Fi (00:09)
Hi everyone and welcome to
the podcast that's unscripted, uncensored and undeniably data. I'm Fi
with me is my co-host Sarah.
Sarah (00:19)
Hi everyone, today we've got someone who's not just a master storyteller, but a true pioneer in the data world.
Fi (00:27)
Absolutely. We're thrilled to have Andy Cotgreave on the show. So if you're into DataViz, you must already know his name. Andy is the senior Data Evangelist at Tableau and one of the most influential voices in data storytelling today.
Andy Cotgreave (00:43)
Wow, that's amazing. Yeah, well, it's fantastic to be here. Thank you. I feel under incredible pressure now to live up to that intro. Also, incredible pressure to swear. You said unfiltered. we'll see if there are F-bombs around. I don't know. It depends how passionate I get. Let's get to that.
Fi (00:55)
Yeah.
Let's see where we can take you.
Andy Cotgreave (01:04)
It's also 7.30 in the morning for me and I don't know if I'm more or less sweary before the coffee has kicked in. Let's find out.
Sarah (01:09)
We'll see as we progress
through the podcast and through your coffee Andy.
Andy Cotgreave (01:15)
Absolutely. I've got a whole pot so I'm ready for anything.
Sarah (01:19)
Nice,
nice. Now I first met Andy in Singapore quite a few years ago when he was doing a little evangelist Tableau tour around the globe. and I was really excited when he presented Iraq's Bloody Toll, which was, is actually today, one of my favorite vizzes showing what you can do with title, color and orientation. But I feel that is really the tip of the iceberg.
because Andy also does so many other things. So many other things.
Andy is also the co-host of the popular podcast Chart Chat, where he dives deep into the art and science of data visualization. And if that wasn't enough, he did co-author The Big Book of Dashboards, a must read for anyone serious. And by the way, I've heard he's got another book coming.
Andy Cotgreave (02:16)
There is, it's gonna be the same size as this, but it's gonna be Dashboards That Deliver. So basically this book was, here's a bunch of dashboards. And then a lot of people said, well, what's the process? How do you build these bloody things? there you go. so we, Dashboards That Deliver not has 15 more real world dashboard scenarios, but a whole front section on the process of building them. And we were joined by Amanda McCulloch.
Sarah (02:21)
Nice.
Andy Cotgreave (02:43)
of the Data Viz Society as the new addition to the authoring team. And can't wait. It's coming out in September 2025.
Sarah (02:50)
Awesome. Looking forward to getting my copy.
Andy Cotgreave (02:54)
Yeah. So I want to talk about Iraq's Bloody Tile, right? So I first presented that chart, right, in about 2012. The original chart from Simon Scarr, so just Iraq's Bloody Tile. That came out in 2012 and I had this little idea, Flip It, to make this powerful point about storytelling. And I think I've used it in 99 % of presentations since. And I reckon if you, it's been cited in...
Fi (02:56)
And yeah.
Sarah (03:05)
Yep.
Hahaha
Andy Cotgreave (03:21)
loads and loads of academic papers and it's in about eight or nine books maybe loads of books on data viz and so obviously I'm really proud that what did i do i had this idea to flip a chart upside down and replace colour in powerpoint and yeah that's so if you distilled yeah yeah i'll the title and yeah so if you distilled my career down to two slides it's those two and that's the secret just just that was it
Sarah (03:35)
And the title, don't forget the title.
Fi (03:47)
beautiful. I love it. I love it that
that you managed to do this without Chat GPT.
Andy Cotgreave (03:54)
Well, of course, right? Yeah, Fi thanks. That makes me feel great. But I mean, I have sought a newer chart that enables you to make that flip so quickly and effectively as that has been the Iraq chart, right? You know, there are loads of ways you can make that point with other visuals, but...
Fi (03:56)
Right.
Andy Cotgreave (04:19)
none is just so punchy. So anyway, I'm very grateful to Simon Scar for doing the original chart.
Fi (04:22)
Mm, mm.
I love it. Well, what a great anecdote, which leads me onto my next point where you've been helping people see and understand data for, over two decades
And whilst you've been speaking around on all those stages around the world and advising Fortune 500 companies, your blog, Gravy Anecdote, is actually legendary in our community. And a fun fact, if people don't actually know or don't know you well, Gravy Anecdote is actually an anagram of your name.
Andy Cotgreave (04:38)
Bring it on.
Sarah (05:00)
I didn't know No, I didn't. There we go. I've learned something already.
Andy Cotgreave (05:00)
It is indeed. I mean, you didn't know that, Sarah. So,
yeah. Andy Cotgreave is also an anagram of converted gay, which I didn't think would quite hit for SEO purposes, the audience I wanted. And Andrew Cotgreave is an anagram of waved erect organ. Again.
Probably not good for a data storytelling blog, so I went with Gravy Anecdote. There you go.
Fi (05:30)
really suits our explicit podcast.
Andy Cotgreave (05:33)
Yeah, I realised
I'm going not safe for work straight away.
Fi (05:38)
I kind of did some research for you on what other anecdotes you could use but they didn't quite go in that direction. I got Voyaged Trance or Voyaged Nectar and for me, mine was really average. A Drongo, if on, so maybe not so great for hosting a podcast but I have to tell you, I loved Sarah's Urban Threats. It sounds super exciting.
Andy Cotgreave (05:41)
Bye.
Mm-hmm.
Right.
Sarah (06:03)
yeah.
Andy Cotgreave (06:04)
nice
yeah Urban Threats that's awesome
Sarah (06:07)
Yeah.
Fi (06:08)
Yeah,
Sarah (06:10)
Yeah, I feel, I feel...
Fi (06:10)
it's got Sarah's name. sorry.
Sarah (06:13)
feel like I could go into a whole other career with it.
Andy Cotgreave (06:15)
Yeah.
Fi (06:33)
No script, no script.
Sarah (06:35)
That's good.
Fi (06:38)
Maybe a few notes there, obviously, with some research, but I do think that Sarah's kind of suits her being the Urban Threat.
Andy Cotgreave (06:46)
Urban Threat. I love that.
Sarah (06:49)
Okay, Fi, I think that's enough of a side quest. Let's get back to the podcast. What I love most about Andy is his no bullshit approach to data. He's all about making data relatable, impactful, and let's face it, if you've seen him at Iron Viz, most of his outfits steal the entire show.
Andy Cotgreave (07:09)
Well, that's very kind. I'm professionally obliged to say it's our contestants that make the Iron Viz show. hosting Iron Viz since 2018 has been an incredible privilege and as is the fact that, know, little secret, we get a wardrobe budget. Each year I get to go in like...
Sarah (07:31)
And do you max
it every year or? Like, I'm just trying to think, like.
Andy Cotgreave (07:35)
I actually, I
ended up spending my own money on But the epitome, I mean, the shoe that you can see, those of you watching, the custom shoes we had last year were absolutely a top notch. Really enjoyed getting those. But yeah, Iron Viz since 2018. And yeah, it is, it it is.
Fi (07:42)
Shoes.
Andy Cotgreave (07:57)
It's such a joy presenting it and the last two years in some ways have been the best because as I've got more and more comfortable presenting it, you know, I'm less, I mean, I'm still nervous, excited, but I'm less worried about things that can go wrong. And so now it's just a, it's a brilliant opportunity seeing the journey of the contestants as they arrive on site, introverted, terrified, wondering why the hell they entered this feeder competition.
Fi (08:25)
You
Andy Cotgreave (08:25)
thinking is my viz any good, right? And then throughout, because they're sort of backstage all the time of conference rehearsing and then just watching them change from nervous anticipation to shining on stage. It's just brilliant watching it, I love it.
Fi (08:42)
So Andy, one thing, if people don't know what Iron Viz is, can you just do a quick explanation?
Andy Cotgreave (08:47)
Yeah, Iron Viz is the world's biggest data viz competition. That's certainly what the Tableau marketing team tells us. I would say, I think I'd probably agree it is. What is it? So we, every year we ask Tableau users, Tableau community, the data fam to enter a feeder competition with a theme. The top three as voted by community members win the privilege
of coming to conference to hope to be take part in the Iron Viz grand final. They are given a data set. They are given 20 minutes to build a visualisation live and then three minutes to tell a story about that visualisation. They get judged on analysis, design and storytelling and the winners, I mean, now receive a hefty cash prize. And it's an absolute party at Tableau Conference. It's grown from a little crazy little breakout
room idea in 2012-2013 into one of the flagship keynotes of the event. even though we crowned one Iron Viz winner, we're getting 200, 300 entries every year and huge community growth as a result. So it's just an amazing platform for growing the field of data storytelling. I love it.
Sarah (10:08)
Yeah.
Fi (10:09)
Throwing this one in there, have you got a favorite Iron Viz winner of all time?
Andy Cotgreave (10:11)
Hmm
Okay, so the first thing that comes to mind there is actually Lindsay Poulter 2019, 2019 Iron Viz. So in that, that was the year there was a tie and she was not in the tie. So in fact, we had two winners and Lindsay didn't win. What I remember, she'd done a very crisp sort of a business style dashboard, but with absolutely gorgeous formatting.
Sarah (10:23)
Yeah.
Andy Cotgreave (10:44)
And what I remember is somebody just walked on stage. They're not allowed on stage, but somebody just walked on stage and offered her a job. But as soon as the lights came down, it's like, come and work with us. And I think that's one of the huge benefits of IronViz. Maybe for any employer whose employee gets on the IronViz finalist stage, you had better be giving them a good reward or a good incentive to stay, because
being on that stage is fantastic resume, CV thing. So yeah, I'm going to go with Lindsay.
Fi (11:17)
think JLL,
my former employer, has had five people in the finals. Yeah, totally stacked.
Andy Cotgreave (11:23)
Yeah, well Chris was
the first winner last year and I mean Chris was great last year because he... the story part of Iron Viz is so important. You know, can build something mind-blowing but you have three minutes to tell a story. I mean, you know, this is fundamental to that storytelling. It's like you can build the most effective beautiful thing but as a human being bringing it to life through the drama and the story arc can transform it.
Chris just had a really well-timed, humorous, insightful story, which I thought was...
Fi (11:58)
And then Hesham was also in our team at the time, although he was working for The Information Lab, but then was also at JLL at the time. So we tend to claim that one as well. Or they tend to claim it. you go. Slip of the tongue in terms of former employer.
Andy Cotgreave (12:12)
Yeah.
You claim whatever you need. Yeah, Hesham did that. That was a really difficult data set. my God. We could talk for hours about the Iron Viz data sets, that was a data set without a timeline, which makes it really hard to find and tell a data story. they managed. Yeah.
Fi (12:34)
Mmm.
All right, let's wind this up about Iron Viz for now then and
move on to the main topic of the day. So Andy, for those playing along at home, what does data evangelism mean to you?
Andy Cotgreave (12:49)
What does data...? Right. I'm senior data evangelist at Tableau. So I'm going to talk about, I'll answer that first specifically from the perspective of working at a data analytics vendor. So an evangelist in tech industry parlance is for me, somebody who is passionate and authentic.
and an expert in a field, right? So I'm extremely passionate. If you've ever seen me speak, you'll know that's true. As you very kindly said, I'm no bullshit. yeah, I am no bullshit. I will call the bad things out about the company I work for. I will call the good things out about competitors. I think that builds a huge amount of trust in what I do. And leader in the field. mean, I find it...
I find it strange to be called a leader in the field, but I did write a book and I have presented for many, many years and I've got loads of followers, right? So I guess that the evidence I have to say is leadership. So they're the three criteria. And then by being a passionate, authentic leader, evangelists are creating content or speaking to others or talking to across the industry press or beyond about the field that they're in.
data visualization, analytics, storytelling with Tableau. And through that authentic leadership, we lead people to the platform we sell. So I am a data evangelist for Tableau, Primarily, I'm a data evangelist because I love data storytelling. I happen to also think Tableau is just light years better than every other platform on the market.
And so the perfect match for evangelism is when the evangelist person is just doing what they're doing because they're driven. And that is so true about me. I'm just like, I can't really help myself. And is doing that for the platform that they support. I'll stop there. And does that make sense? Because I could keep going. What do you think? Is that a good definition?
Sarah (15:07)
Yeah,
yeah, that's a great definition.
Ha ha.
earlier on you spoke about your love for storytelling. So can you tell us a little bit more about your journey with storytelling and maybe some of the common mistakes that people make when they think they're doing storytelling?
Andy Cotgreave (15:29)
So I'll tell you the first dashboard I ever wrote, I ever made sort of started me on this journey. And I actually wrote this, this forms half a chapter in the dashboards that deliver. Cause I was like, when did I actually build my first dashboard? about 20 years ago, 20, I was working in a small business research company and we went out and interviewed CEOs of fast growing private companies.
and we'd go along with somebody from a major consultancy to do an interview and essentially it was a way to build that connection. But we published league tables. The boss, he was something of a sociopath. with regularly, I mean, I've never worked at a company of 40 people, well, it a company of 40 people, the turnover rate was about 50%, right? This guy would reduce people to tears by just walking into one of the rooms and just tirading by.
piling on somebody with absolute anger. And often it would be like, what's going on with your project? And I'm like, well, how do I defend myself against this? Yeah, how do I defend myself against this? So each individual project was run off this enormous bloody spreadsheet. And I was like, you know what? He always asks the same fricking question. So what I can do, I'll just put some charts on a piece of paper so I can just go, Here, have this.
Fi (16:54)
You
Andy Cotgreave (16:56)
Anyway, I did that. I just, so if you ever came to try and do it to me, I was just like, I have this, this is a summary, piss off, right, and go and leave me alone. And it was really, it worked, right? He was like, oh, right, I was one of his star employees. Anyway, so then I passed that idea around the other team. And then later it was like, oh, one page, a collection of charts that fit on a single page, turns out they're called dashboards. I was like, all right. So.
through there, I discovered that was Dashboards and then got the job at Oxford University and then discovered Tableau in November 2007, downloaded Tableau. So that began the path, right? Through that, I learned about the cognitive science of data visualization, learned about building Dashboards. And then storytelling specifically. I'll tell you that one of the...
One shining moment I'll remember. I was very much into storytelling already by this point. But in about 2016, I was sat in a Tableau on Tableau presentation being delivered by Elisa Fink, absolutely incredible CMO. She was the first CMO of Tableau, just inspiration. And I'm now going to criticize her, but through the utmost respect. mean, she is a brilliant individual.
Sarah (18:13)
Ha ha ha.
Andy Cotgreave (18:17)
But she was doing this Tableau on Tableau presentation. And she was just showing dashboard after dashboard. And she was standing in the corner, and the screen was over there. And she just go, as you can see, the data shows da-da-da. And then da-da-da. The data's doing this, and the data's doing that. And I'm just sitting there going, I have no fucking clue what you're talking about. I can't see anything. You're just standing. You've got some crappy, complicated dashboard. You're standing there. The slide's not telling me where to point. You're not even pointing.
Sarah (18:35)
You
Andy Cotgreave (18:46)
where I should be pointing. I was like, this is rubbish. I spent that whole session going, well, how do you tell a story in a presentation? Better. Anyway, so that became a presentation called Clear and Presentation Danger, which I will one day turn into a book about how to present data. But I...
I remember that as a really important moment, this epiphany of like, I have sat in meetings my entire career and every time somebody puts a chart on the screen and they go, as you can see, blah, blah, blah, I can't see what they're talking about. And if I can't, I bet nobody in the room can see. And it's like, well, hang on, come on, if we try to, if when we present information, we want to be, we don't want to look like chumps. And I realized every presenter looked like a chump in front of me.
So there you go. That was kind of that was one epiphany for the storytelling in presentations perspective. I'm not sure if I answered your question, but I hope that was fun.
Sarah (19:40)
Love it. Well, I think,
yeah, I mean, make sure your data tells a clear story, right? If it doesn't, you're going to lose your audience.
Andy Cotgreave (19:46)
Yeah, right.
Yeah, and the simplest things, you show a bar chart with five bars on it on a screen, highlight the one you want people to look at. It's that simple, right? mean, that's the simplest thing. If you just show a rainbow with all the colors, people are like, where should I look? Where should I look? And it's not hard. It's not hard to make that decision.
And I will challenge people. When people say, can't be bothered. I'm like, well, you can't be bothered to actually have your audience think you're an effective communicator. You can't be bothered to do them the service of letting them see the insight you're trying to share. You actually are so lazy, you're going to make them do the work. At which point, people are like, Andy, you get a bit cross. But at that point, I'm like, well,
Are you trying to be a professional or are you just trying to look like a chump? Wow, this coffee's getting, I'm off, right, we're going now.
Fi (20:46)
I love the passion
in there, Andy, and it leads me in a slightly different direction and something that's been top of mind for me recently, now having to present more in our client space and I still don't feel really confident or really like I'm nailing it, I've done many presentations before internally in corporate land.
Do you have any recommendations for where I or the listeners can go in terms of learning the tricks of the trade? Because I was watching, for instance, Heidi Kalbe, who presented at the Sydney Tableau User Group last week, I think it was, and she could zoom in on certain elements and come back out. Do you have somewhere that I can go and learn all about that or that you would recommend?
Andy Cotgreave (21:35)
Mm-hmm.
My god, you know, I...
shockingly, for the presentation side of things, I don't entirely know. I mean, so Cole Nussbaumer Knaflic's books, storytelling with data will give you a great amount of detail on crafting the chart for slides. So that's a place to go and learn about how to build the charts. If you Google me, I have some videos on how to present product, how to do product demos effectively, and also how to build charts.
Fi (22:03)
Mm-hmm.
Andy Cotgreave (22:16)
I've got to write the book on this, because I don't think there is something... Stephanie Evergreen's got a book on presenting data, but it's several years old, it could probably do with an update. I don't know, Fi. I'll go and write a book, yeah.
Fi (22:32)
I stumped you, I stumped
Or you could write a training program and we'll host it on our Dub Dub site and sell it. Because I think that it'd be a really good, it doesn't have to be a long course. I think that you could get through it relatively quickly, but it could be just one of those nice bite-sized chunks.
Andy Cotgreave (22:49)
Let's talk.
It's one of those things, I remember learning the cognitive science of data visualization, right? It's not actually very complicated, but nobody's taught, nobody tells you these things. And, you know, the way you can design a slide, the way you can use color, the way you should use titles, it's not complicated, but yeah.
It's just, it's something that should be learned. I mean, I was very lucky. I got taught presentation skills in my first graduation job, first job after graduation. And then Tableau invested in a huge amount into presentation skills right from day one. I got the opportunity to speak and then lots of training, but the presenting data bit, nobody really was writing about that. I have to fix that.
Let me know on LinkedIn if you
Sarah (23:40)
think you do.
Fi (23:40)
You have to fix that.
Andy Cotgreave (23:41)
would want that course
Sarah (23:43)
Yeah, yeah, I feel that as well.
Fi (23:45)
I want it. That should be enough.
Andy Cotgreave (23:46)
Alright, okay. So you think...
Sarah (23:47)
You
got two people.
Andy Cotgreave (23:48)
Right, that's enough. Yeah.
Fi (23:50)
You got two? I'll go out and I'll do a poll.
Andy Cotgreave (23:54)
Alright, yeah, let's do it, let's do it.
Fi (23:56)
Awesome.
Sarah (23:58)
Yeah, I'm just thinking as well that you must give a lot of feedback to people and what's some techniques that people can do when you're giving feedback on say data visualization.
Andy Cotgreave (24:16)
We were doing feedback for the people were getting ready for Tableau conference sessions and one of our colleagues was delivering this presentation I can't remember maybe it was about mapping and every slide was a bullet and I'm like and it was a 45 minute presentation and I was like What is he doing? What is he doing? It was like it and I was like it's like he's built a manual and he's just reading the manual
And at the end, I was so cross. I didn't give feedback in the most effective way. I was just like, that was absolute garbage. You've just read a thing out. And he was like, well, I've created a handout so people can use it. I was like, well, don't present the fucking handout. Build a presentation and then give the handout, right? But it was.
Sarah (25:02)
Hahaha
Andy Cotgreave (25:11)
I was so, it was delivered in a very, well not non-professional, was delivered in a very assertive, he was somewhat taken aback at the time. How do you give feedback?
Sarah (25:22)
I bet he's never done
Fi (25:22)
Did it.
Sarah (25:23)
it
Fi (25:23)
Did it.
Sarah (25:23)
before, I bet he's never done it again.
Andy Cotgreave (25:25)
No, I bet he's never done
it again. mean, he had to, he was not a bad presenter. He was a pretty good presenter, but he'd obviously, he'd just like, this is a really important topic. People need a manual. I'm like, they do need a manual. Just, they don't want to hear it in a actually one piece of advice, in a 45 minute presentation, people can remember three things, which is shocking, right? Maybe four or five, but.
Fi (25:39)
not in the presentation.
Andy Cotgreave (25:50)
I'm talking about a breakout style presentation rather than perhaps a client meeting presentation. So if you're informing people at conference or any thing, people can remember very few things. what are they gonna leave remembering? Start with that in mind, right? And then you can put little contextual details behind each of the three pillars, but you're gonna learn this, this, and this. And then,
What have you done? You've given them inspiration to want to know more. That's another goal. So they've learned three things and they want to know more. And if they want to know more, where do they go? So this is what I said to that person. well, you've written a manual, so you can at least have a link or a QR code on every single slide going, if you want to know more, go here. So I always recommend, your presentation is just a amuse-bouche, a teaser, a trailer.
Sarah (26:25)
Yeah.
Andy Cotgreave (26:42)
But if people want more, where do they go? So have supplementary materials linked and ready on your first slide, each agenda slide, and your last slide to generate some continuity. Particularly in an organization like yours, you're also evangelism is about selling and marketing, right? So that creates the funnel. If you're a member of the Data Fam trying to build your own brand,
It creates the funnel, it gets people to your site. But it also helps the audience learn more. That's in a breakout style presentation. In client presentations, so if you're pitching to clients, you're gonna be doing that probably in a lot shorter time, but I would still think you're gonna get people to remember three things, right? And at that level, it's like you've got a maximum of six slides.
Sarah (27:15)
Mm.
Andy Cotgreave (27:35)
Nine slides, three slides on each one.
Sarah (27:35)
Yeah, yeah, yeah.
And I think the key thing for them is what's in it for them as well when you're talking to clients, right? So you've really got to frame it of how you're going to make their life easier.
Andy Cotgreave (27:42)
Yes, yeah, yes.
Let me... There's a certain company, perhaps the one I work for, whose first slide in every keynote is a line going upwards, right? Do you know the chart I mean? And what does the line show in these keynotes? Do you know it?
Fi (27:55)
Yes!
Sarah (27:56)
Yes!
Fi (28:12)
no, I was thinking that it was
the forward-looking statement and I was just like, mm.
Andy Cotgreave (28:16)
well, it's okay. It's forward-looking statement. It's
Sarah (28:17)
Hahaha
Andy Cotgreave (28:18)
thank you. And then it's a line chart showing company revenue. Right. And I give this feedback back to leech it. I'm like, why are we standing in front of customers saying thank you for giving us all your money? You know, that line chart could be anything. It could be customer served. It could be answers, insights delivered. It could be, you know, sustainability, environmental guides. That slide says everything we're doing is for us.
Whereas I'm like, well, no, when you're presenting, when you're pitching to clients, everything you're doing is about how you're helping your customers. So yeah, that's, I'll probably get fired now, but anyway, that's a concrete example where I think it contradicts what you're trying to do when you're pitching to people to be involved.
Fi (28:53)
It's true.
I think that slide would be really great in the shareholder update.
Andy Cotgreave (29:09)
Yeah, absolutely. And that's not a keynote at major events. yeah, I mean, I've, yes, you're right.
Let's move on before I forget myself.
Fi (29:20)
Very interesting.
Well, this is maybe back into the Tableau space. Obviously, Tableau would be your tool of choice. As you said earlier, it kicks a few goals. But do you have any others that you use to supplement your analysis?
Andy Cotgreave (29:43)
I still use Excel. mean, before using Tableau, was presenting at Excel user groups and I was a VBA. I could do everything with VBA. every day, every day I was using the spreadsheet, my goal was to not use my mouse. I would measure my skill at Excel by the ability to not use a mouse. Cause it was, was like only through keyboard shortcut wizardry would you be a true guru, right?
Fi (30:12)
And in
Andy Cotgreave (30:13)
Yeah, and I would judge people harshly if they like put their hand on their mouse to do something. I'm like just use control or three two one Zed, right? So I used to use Excel all the time I mean I use Excel a lot less now because I can do most of the things in Tableau But clearly I can't do a lot of the job or a lot of data analysis without using Excel
LLMs, Gen.AI are creeping in for idea generation and well largely for idea generation and just talking around getting ideas. Yeah, we can talk about AI a lot more, that definitely being, I mean, I find them incredibly useful for analytical help, but they are also incredibly dangerous and incredibly
shit, right? So the chance of the making a heinous, like dangerous error is so high. I've not done anything close to production. I'm just like, I'm playing a Tableau or I'm goofing around. Give me an idea. What do you think? Right? And then I go and build myself my own visualizations.
Tableau Tim did a video a few weeks ago and in it he said, you know, we've spent years as analytics teams building trust with our organization. You know, one hallucination, one insight saying sales are going up when they're actually going down. One error like that, it's all gone. And the LLMs are really good at making those mistakes and they're really good at... absolutely. it's just my, you know, all right, let me...
Sarah (31:50)
Yeah. And they say it with conviction, right? It's like, yes!
Fi (31:54)
I've got one!
Andy Cotgreave (31:59)
Have you ever been in the situation or talking to an LLM and then you end up with this guy's, and it's 32.5 and you go, I think you're wrong. And then you get this response going, you're absolutely right. And then it goes, you come across that, right? And you're like, it shouldn't be me having to tell you this. So that blind optimism is so dangerous. So yeah, I'm using those tools, but with very, very wary.
Sarah (32:13)
Yeah.
Andy Cotgreave (32:29)
constraints.
Fi (32:31)
I've an example
I have to cut in because I'm just dying to give you an example of that. I was researching you for our podcast and I was using our friend ChatGPT and ChatGPT actually told me that you were originally an archaeologist.
Andy Cotgreave (32:36)
Go. Yeah.
boy. Yeah.
Brilliant.
Fi (32:51)
Indiana Jones?
Andy Cotgreave (32:52)
Yeah. Fantastic. But yeah.
Fi (32:56)
Maybe you can dress
up at Tableau conference, you know, with your Indie outfit or something. Maybe that's some Iron Viz inspo. But it was hallucinating, wasn't it? Yeah, okay.
Andy Cotgreave (33:01)
Yeah. Yeah.
not true.
That was hallucinating. I actually updated my... here's a LLM tip. I updated my LinkedIn profile a couple of weeks ago and I put pasted it into Chat GPT and said critique this and it gave me my critique and I said okay that's fine roast me give me the harshest critique you can and holy cow it was really it's like
Sarah (33:08)
as hallucinated by Chat GPT there you go
Andy Cotgreave (33:34)
You're just using non-words, none of this is substantiated. It's all industry bullshit cliches phrases. And actually it was really good. So if you prompt it to roast you, it was actually very helpful. But Sarah, you're dead right, that blind optimism. then, you're absolutely right, I'll do it again. It drives me nuts. So Excel and a bit of AI, I think.
Fi (33:49)
think that's good one.
Andy Cotgreave (34:03)
And a lot of pen and paper. Can't beat pen and paper.
Fi (34:06)
Mmm.
Sarah (34:09)
Yeah, I agree. Love the old lo-fi, get back to the pen and paper. And I like to use about 100 post-it notes as well and just like, you know, have them everywhere. Yep. Yep. Yep.
Andy Cotgreave (34:18)
Yeah, post it. I love them. Love them. Absolutely.
Fi (34:29)
it.
One question that I'm curious about is how do you think that leaders in organizations, and so we're talking the Cs or the head-ofs, those sort of senior directors, what do you think their role is in data evangelism?
Andy Cotgreave (34:46)
Mm-hmm.
Yeah, we've largely focused so far talking about evangelism in products, from the product side. The evangelism for data leaders is a never-ending, brutally hard job of influencing culture, right? Chief Data Officer, you've got four things you have to do from a sort of product strategy
You need a data layer, a semantics layer, an analytics layer, and an insight layer, right? So can go and buy those things and implement those things. But you could buy all those things, implement those things, and absolutely nothing changes, right? Because you've got to go and sell the value. And it's really, really hard to sell the value of all this.
It's really hard to deliver and make people realize that has been a valuable investment. And so the evangelism job for the leaders is about endlessly banging that drum, getting the CEO to use the data in their presentations, getting the CEO to understand how long any chart he or she uses has taken and who built it.
That's a full-time campaign trying to make that happen. Then getting people further down the chain to recognize that they can use data to make decisions. Well, how do you do that? You've got to make access to this stuff so frictionless that they do it. That's really hard. But making things frictionless is nearly impossible. So then you've got this job of convincing people that...
It is worth logging on and filling in the 2FA form to get these answers. Because people are lazy. Humans are really lazy, right? we had a customer. It was awful. They had this line of dashboard usage going up on their server. they'd bust their ass. then the IT team implemented 2FA on the server, two-factor multi-factor authentication.
and usage just dropped off a cliff and they couldn't get it back. And this isn't to that internet security isn't important, it clearly is, but that friction destroyed their data culture trends, right? So what was the fix? So they had to rebuild that. it's about, you've got to be convincing everybody using it. And when you say leaders, it's about the CXOs and the heads off, but also you need to build the army.
of people at all levels across all parts of the business. And I think it's really hard. I've not been a data leader. I've been very lucky that I can turn a chart upside down, change the color in title, and see people do things differently as a result. That's a lot easier, I think, than a chief data officer implementing an infrastructure and then just keep on talking about it.
I mean, you two actually probably have more experience than I do. Mean, what would you say?
Sarah (38:04)
So I find some of the big challenges is quite often the final point is the final point. Like that insight, which to me is the most important point, because why would you do everything else behind it if you're not gonna see it, is often forgotten about, or it's run out of budget.
Andy Cotgreave (38:13)
Hmm.
Yeah.
Sarah (38:22)
or everyone's burnt out from getting the data perfect, they're not interested in the insights. And that's the bit that's like, you've got to reverse engineer it. You've got to start at the end and work backwards, but a lot of organizations don't see it like that. A lot about our role is trying to get in there before it turns into the wrong way around.
Andy Cotgreave (38:43)
You know, two things from that. There's an amazing book called How Big Things Get Done. The author's, two author's names I forget, but it's about, large, it's mostly about large infrastructure projects and why they go disastrously wrong. Sydney Opera House being one of the main case studies in it. It's an absolutely brilliant book, but it's, and there is the Sydney Opera House in behind me. And, and,
The great thing about the book is it's also just a really good project management advice book. Great read. But in it, the authors talk really about filling your planning from the right-hand side, which I don't know if that's the right-hand side of my screen. What are you driving change towards? And then you go back from there. I mean, God, this sounds so basic, and it's so hard to do in project. And then the other thing is...
You know, we've talked, there's questions you know you're going to answer and there's questions you don't anticipate. Donald Rumsfeld had that famous unknown knowns poem, but he was a bit of a monster. He actually stole that idea from NASA. So it was originally a NASA idea. And, you know, again, the power of Tableau that I love is that I can build dashboards to answer the known questions and I can rapidly respond to ad hoc questions as they come up.
Fi (40:09)
Mm.
Andy Cotgreave (40:11)
I actually wonder how many businesses, right, what's your experience? I've never really thought about this. If you were to ask all the questions, if you were to somehow understand all the questions being asked inside a business on any given day, how many of them are ad hoc and how many of them are known? I'm asking because I wonder if actually ad hoc questions are more common. Hmm, I don't know.
Fi (40:34)
I think they
are like anecdotally, I would say the weight that I throw towards the those questions off the cuff questions so much higher than whatever we've developed in the standardized dashboard that's there. Now the standardized dashboard is great because it's gone through the governance, we've ticked off. This is the way that we are calculating these metrics. it's really gone through that design processes and it's really robust. But
Andy Cotgreave (40:42)
Mm-hmm.
Yeah.
Fi (41:03)
that often leads to more questions than answers. Why is that dropping? What's happened here? Tell me about that. And they need to, and it's impossible to build that all within a dashboard. Otherwise your dashboard becomes bloated, overweighted, and too complicated to use. And that complexity then drops the adoption. So I believe it's more ad hoc than standardised questions.
Andy Cotgreave (41:10)
Yeah, yeah.
What do think, Sarah?
Sarah (41:33)
Yeah, I tend to agree. I think if you've got your standard stuff running nicely, and everyone knows this is this is our sales, you know, month to date, and we can see all the history I think that builds the trust. And then I think it's the insights, the actual
Andy Cotgreave (41:45)
Mm.
Sarah (41:50)
ad hoc is way more important going back to building everything behind the scenes and using a tool like Tableau, you should be able to build those ad hoc pretty quickly and have the right people in the business that can do it. So it's not slowing a department down. It's something that everyone can do themselves. I think that's where you get the real aha moments from a tool like Tableau.
Andy Cotgreave (42:14)
Yeah,
everything people can do it themselves. I think that's the hardest bit. It's like, right on a given moment, you know, it's a person in marketing who's trying to run a campaign for a conference, right? Or a person in sales who wants to know which person's call next and what they've been doing, right? That's a spontaneous question and getting the answer to that. It could be so hard to build a data source that can answer those questions. then, you know, we...
The three of us, we love Tableau, we know how to drag and drop. We know we just need to open Tableau, connect to data and do it. But those three things are actually hard if you're not a data analyst and if you're really busy trying to do a non-data analyst job. And so solving that gap, mean, that's what people have been trying to do for over 120 years. I still don't think we've solved it.
Sarah (43:07)
Yeah, I think we have got some ways to go, but we're getting we're getting more and more even in the couple of decades that I've been around data, I've seen it more in the hands of business than ever before. it's increasing, exponentially I really see that tech people are good with the with the
Andy Cotgreave (43:11)
Mm-hmm.
Sarah (43:27)
with the kind of tech side of things and let the business do the development and so forth and build those reports out. And if you can evangelize the tools as well and get everyone with the power, some ability to at least make a start and feel confident in it.
Andy Cotgreave (43:32)
Yeah.
Yeah, indeed. Indeed.
Fi (43:47)
Okay, Andy, this has just been absolutely amazing and I can see our time is creeping up, but we do have a few rapid fire questions that we have prepared. So if you don't mind, we'd like to switch gears a little bit. All right, so we spoke a little bit about Iron Viz at the beginning of this call. What's one thing about Iron Viz that no one knows about?
Andy Cotgreave (43:53)
yeah.
Yep.
Do it.
Mm-hmm.
well, we talked, actually, we talked a little bit about the progression of contestants, you know, that genuinely is seeing them walk. And that's been epitomized. You know, at Iron Viz, each contestant has to walk onto a big music fanfare, right? And then they've got this enormous stage in front of five to 8,000 people to fill. That's a really intimidating thing. Watching their first rehearsal, they're always like this, right?
And then to being on stage with open arms, that's amazing. I also alluded to data sets. Finding an Iron Viz data set is absolute bloody nightmare. Because one of the curses of working in business intelligence or analytics is, can you just build a chart? I'm doing this thing, can you just build a chart? I'm like, yeah, how long do you think that's going to take? Well, you're just building a chart. that's easy about,
three hours, I'm like, well, no, actually, I got to find your data, or I got to clean your data, or I got to ask you what you want, and then I got to build it. It's actually a week just to get a chart. Oh, I haven't got a week. It's like, well, should have thought about it. Finding a data set is, for Iron Viz is the same, because it has to be clean, it has to have a really good timeline, it has to have loads of dimensions and measures with good granularity, and it has to be within a budget and legally approved. And holy cow, it's so hard, it's so hard.
And something people don't know is we always start the process too late. Every single year, it's an absolute, they're like, oh, dataset, oh God. So thank you to Mark Bradbourne, who has helped valiantly over the years with me scrambling many times to find some data. I'm very excited about the 2025 dataset, but I can't tell you what it is. But I'm really excited about it. Can't tell you. No hints, nope.
Sarah (45:50)
Ha ha.
Nice.
Not even a hint.
Fi (46:04)
No, come on.
Sarah (46:06)
No, their
heads.
Andy Cotgreave (46:08)
Okay, I'll do my, come to Tableau Conference in April 2025. Join the data fam. There you go, that's the plug.
Sarah (46:18)
Nice. What is your favorite data visualization of all times, Andy?
Andy Cotgreave (46:24)
Well, it's got to be the Iraq Bloody Toll from the South China Morning Post, early 2012 by Simon Scar, because it enabled my career to flourish. think, yeah, that's the favourite. it just, not because of the impact it's had on me. I mean, it is the perfect epitome of somebody seeing the shape of data and recognising that they can...
control the story with very simple choices.
Fi (46:55)
I mean, it's super powerful. You know, the impact that you can see that color and position have on that visualization is just breathtaking.
Andy Cotgreave (47:06)
Mm-hmm.
Fi (47:10)
I think it's really important and there's so many new people that are coming along in their careers that won't have seen that. So one of the things that my all-time favorite people says is you've just got to keep telling people over and over again.
Andy Cotgreave (47:22)
Yeah.
You know, something that bugs me about this field is every now and then you'll be like, somebody will post a blog or a tweet or a write, you're like, why are we still talking about pie charts? Can we stop talking about pie charts? Or they'll look at Menard's chart from 1869, which is on the wall behind me and be like, yeah, this chart could have been done much better if da da da da da. It's not that good a chart. And it's like, okay, that opinion is the opinion of somebody wise.
Once you've talked about pie charts for five years, you know the good and bad of them. And then you can also think alternatively. Once you've seen them in our chart 10 times, it's like, yeah, okay, Napoleon did this and then he came back. But every year, there's hundreds of thousands of new graduates who never had the benefit that we have of being in the field for five years. So you gotta keep telling.
Sarah (48:13)
What's a book that every data professional should read?
Andy Cotgreave (48:20)
I'm gonna go for two Living in Data by Jer Thorpe He's an ornithologist and data journalist and this, I love this book. It's just about data in society. It's just about sort of a philosophy, politics, the essay about how we use data. And I think, you know,
Fi (48:23)
Yeah, break it.
Andy Cotgreave (48:48)
We all know the practitioner books that we can recommend, but this is a brilliant one to think differently. Then my other favorite one. I mean, everybody, did you recognize this one? Have you seen me present this? Graphic methods for presenting facts. This is from 1914. Oh, you do not know this? Let me read, right. Let me read the first page. So in 1914, Willard Cote Brinton wrote this book.
Sarah (49:09)
Wow. Darn.
Andy Cotgreave (49:17)
And on page one, he says, millions of dollars are spent yearly collecting data with the expectation that the data itself will correct the problems being studied. Page one, right? So 1914, he was like, why are we wasting millions of dollars on data? It's 2025, everybody. Still haven't solved these problems. mean, this book's amazing. It's like, don't use pie charts.
how to build interactive visualizations with index cards, up a data culture. Honestly, it's amazing. It's absolutely brilliant. He also says there's danger in giving too much information to executives of small brain capacity. Truth. Executives who are very busy, I suspect he would say in 2025. Anyway, there you go. There's two books.
Sarah (50:13)
I'm just thinking where I can use that quote,
Fi (50:17)
he had a time machine and actually came to visit.
Andy Cotgreave (50:22)
Well, think here's the thing, right? The first thing human beings ever wrote down on little clay tablets was data collection. We were counting things, right? Cuneiform and boiler pots. And then 110 years ago, we had that. We've got big data. Well, we've got AI now, right? But at all points, human beings have used the technology they've got to do cool stuff. And then as soon as they've done that cool stuff, they have a little bit more ideas about what they want to do, right? So we get to 1914 and we have...
tabulating machines and index cards, but Brinton wanted more. Now we've got massive data storage, automated stuff, and we still want more. So we've just always got ideas, have just always advanced ahead of data and technology. So I suspect he'd be blown away, but he'd still be frustrated.
Fi (51:13)
What's your most memorable data storytelling experience?
Andy Cotgreave (51:21)
I don't know, this is a bit of a cheat. That quote I just read out, I used that at the start of most presentations. So this is a presentation thing. Anyway, so I'm like, here's the quote, millions of dollars. I'm like, who thinks that's true? Hands go up. Well, when was that quote written? And it's almost always the same, 2016? 2004? And then some wisecracker in the audience would be like, oh, it's clearly old.
1980 or 1960 and I'm like no you still must have been 1914 and every time seeing people's like what? You know it's one of the most audible slide drops I ever did. It's not quite that storytelling but it just makes me chuckle every time.
Sarah (52:15)
What's one myth about data storytelling that you'd like to debunk?
Andy Cotgreave (52:21)
that there is a right chart for any situation. if you build a chart and somebody says it should have been a line chart, you will face my wrath because that's just bullshit, right? Let's assume charts are good. Let's assume it's a good effective chart. If it's telling its story, it just...
The right chart depends on what you're trying to achieve, depends on the time available to your audience, the literacy of your audience, the insight you're trying to give away. I coined Cotgreave's Law in about 2017, 2018, that the longer an innovative visualization exists on the internet, the probability somebody says it should have been a line chart increases to one.
Fi (53:06)
See, it's interesting. I usually say it should be a bar chart.
Andy Cotgreave (53:10)
Yeah, or bar chart. Yeah, it could have been a line chart or a bar chart, assuming both the original one is an effective visualization, is a well-built one, a line chart gives you a different insight. You're going to show one thing at the expense of another.
Sarah (53:27)
ever be an exploding 3D pie chart?
Andy Cotgreave (53:30)
I mean, you know, I'm old now. I've been in the field 20 years. Yeah, maybe, right? mean, you know, if you're trying to, you see, I can feel listeners just be like, that Cotgreave he jumped the shark on that podcast. That was when he really blew it, right? But if you're trying to just show on a, yes, yeah, you can make an exploding pie chart with, in the right place.
Fi (53:56)
Wow. Where we, yeah. You did hear it first and where we might debunk that one or where we're gonna take a shotgun to it. I don't know.
Sarah (53:57)
You heard it first!
Andy Cotgreave (54:07)
Yeah, well, bring
it on. Let's argue and I'll show you a case where exporting 3D pie chart is good.
Fi (54:16)
what we do is we give you all the shit charts and you explain why yours yours is better than all the normal charts
Andy Cotgreave (54:18)
Mm-hmm.
Yeah,
well, that'd be good. Well, that's kind of what we do at Chart Chat is we often will look at a chart and I mean, we are always very careful not to say the charts we are looking at are shit because again, we don't know what the end user was trying to achieve, how skilled they were in the field and who their audience was, but we will then remake them with different, you know, just by applying different perspectives.
Fi (54:29)
Yeah.
Okay, moving along.
Andy Cotgreave (54:52)
Think,
yeah, yeah, yeah, yeah, move along. I'll have to come on again. I've got so much more to say.
Fi (55:00)
know, and I'm so, and I'm thoroughly enjoying it. You know, I really didn't want to jump into the quick fires because I just wanted to keep the conversation going. And I know that this means that it's almost coming to an end. What's the best Iron Viz outfit that you've ever worn and what was the story behind it?
Andy Cotgreave (55:08)
You
Yeah.
The first one I think, so in 2018, Ross Perez left Tableau and was like, Andy, do you want to do it? was like, hell yeah. And so if anybody remembers 2018, I had a suit that was blue with black paisley pattern on it. It was absolutely, it was just mind blowing and I had these crazy blue shoes and
and I arrived at Tableau conference, this was in New Orleans with a black dinner jacket and the blue suit. And Amy Barone, who's the head of events at Tableau back then, amazing, worked so closely with Elissa and Ellie Fields, three incredibly exceptional women who were super leaders that helped build Tableau into what it was today. And I was like, Amy,
I'm not sure which suit to wear. So I brought them both to the backstage and I'll put the blue one on. And she was just like, Andy, this is not even a fucking decision, it's the blue one, isn't it? I just needed that affirmation. And I remember walking out on stage, and that was 5,000, 6,000 people. I'd been on stage in front of 1,000 people before that, so it was a big step up. And I just remembered, I sort of felt.
something tangible as I walked out on stage, everyone was like, that's a suit, right? And it just felt like I had a superhero suit on, just, was, that was the mind blowing sense of endorphins and adrenaline coursing through one's body. I'm like, I want to do this for many more years. Let me do it. Brilliant.
Fi (56:50)
Mm-hmm.
I remember it.
Sarah (56:57)
Me too.
Fi (57:04)
I love that. I love it.
Sarah (57:07)
Yes, awesome. So we all know you quite well and you love the old magic trick.
Andy Cotgreave (57:14)
Mmm. I do.
Sarah (57:16)
So if
you had to explain data visualization as a magic trick, what would it be?
Fi (57:18)
you
Andy Cotgreave (57:25)
I don't know, alright,
Fi (57:25)
Can you do it?
Sarah (57:25)
you
Andy Cotgreave (57:29)
I'm rusty so audience if this doesn't work and also I'm on a right I don't even know if I can do this here's a here's a big coin this is a cool coin it's a tableau New York Stock Exchange souvenir coin right I'm gonna make it this right I'm not gonna look at I'm gonna make it disappear on three okay so we'll go one two
Fi (57:33)
We're just gonna...
Andy Cotgreave (57:54)
Three. it's not. yeah, look. All right, it's there, right? So that was cheap, but look, one, two, three, and it disappeared. The coin disappeared too. Second career is ready. There you go. Why? I can stand on stage in front of 10,000 people, but doing a magic trick scares the shit out of me. Right, why does this explain data visualization? Because.
Sarah (57:59)
yay!
Fi (58:10)
Still got it. Still got it.
Sarah (58:11)
Still got it,
yeah.
Andy Cotgreave (58:24)
Data visualization is about using the cognitive system, the eye and the brain to enable people to see things. What's this trick? you go, one, two, yeah. So I'm hiding things, so I'm misdirecting. can't do it. And then when I turn around, I'm just dropping the coin. So you're all looking here, but I'm doing something different over here because you're not looking at it. Magic is about using those same processes to fool people.
data visualization is about keeping people's in the eye at the right place so they're driving their insights.
Sarah (59:00)
like it? Great answer.
Andy Cotgreave (59:00)
can tell you my heartbeat has just
rocketed. Because I haven't done a magic trick for tons. I'm over 100 beats per minute,
Fi (59:05)
If you look
That's just because you're with us
Andy Cotgreave (59:13)
Well, so my heartbeat has probably been at 90 because I'm with you guys and that is just rocketed because I'm like, my God, I've just done a magic trick on screen. A simple one. Anyway, carry on.
Fi (59:25)
That's awesome. Can I have another magic question? Alright. What's a trick, data or magic that always surprises people?
Andy Cotgreave (59:29)
Yeah?
my gosh, a trick, a data trick and a magic trick. I think my favorite magic trick is a trick called triumph, which essentially the deck is in the audience's hands and they predict the entire order of the deck based on its color. So is it red or black? And they have the deck, they shuffle the deck and they get...
every all 52 cards predicted correctly. It's absolutely brilliant. It's a brilliant trick because it's impossible. It's impossible. And one of the great things about magic is the skullduggery of magic is so much fun to know because you just like if you knew, if you knew. So that's one of the reasons I love magic, a data magic trick.
You know, remember in Iron Viz when it was, that was Lindsay as well, when she just dragged and dropped text from a calculated field onto the data pane and it created calculations. It was like, boom, blown people's mind. I think that was one of the best Iron Viz moments.
Fi (1:00:29)
Mmm.
Sarah (1:00:36)
Yeah.
Fi (1:00:41)
That's awesome.
Sarah (1:00:42)
Yeah.
So Andy, this comes to the end of our podcast. I've had a great time talking to you. I'm sure Fi has as well. We talked about so much. I loved hearing more about your journey with data evangelism and some of the tips and tricks that you've taught us along the way. We are unDUBBED I'm looking forward to seeing your book. What is it called again? Coming out in September?
Andy Cotgreave (1:01:06)
Dashboards that Deliver
September 25.
Sarah (1:01:10)
Yes, looking forward to that. And if you don't already have a copy of this book, I'd highly recommend it. And also check Andy out on Chart Chat, which is a very exciting podcast. It's in collaboration, isn't it Andy?
Andy Cotgreave (1:01:24)
Yeah, that's Amanda McCulloch, Steve Waxler, Jeff Schaeffer, and myself, rabbiting on about the minutiae of charts. I love it. Episode 59 is live at Tableau Conference.
Sarah (1:01:39)
Awesome. And talking about if you want to see Andy in his next sparkly suit, that will be coming as well next month at Tableau Conference.
Andy Cotgreave (1:01:47)
Yeah, we're really excited.
We've got some great contestants for Iron Viz and a dataset that's very close to my heart.
Sarah (1:01:55)
Awesome.
Andy Cotgreave (1:01:56)
Teaser.
Fi (1:01:58)
That's awesome. Well, thank you so much, Andy, for joining us as a guest today. I've had an amazing time and I really wish that we could just sit here and talk the night away, although I'm sure that Sarah might want to pop off to bed at some point. For everyone listening, please like, subscribe, and leave us a review along the way. And we look forward to hopefully welcoming Andy back sometime as well. Sounds great. Cheers. Bye.
Andy Cotgreave (1:02:20)
Yeah brilliant thanks very much.
Sarah (1:02:22)
Thanks, Andy.
