission for special filming on the Underground. And as a thank you, they gave TFL a special preview. So we went to it, it was cool, good movie! And that's the last time. It was at Leicester Square, you know, it was all pretty cool. But one thing I was saying is my colleague actually is -- I'd like to see our company giving employees some cinema tickets each month and for them to go and write a film review for our website.
Tracy: Oh yeah, that would be cool. And did you say you had used Tableau before this company?
Mark: Yes. Yes. So in my old life, I used to work in London for the NHS, which is the National Health Service. I was a data analyst.
The NHS is very cost pressured. There's simply not enough money, so cost is a real driver and decision. So we used Microsoft for everything. The stuff I've done in Access, my gosh, I've made Access jump through hoops along with Excel. I had automated macros - you'd start it in a batch file like you would in Tableau Server, it would start Access, run some stuff, it would pass that to Excel, run some stuff in Excel, e-mail it. And all that from a batch file.
You end up having to do that with the tools you're given. But then, we started using a demo version of Tableau to look at accident & emergency data and really try and understand what was happening.
By that I mean -- who's coming in, what ethnicity, what problems, where are they coming from? And just starting to just really understand and from those initial insights, it became obvious to management, as it were, that we could expand this across other functions across the hospital and certainly benefit from the increased insight that you could get.
Unfortunately, I happened to leave at the point where I'd done the research and I'd recommended Tableau and then I left. And it was just at the point where they were purchasing Tableau and I happened to leave. I was leaving the country. It wasn't an anti-workplace thing, I was moving country.
But my old boss, he came in as the information manager. He had a lot of experience with R -- he's very statistical, math minded, as it were. And he came in -- he was the one initially that raised the idea of Tableau. And then we all took it away, different products to go and research. And I happened to pick Tableau to do. So that was my first.
My first thing was just to attach it to an Access database, actually. What I was trying to do is just create a very basic -- well, a very complex, in Excel, pivot table and some various things. I connected it to Access, brought it through. Literally in under five minutes I had what had taken weeks in Excel to do and it was a lot more useful. Within that five minutes, I knew that that was the future that I would personally end up spending more of my life with it than I'd planned.
Tracy: Fantastic. I don't know how you think that your story is not interesting. I'm fascinated right now!
Mark: Yeah. Well, one thing that I do enjoy is seeing someone's reaction, particularly when someone first sees Tableau in action. I just had that today with my colleague who is in town.
And we were just talking, and it seemed like he knew a bit about the space, et cetera. And then as I sat him down, I just had my first chart, which is purely just top grossing movie titles with a relative date filter, just shows admissions, box office, number of free tickets, and the average ticket price. So just quite basic. It’s just some stacked bars in Tableau.
And from that point, you could tell straight away he was like, “Wow, just wow.”
I showed him the data at show level and then I take it all the way up to country and past country. He was amazed that we can literally go into Star Wars, for example, and then we can take a day, a cinema, a particular screening. He can go from seeing how Star Wars performed internationally to how it performed in a cinema in Michigan, for example. And he's just sitting there blown away.
To be able to offer people that kind of insight into their information is, you know, it is exciting.
And, you know, it's their reaction where you see the cogs tick, as it were, and they start saying, hang on, and you can see them still working through the benefits themselves and how they can leverage other data, And then just suddenly there are a ton of ideas, you know, within a couple minutes. Oh, actually, what about this? What about that? Hang on, maybe I've got that.
Then they’re telling you -- you’ve got to do this, you've got to do that. Just within a sheet or two, they were already coming up with ideas of their own.
The next question was, can I have a user name? Can I have a password? That's normally the start of -- the conversation, and then try and rein them back in a little bit.
But that's the thing, the product speaks for itself. And I've built it in a way that is about the end user to interact with it. It's built in that way. It's built so that you've got questions, you can find the answer. I'm trying to make ours a sort of Siri for our part of the movie business, as it were.
Tracy: Yeah. That's so cool. So what is the data source that you have, like, all of that information in?
Mark: Well, that's mainly our own product that we're actually selling. Or, as I say, it's actually information from other sources such as social media or international box office information.
Tracy: And then do you put it into a database or how do you connect to it in Tableau?
Mark: Well, it's a mix. Yeah. Some of it we write -- our own stuff, we write to our own databases. Some of the others -- I can use the Web scrapers, for example, and the Web connectors. And that's an area I still need to really explore fully, but yeah, we need to take the data and then store it locally or retrieve it from an API, for example, on the website or as I say scrape it.
One thing that was interesting -- I haven't yet looked at it in version 10, Tableau can connect to JSON files. That potentially is very interesting for us.
Lots of the APIs are surfaced in a JSON format. That was a really good move by Tableau.
The JSON format, effectively, is like question and answer. So you just simply have a question -- might be title, and then answer is the actual title. And so you effectively build a series of tags. You simply just loop through the tags and just pick out what you want.
And that takes me to my third hat that I wear at work, which is I'm managing our mobile applications. We make a mobile app for the Google and Apple store. So I'm managing that. I'm quite busy.
Tracy: Yeah, just a little bit sounds like!
Mark: Yeah, that's quite different because that's all written in C#.
Tracy: So are you building the app yourself or you're just project managing it?
Mark: Yeah, I dip in and out, but I'm doing all the project management and I'm the marketer, I come up with the name, the icon, everything else. So it's been interesting. App development is something entirely new for me. Actually seeing the back end of some of Apple and Google's processes is quite interesting.
But bringing it round to Tableau, it's one thing I'd like to do now is connect Tableau to our app usage data. What I actually want to do is I want to use Tableau so I can start to understand how our downloads are doing, and how are end users are using the app. Because it's interesting for us to know.
Tracy: Yeah. Because you want to know if you're spending your time wisely and building it for the right audience.
Mark: Right, if people are happy with the app, then we can see how often they're using it, for example, and also we get some information in terms of if they download it in one to three days, are they keeping it or not? One to seven days? So this lets us know about retention. I know what it’s like, I can be like a butterfly in the app store -- Oh, I'll download this, oh, I'll download that.
So it's whether you actually end up staying or not, which is certainly interesting. But, yeah, as I say, it's an entirely new space for me. But, yeah, quite interesting. And as I say, Tableau still has a role to play in that.
Even though it seems like there's nothing particularly relevant, it certainly is and certainly will be helpful for me going forward. And, you know, as the only BI person in my company, I have to be the whole stack. So I have to be ‘Toby’ (Toby Erkson) and I have to build the server stuff, I have to do everything else as well. So it's been quite a steep learning curve. And particularly the forums help me learn Tableau Server.
Tracy: Yeah, that's interesting. I didn't really realize you were a server guy.
Mark: Yeah, I have to do both. I'm the whole stack, as it were.
Tracy: So you really do have your hands full!
Mark: I really do! -- going back to my example earlier with the automation of stuff, I taught myself all the tab command stuff because I'm used to that level of automation, I need my time back for other tasks! That's one thing I do look out for on the forum, you'll probably find a couple examples of where I really take out literally line by line and I put with arrows, right, this is doing this, this is doing this, this is doing this, and the order matters.
So I'm trying to educate people not only the person who actually asks the question, but anyone who happens to stumble on the thread because I know -- because I taught myself only recently, and if I can move someone up a few notches quicker, all the better. Maybe they can teach me something in return.
Tracy: Totally. That's awesome. So, one, what's your favorite feature of Tableau?
Mark: I think the dashboard actions. And by that, I mean the ability to take, as I say, the seemingly different data sources and create a relationship. Then see them interact with each other and see what that means in terms of the relationship.
Not only is it a Tableau feature, but it's something, you know, that the end user can actually use. Do you know what I mean by that? Not only is it a great Tableau feature, but it's great for end users. You know, pulling up my dashboard today and just clicking on some film titles saying, well, look how it performs across the country and then being able to reverse that.
One of my favorite things I’ve done with actions is that we were looking at a map of Holland. And there were three really big cities. So when we were looking at a map of the box offices, they were just three big blobs. I excluded those because what you'd then see is what I call the tertiary -- the secondary and the tertiary cities all suddenly pop out that you'd never know were there. And then you can start to look at the dashboard in an entirely different way. Once you start to peel off that layer of information, that's when people really start to understand how the data can help them.
Tracy: Yeah. It's one of my favorite features, too, because, as you said, it peels back the layers and just makes you ask more questions.
Mark: Exactly. I mean, there are loads of great features and they all play a part. But I think that actions in particular are just a real crux because that's where for me the developer and end user come together at the closest point.
It sounds a bit odd-- but, you know, the developer's really trying to focus the end user and the end user can then really start to really dig into their data rather than just have the usual canned reports.
That's the thing from the development side that’s interesting. To build and start to think about that, but then to leave it and watch the end user actually interact with it, and see how they start to draw meaningful answers to their own questions.
The dashboard largely is built to answer the questions. And that dashboard action gives them the answers that they need to make business decisions. That's why I think it's so key.
Tracy: Yeah. I completely agree! Okay, what about the flip side? What would be your biggest feature request?
Mark: Apart from the connection to Apple's stats. I have to confess, hand on heart, my issue with Tableau is their licensing. I don't know if that's a popular topic or not, but a company of our size, to be able to take Tableau to where we want it to be, it's a big jump.
The licensing, we find, is for the smaller end of market, just a little too restrictive.
Tracy: Yeah, that's frustrating because you just -- as a company, you want to be successful using the tools that help you, but you don't need a huge implementation. And so there's got to be a middle ground.
I hear your frustrations. I will continue to pass this feedback up the chain because I know you're definitely not the only one in this situation.
Mark: Yes, please help me, help me!
Tracy: I wish I had more control over that situation.
Mark: It's not something I actively raise in particular on the forum. I'm quite conscious of being overtly negative or anything like that, you know, that's not our purpose. You know, as ambassadors, I think we have a role to set but in a conversation with yourself when you ask the question I have to give you as honest answer I can.
Tracy: No. I appreciate your honesty. I do. So thank you for being honest. Okay. How did you discover the Tableau community?
Mark: Well, via the website I think initially. I was a member originally when I first looked at Tableau at St. George's in London. And then coming back to Tableau and being the only BI person there, the forum was a natural go-to place. The forum was also one of my key decisions for choosing Tableau because I'm conscious that it's just me. And I can only self generate so much creativity and inspiration and clever formulas and stuff. Having such a strong community was what also attracted me. The other side was Tableau Public, it provides me a lot of inspiration – not only the industry related stuff but the wealth of amazing vizzes.
It was a big part of our buying decision.
Tracy: That's awesome. What is your favorite part of the community?
Mark: Most often, I'm likely to go to unanswered questions, to be honest. Ideally, I don't want people to leave without some kind of answer. The tradeoff with that, I worry sometimes, that is I spread myself far too thin, and then I don't finish helping people. But, at least, I hope they have had some interaction, from a friendly face!
Tracy: Yeah, totally. I mean, I think people are grateful for getting any response sometimes.
Mark: And that's in part, you know, being on both sides -- particularly someone who answers a lot, I know what I'm looking for to help me answer. So what I try and do, and you'll see it probably quite a lot, is that you'll see a question and I effectively just ask them a series of follow-up questions to really set it up so that either I or someone else can then -- it's that first reply or their second post that actually contains all the useful stuff that someone needs.
And maybe that would be an example or some context or just some -- quite often, I'm just literally just trying to tease out the details that either I or someone to actually answer the question will need. And so you'll see that quite a lot.
Sometimes, you know, I'll try and finish it off if I can, otherwise, someone might well jump in, good luck to them. But I at least I like to think that I know what I would need and what others would need in terms of trying to answer that. Could you clarify what you mean by this or you know? How does this work in relation to this? Or, I don't quite understand what you've explained here.
At least if I can get a question out to that stage, and then someone else can answer, it may not be me that actually gets the green tick, but at least I've helped get that answer and that's enough for me.
Tracy: Right. You've started moving it on its way into the right direction.
Mark: Exactly! And also just being that first responder friendly face, as it were.
Sometimes you can answer it in one go, but other times it's just teasing out some more details. Yeah, someone can deal with it. Sometimes it just, you know, it just takes a completely different turn and, you know, turns into something completely different. In which case, I run away. (Laughter.)
Tracy: Haha. Let someone else deal with it.
Mark: Normally let Simon (Simon Runc) take a look at it.
Tracy: Yeah. I know Simon will take a stab at anything.
Mark: Yeah, to his credit, and he's quite mad. At the point where I get stuck, Simon just carries on. He's like the Duracell bunny.
I've run out of battery and he just carries on. As I said to someone the other day, when I grow up, I want to be a bit like Simon. And when I grow up, if I can be just a little bit more like him, I mean, and that goes for a lot of people in this community, if I can just take a little bit of what they've got, I think I'll do very well for myself.
Tracy: Yeah. We have some smart people in this community, including yourself.
Mark: I do all right, but yeah.
Tracy: Yeah, you do more than all right, I think. So what advice do you have for new members of Tableau or of the Tableau community or both?
Mark: Enjoy it.
Tracy: I like that.
Mark: Yeah. I can advise you read this, or watch that video, but just take it in and have fun with it, you know? And that should hopefully, naturally, spark the ones to then go and find out more. But I'd say initially just have some fun with it. That's the beauty of the tool, it can be really business focused, really hardcore, boring, dry stuff. Or it can be about something that interests you.
And to some degree, if it takes someone to do something a bit more fun and a bit less businessy, but it actually sparks their interest, all the better I think.
Not many people can get really into designing a new profit and loss spreadsheet. Do something with your favorite football team or your favorite film or whatever it may be. You know, just something that really interests you, then naturally everything then follows after that.
Tracy: I like that answer. That's different than everybody else's. It's so simple and succinct.
Mark's Profile: Mark Fraser