Lazaro Gamio is a deputy managing editor at Axios, where he oversees a group of visual journalists that make charts, maps, interactive graphics and editorial illustrations. He previously worked at the Washington Post as an assignment editor on the graphics team, and before that, at his hometown paper, The Miami Herald.
In this week’s episode, Laz and I talk about his work at Axios and the team he leads. We also talk about the different kind of layout and feel at the Axios site as compared with other news organizations. We also talk about some of the non-standard graphs he and his team has created, what’s drives him to these charts, and how to help readers understand those chart types.
If you would like to support the show, please consider leaving a review on iTunes (or your favorite podcast provider) or becoming a Patreon supporter. With just a few bucks a month, you can help me continue paying for sound editing and transcription services.
By the numbers: How Trump properties profited from his presidency
Hi everyone. Welcome back to the PolicyViz podcast. I’m your host, Jon Schwabish. Thanks so much for tuning back into the show. On this week’s episode, I’m really excited to welcome Lazaro Gamio from Axios. If you haven’t checked out the Axios site, you really should; they are doing some really cool work with data and data visualization which is, of course, where I focus on, but also on the news reporting as a whole. Laz leads the data visualization team over there at Axios. He’s worked at the Washington Post and The Miami Herald. So we talk about the work he and his team are doing over there, some of the nonstandard graphs that they’re using and also about the sort of non-traditional different type of layout and feel at the Axios site. You’ll also notice that we have a little fun at the beginning of this episode.
Sometimes I just cut that stuff out when I have a little conversation with folks before we actually do the interview, but we had such a good time this time around I thought I would just leave that in for you. So hang in there for a couple of minutes until we get to the actual content. One last thing before I get to the interview, if you would like to support the show, please consider leaving a review on iTunes or on your favorite podcast provider. Or if you’d like to become a Patreon supporter, that would be great. Just a couple of bucks a month will help me cover the costs of sound editing and transcription costs, and also, of course, to pay for all the web development and web hosting needs. I’ve put links to both of those, iTunes and the Patreon page on the show notes page. So please do consider supporting the show. So here is my chat with Lazaro Gamio from Axios.
Jon Schwabish: I am not even going to introduce because I’ll do that later. I’ll just stitch it on. And the only rule is like and I’m bad at this too, like not trying —
Lazaro Gamio: Oh, yeah,
JS: Pick a distance.
LG: So maintain the same distance.
JS: Maintain the distance, but other than that it’s not new.
JS: I’ll do this one.
LG: This feels like I can maintain.
JS: You feel that you can do that?
LG: I can maintain this for 20 minutes, maybe, maybe not.
JS: How long can you hold that? How long can you hold it, right?
LG: I am extremely fidgety. I think I cannot stay in the same position for like 15 seconds.
JS: Do you need something like I’ve got like or something.
LG: You know, it actually would be fantastic.
JS: I’ve got the earth.
LG: Oh yes.
JS: Then I’ve got this from – originally, I got this thing.
LG: Oh, look at that. Yeah, I just love to squeeze the earth between my fingers. This is a metaphor. It’s a metaphor. All right.
JS: All right.
LG: All right.
JS: Let’s just do this. Let’s just give it like five seconds of quiet for the air so we can pull the air out of it.
JS: And then I am going to start.
LG: That’s fine.
JS: It’s mid-December, right? Well, I feel like everyday is like —
LG: It’s a reflective time of the year, this is a perfect time to do the podcast that I will do for a living.
JS: So this will come up, like I think I want to do like second week of January, some like that.
LG: Great! Plenty of time to look for new job after my boss is here, he is saying like what the f**k are you doing? Oh, can I curse?
JS: Oh yeah, you can totally curse. Oh, you can do whatever you want.
LG: This is great.
JS: I love the little explicit.
LG: That’s great. How many of them have the explicit thing?
JS: Just one.
JS: Kim Rees, I think, is the only one who swore.
LG: That is bulls**t because all these f**king people curse every single day.
JS: Oh, all the time.
LG: Yeah, like I can —
JS: Because they try to hold it on it.
LG: But you told me this is not a professional podcast.
JS: I mean it’s professional, but it’s not, you know, I mean, look at me. I’m just like, you know.
JS: I’m slapping microphones down.
LG: So for listeners at home, Jon’s not wearing pants.
JS: You know, it’s like doing a webinar. It’s pencils teaching.
JS: As long as you don’t stand up halfway through, you just have to look at on top.
LG: Oh man, life is a facade.
JS: Let me just start.
LG: Go for it.
JS: I mean we’ve already started. I’ll put all this in there. I’ll just have you.
LG: It’s great. Yeah.
JS: Do you want to tell folks about you yourself?
LG: Yeah. So I’m pretty sure you can just read my name on the website, but I’m Laz. I lead the visuals team at Axios and, I used to work at The Washington Post, I was an assignment editor at the graphics department there. Now my job is a little different. So like, I went to Axios and I was like, “Hey, can I just make charts?” And they were like, “Hey, how about you edit this team?” and I was like, okay. One of my problems is I never say no. So here I am. And the team is a little different than the one at the Post because I’m in charge of both data viz and illustration.
LG: So I’m sort of helping art direct the look of the editorial illustrations that appear on the site and also edit the data viz it appears. So it’s like I’m using both sides of my brain, the sort of art side and then the sort of more technical side that you have to employ for databases.
JS: How many people are on the team and then what’s the split between like the illustrator side and then the data viz developer side?
LG: Sure. So right now in total we are eight folks, including myself. So I’m the visual editor. We have four folks doing database. Those are Chris Canipe, Harry Stevens, Andrew Witherspoon, and Naema Ahmed. And then we have three folks doing editorial illustration. So that’s Sarah Grillo, Rebecca Zisser, and Aïda Amer. So it’s a four-three split right now and we’re hiring one more person for the editorial side.
JS: Was it that size when you got there?
LG: No, we started and it was three of us. So in two years since — you know Axios is as old as the Trump administration, so about 10 million years. Since we started two years ago, we’ve essentially tripled the size of the team which is fun.
JS: So tell me about the different set of between since you’ve done like the big standard Washington Post, the hundred years, and then like not a startup, but essentially a startup news or young newsroom.
LG: We’re an old startup.
JS: Old startup, yeah, old startup, how do you view the differences between the two?
LG: Yeah, it’s interesting. I think when I left the Post I was looking for something a little bit more dynamic which is why Axios is so attractive to me. It’s like there are no rules. Like it was like no bureaucracy, no nothing. And there wasn’t, right? Like it was a wild west. Like I sort of just did whatever I want and we had success. Now that we’ve built more structure, it’s still very different I think from what your traditional newsroom is like. I think the biggest difference between the work I do at Axios and the work I’ve done in the past, it’s just that the format it acts is very different. Like we’re all about the smart brevity, keep things small and concise, which I never really understood the luxury of having space.
LG: Like having a blank slate, every time you open a graphics template at a big shop, it’s like, oh, I can do whatever I want. I can have like a big photo, like a little tiny chart here and just sort of take my time, because we’re taking this very sort of rigorous reader-first approach where we know for a fact that when readers open website, they, most of the time will just close it without reading anything.
LG: And they do read, they’ll maybe can’t get pass like one scroll, right?
LG: So taking that rigorous approach means that we have to be equally rigorous, as rigorous as we are on the writing with our visualizations. So what’s the one chart that we can do to sort of communicate we want to communicate and that’s it which is kind of hard. It’s so much easier to do more than it is to do less, there’s so much time that we spend just trying to figure out what shouldn’t be there, which is a different type of labor.
JS: So I want to come back to the look of Axios because it is like really different from that place at this point on, essentially more versus less. So I’ll give you a statement that I gave to a previous guest.
JS: Okay. So technical considerations aside, it is harder to create a good effective static graph than a good effective interactive graph.
LG: Is this just true or false?
JS: Maybe. True or false with a explain-your-answer-below.
LG: So they’re not really inherently different. If you think about it, like an interactive chart is a static graph until you interact with it. So it could just be sh*tty. There you go. You can put explicit on it now. So it can just be a sh*tty chart the same way like a static chart. I think the tools that you have maybe make it easier to make a better chart because something like Adobe Illustrator is much more expressive than like D3 can be. D3 is expressive in how you can use it to do anything you want, but then the annotation layer is a little bit more laborious.
I mean I think technically like if you’re talking about like the knowhow required to make an interactive chart, yeah, definitely, it is much harder to do it in D3. Once you’ve got your skills in place, like…
JS: Well, I guess some people, I mean some people would argue that Illustrator or InDesign like those are hard. I mean they’re clearly at the menu, but they are heavy, dense tools.
LG: So what’s the easiest tool to make a chart?
JS: I mean this is like an ongoing question now.
LG: It shouldn’t be easy.
JS: It shouldn’t be easy to a point. Like, I want the tool to allow you to do everything. Like there’s always this sort of maybe goal that you’d have a tool that like you could do the data analysis and you can make the graphs and you could do the interactivity and make good design. And I worry about like that tool, like the one to rule them all because I don’t like, I don’t want statisticians to be making design decisions easily then. And people who don’t know anything, don’t know enough about data to be like doing statistical analysis. But it’s easy enough now to like do a regression, but not everybody should be ready. So what’s the easiest tool? I don’t know. So what do you guys use?
JS: And what about the data analysis part?
LG: So from there we do, I mean, I work a lot in Python. Some folks work on node. We do a lot of R2. So it just depends; everybody has their own skillset and like as long as you get to where you need to go.
JS: So how does the illustrator side work because I’m not as familiar with that. Like if I have to make five different versions, is that a manual process?
LG: Oh yeah.
JS: Yeah. So you’re creating a new artboard of the right dimensions and then —
LG: Oh, we have a template. That would be madness to try do that each time. Nobody wants to work that hard.
JS: Yeah. And so what are the templates that — so you have templates for all the style decisions?
LG: Yeah. I mean it’s all like there are a lot of good guidelines there. What do you call it? Guard rails. So like all the fun, you have right type faces, right colors. I think we’ve made, I mean I’d say given a week we make about 20 visits. So like we work, we’re a high volume shop. So doing that, we have a good corpus sort of lean back on and be like what was that, how do we do this last time we did it. So like we have a design language that is ongoing and evolving and sort of everybody is participating in.
JS: And does that include the reporters as well after they get to chime in and say I wish this chart had a different look to it or something like that?
LG: Well, I think so when we get feedback from reporters, like I think feedback like I don’t like red is not feedback. That is just like okay, cool.
JS: Like that’s just a complaint.
LG: Yeah, exactly. I think feedback from non-viz people is valuable in gauging the readability of the chart, where like, I don’t get this. It’s like if you have an expert in any given field that asks you for HR and they don’t understand what you did, I think it is time to re-evaluate what you’re doing. So I think we have, again, like a pretty robust feedback process where like we were really big on slack. So you’re making a chart and you post it in a room and then everybody yells at you for five minutes. And then often you’ll show it to the reporter and then it’s in a really good spot by then and there shouldn’t be any problems. But obviously a lot of those discussions should happen beforehand. Lots of dialogue makes good work.
JS: Let me ask you about the bigger project. Maybe they’re not bigger projects, like you especially I think I’ve seen a lot of the stuff that you do is like a lot of like I’ve seen beeswarms plots a lot.
JS: I think that’s mostly will be in the examples but I’ve seen a lot of beeswarm plots.
LG: Yeah, which is the hot new chart of 2018.
JS: It’s the chart of the year.
LG: That trend is dying.
JS: So what do you like about the beeswarm?
LG: I think a big problem with anytime, anytime you do look on dot viz, you’re going to have clustering. It’s a big problem with the scatter plots where like there was a whole Twitter debate that I may be part of, about occlusion. So you have a lot of occlusion. So beeswarm plot is all that because you can see each dot and you don’t have to worry about hiding information. Now obviously they can also be really problematic, especially when one of your axes you’re relying on one of the axes, x or y, to encode data and you’re obviously distorting it.
But I think they’re good in that they give you the overall sense of the shape, which is —
JS: It has all the individual points in another shape.
JS: Do you think people get that or do you think your readers get that?
LG: I think they do because I think your normal reader isn’t going to put a ruler up to the screen and be like, hmm, this one’s off by six pixels.
LG: Normal people don’t do that. We’re not normal, but you know.
JS: But even we don’t do that.
LG: Yeah, exactly. So I think there’s like you have so much power to be exact and often, the need to be exact outweighs the necessity to sort of just be clear. Like I think it’s more important to be clear than to be exact all the time.
LG: As long as you have a good reason for and as long as you can explain it.
JS: So there are other nonstandard, I don’t know what the right term is for this, but like, nonstandard, let’s just say, nonstandard charts like outside the lines and bars and pies?
LG: Yeah. I mean like lines and charts and pies, well, we never do pies. It’s a crime against humanity. But I think lines and bars or your meat and potatoes. I think because we have this sort of a space limitation, our sort of gold standard is like make one really crazy thing at the top. We often just experiment, like, hey, what’s the weirdest chart that I can put up there. Like I’ve done a couple of connected scatterplots that some of them have been good, a lot of them haven’t. I’ll be the first one to tell you, we do like a lot of nonstandard encoding. I made this like wildfire thing a while ago, Chris Canipe did this really great hurricane’s piece when he started that just shows, I think like what is it, like wind speed over time.
So we’re really big into — I think having that space constraint has just forced to be very creative about how we encode. Then it obviously creates a density problem where it’s just like a lot to read which is this is like a constant sort of a struggle when you’re making charts. Charts are meant to be looked at or they are visual. So you look at it first before you read it. So whenever you have to read sort of ingest all the information, you’re asking the reader for a lot. So we’re constantly trying to tread that line. We’re like, hey, if you look at this, you’ll get a lot from it, but you have to make that investment.
JS: So explain a little bit about the layout because when I go to like axios.com, it’s definitely a different look than Washington Post. , It’s not set in the three or four columns with the thing at the top and the picture’s everywhere. It’s more sparse; it’s a little more direct maybe, I don’t know.
LG: Yeah, our sort of thing is smart brevity trademark. Our goal is to provide the best information for people so they can get smarter and faster. It’s about having that respect for the reader to just say like, hey, you need to know this and here’s why it matters and if you want to read more, click this button; if not, move on to the next thing. So that is the goal. It is ruthlessly reader first. We’re like, we’re not going to make you read 800 where’s the matter after the first or second paragraph so you can know more. If you want to, go deeper. And we do that on every single story where we have a button that says read more.
JS: Continue reading.
LG: Yeah. And that is it. I mean that is the fundamental mission. There’s a lot of information out there and we’re trying to deliver it to readers in the best way possible.
JS: As the graphics team, you are thinking a little bit differently, packing more in a tighter spot. But I imagine it’s even bigger switch for some little reporters.
LG: Yeah, it is tough to write short. It is something and it’s something that we’re constantly perfecting.
JS: It’s like short with the option of long.
LG: Yeah, but even the long is like maybe 400 words. So we’re putting essentially two paragraphs before the button and then even getting it down to like 300, 400, 500 is long. So yeah, you just have to learn to — you just have to sort of kill your darlings. Like what is essential here, like are you telling people something they already know, like do you really need to say this. In charts, are you putting this to so you can feel okay with what you’re doing, like really long notes. It’s like I’m going to caveat the seven ways and it’s like, well, it is or it isn’t.
JS: So tell me what I want.
LG: Exactly. Yeah.
JS: What about the bigger pieces? So you’re doing like 20 visits a week.
LG: Yeah. Smaller things.
JS: How do you organize and create the bigger projects?
LG: Sure. So I think our benchmark for bigger projects is usually they should fall in one of three buckets. One, you should do a big project so you can learn something technically new. It’s like that is important for us as people who make charts where you should be challenging yourself technically, constantly. So a big project can fall in that category. Two, are you going to be doing something unique with the data set that already exists. Like this is where maybe you can push the field data visualization forward by coming up with a crazy new chart like that is worth doing.
LG: And three is, is the data newsworthy like do you have data that nobody else has either because we get exclusively from somebody or you constructed it by yourself. So you have those three buckets, like those are three checkmarks that I look for when I say yes or no to a bigger project. And then from there, we just have to make time. We are on a daily news cycle. I think I the longest we spent on a project is maybe like two or three weeks which is not a long time. I think I worked on previously, I used to have less than a week. Once we say yes, it’s like, well, let’s make the time, and then, it’s no different than a small one. It’s just you work and you show and you work more. And then eventually we get to a place where we have buy-in from everybody else and we just publish it and then get all the likes on Twitter.
JS: Which is always the goal.
LG: Which is the goal. We’re all working for everybody else on Twitter.
JS: So your staff would pitch ideas to you?
JS: Do you ever get stuff from like other parts of the newsroom?
LG: Yeah, of course. I mean, it’s very everything just percolating. I think what’s one of the benefits of working at a small place is it’s a village. So like everybody talks all the time and there’s a lot of contact between the graphics folks and the reporters. In those situations, you’ll have ideas that bubble up or news bubble down and these ideas come from everywhere. There is not enough time for people.
JS: You just need to slow it down.
LG: Yeah, that’s not going to happen.
JS: You had mentioned earlier about Illustrator and making these multiple versions for mobile, desktop and tablet. What about interactives? I mean I know you can code them so they are responsive. How do you think about people on their phone interacting with the thing versus people on the desktop interacting with them?
LG: Well, that’s what I said before. It’s like every interactive chart is a non-interactive chart until you interact with it. So like when we make interactives, first of all, interactives are expensive. So if you’re going to make something interactive, you better blow my fucking mind. Like you’ve better click a button and literally change my life. I mean we do this all the time where like, we’ll make an interactive that has lightened attractivity where you can hover over a dot, and get information on stuff.
But it should work well without that. Like a good interactive chart has to be a good static chart first.
JS: So when I’m scrolling through it on my phone and I see the scatter plot, I get the point even though I’m not going to try to put my thumb on that little point.
LG: You have to assume that people will not click. So like I think if you are forcing a reader to interact with something, it has to be spectacular. I mean that’s the benchmark changed my f**king life.
JS: And you’re seeing that on usage and stuff like that?
LG: Yeah, I mean, this is known, like nobody clicks on anything. Like people just scroll and like If you spent like three months coding up in interactives, like I’m sorry, nobody’s clicked on it. This is the truth. And I think the reason to do it still is like I think [indiscernible] had a nice long write-up on why you still interact with things. I think it’s good to provide a high level of transparency to readers. There are readers who like, well, look at your source data and like call you out on your bulls**t. And I think those are great readers, like those are very engaged readers. You want to give them that level of insight into like what exactly you are doing.
LG: So it’s a trade-off. Luckily, we are amazing at it.
LG: I sound like an a**hole. But we’re pretty good at like making interactive stuff very quickly. We’ve made him enough now like we have enough tools to rig it up pretty quickly. But you also give up the sort of expressivity that you have to look. It will straight it. We’re like, you want an arrow and a label here and —
JS: This is the next spot.
LG: Exactly. We make it work.
JS: Right. We talked about the beeswarms and we talked about connected scatter plots and scatter plots. So what other crazy nonstandard chart forms do you do you like that you think people get? Because I always come back to when I talked to like students and especially professionals, they are like show them a dot plot or something. They say, oh that’s really cool, but I couldn’t put that in my report because my manager would never get it. Well, let’s look at the New York Times, they have this scatter plot or bubble plot with like a million points on it. An average New York Times reader doesn’t immediately get that until they engage with it a little bit and then they learn the chart type. So are there other chart types that you like to use that you think people now or your regular readers especially like getting.
LG: I think one that I’ve seen that sort of is crossing the threshold is a cartogram. I think if you made a cartogram like five years ago, I think you would still get like, “Well, what is this? Is this United States? That’s bulls**t. No.” But now I think back to these midterms, like most people just defaulted to the cartogram view, which I think is f**king fantastic. It is the better way to look at congressional results. I mean that makes me happy, like equal area representation plan doesn’t vote. So that’s one that I think is bubbling up.
JS: But do you think those translate? I always wonder about these translating to the world, like I did for fun, I did like a tile grid map of the world which was stupid. I mean Russia can’t be the same size as Barbados that’s kind of ridiculous. So do they scale up you think to the world?
LG: I think it depends. Like, I think the reason they’ve worked for political data is because often like one congressional district is one congressional district. So like there is a one-to-one thing regardless of land. I think when you have world data, it’s like it depends on your mapping. Are you mapping something where like one country truly is one country and it has equal weight and what data is Barbados equal to Russia. Nothing. So I mean it’s actually a congressional district cartogram is much more complex in the world because it’s like about 435 versus how many countries are there, 250.
JS: 250, it’s been 28 or something like that.
LG: Yeah. I don’t know. I don’t know.
JS: Yeah. I mean, I find the cartograms of the world usually the ones that I see that I kind of like are they are essentially bubbles. They’re essentially the packed bubble map and it’s not even about the geography per se. They’re just like, here’s the North American countries are here in blue. Asian countries are in purple. And it’s like, it’s not even so much about getting the geography really remotely right. It’s just in the region. Well, obviously, it’s like a US-centered view to begin with. The US is over here on the left and Asia is over there on the bottom-right, Oceania is down bottom.
LG: I think what it means, I think eventually somebody will figure out a really great way of doing it. Then we’re all going to copy that person.
JS: And that’ll be the same.
LG: That’s it. I mean the same way, I mean even the congressional district cartogram is like, they’re still very boutique. Everybody’s got their own cut which is nice.
JS: It’s nice. So I agree that it’s nice, it also is like if I look at I know we’re not supposed to look at Mercator maps, but if I look at a Mercator map projection, I know where Virginia is. If I look at a Telegram map, Axios is going to be different than the Post and the Times and 538 and that’s not the way geography works.
LG: Maybe we should just standardize it, and maybe there should be like a cartogram consortium.
JS: Just agree.
LG: Everybody just decide, we’re going to do this one and then we’ll just fight for two weeks and not agree on anything.
JS: I like, they just pick one. It was just, that’s it.
LG: But they’re all, they all have their —
JS: –little thing.
LG: Yeah, I don’t know. I think it is at a nice spot where like you’re living through something where like you can sort of — you can see design decision to be happening in real time, which is, it’s a nice joy in life.
JS: So what do you foresee for 2019, both for you and for Axios and dataviz in general?
LG: Well, I mean for us, like I think our goal for 2019 is we have such a great fast gear, like we can just produce and produce and produce. I think the challenge for us is like slow down. What do bigger projects like for us, and I would do them like with much more intentionality. For dataviz in general, like I’ve given up guessing to be honest. It’s just I think something that we’re in the midst of right now is the sort of mobile revolution. I think back to like what 2012, 2013 where like The New York Times was producing these fantastic interactive charts, they were all 900 and 100 pixels wide. It fit perfectly on your desktop and then they look like nothing on mobile which is nobody was thinking about mobile because you had like the iPhone and who cares.
Now I think, I would say that more than 50% of anybody’s page views come from a mobile device. How do you translate something really dense into a smaller screen where like you have a shorter attention span, you have less real estate to display your information? I think it’s hard. Like it’s just really, really difficult. I mean that’s something that we struggle with every day. Whenever we make a really crazy chart and it’s like, oh great, now make it read on 300 pixels wide, this is why we have jobs.
JS: So I think it was in Axios 1 you guys did a beeswarm plot that was vertical. It was line-up vertically. It was time, it was a Twitter data or something like that and it was beeswarm and it was time going from the four pass at the top and then you scroll down closer and closer. Have you ever done a piece like that where you’re like, okay, so on a phone it makes sense to go vertical but maybe on a desktop?
LG: Yeah, all the time. Switching orientations is like a really common trick where on desktop like height is probably your most important dimension because you don’t want to be too tall, but you have the benefit of width, whereas on phones, like you can scroll as much as you want. But like you got to pack it all in like tiny little narrow screen. But even then nobody’s going to scroll that far.
JS: That’s right.
LG: It’s hard. It’s just like you have to strike the balance between density and brevity just because you’re taking up space doesn’t mean you’re saying anything. I have a couple of decent practices, but then the thing is if you design something for mobile, then does it just kind of look sparse on desktop. And then you have the conundrum of like, well, why are you putting so much information on desktop and short-changing your mobile readers? Do they not deserve the same information?
But then some people will say like, well, nobody wants to read that much on mobile. It’s okay to take some things away.
JS: Do they do the same for the writing?
LG: No, I mean I think the writing is the same everywhere. But I think writing, text is very responsive.
JS: I am just curious. Like if you think that people are only scrolling up with her thumb so far, but I guess you are just rolling.
LG: You stop scrolling or like again like our text is like you’ll have two paragraphs before to read more. It’s like we do that automatically for you. So it’s already sort of built in. We sort of write and edit and package our content with the acknowledgement that like you are busy even when I have time and you should inform yourself.
JS: Last question. I don’t know if you know any answer to this though.
LG: Probably not.
JS: So in all the post does like AB testing on their articles. So they’re like mix up the mix, the headline and also the picture that does that. Do you guys do that sort of thing? And kind of more to my interest, have you tried doing that with like dataviz?
LG: Well, we’ve done limited analytics with dataviz just to figure out. I mean we did one test where it’s like, are people having on counties? Surprise, no. So we’ve done some limited tests like that, but we have not turned around to do AB tests. But that is fascinating to know. I mean I’m personally interested in knowing if folks interact with like really odd chart types.
JS: Right, exactly.
LG: Because like assume that like a baseline of folks are just doing a scroll by passing. Of the folks that do stay there, are they perplexed, are they frustrated. How can you gauge that?
JS: So there’s the question of why they’re sitting on this complex whatever. So there’s that question, but there’s also the question of there’s a beeswarm chart versus a bar chart and like are people more likely to hover over the bar chart?
LG: You wouldn’t even hover.
JS: Not hover, but like stay. More readers reading the bar chart than the beeswarm and do they spend more time with one or the other. And then you get to this question of why are they spending more time on the one or the other? So I don’t know, I mean, I haven’t heard a lot of people trying to do this AB testing. It’s really hard to do.
LG: Well, you have to be to figure out how to measure it. I think often with analytics, it’s like, are you, the thing you think you were measuring is probably not what you think you’re measuring. So I mean even with a chart type, maybe the point is for people not to have to spend a huge amount of time reading something. In that case that is success and your beeswarm is not successful.
JS: That’s right.
LG: I personally think like I would not like to live in a world where like everybody just decided we’re only going to do bar charts and line charts. That would be sad. I think it is important to just like push the field, like I think it is good that weird charts exist just to show that they are possible. Even if we don’t make it again or it sucked, it’s good that you did it.
JS: Yeah, you tried.
LG: You got it out of your system.
LG: And maybe somebody like 30 years from now we’ll find this and be like, oh, I’m going to improve upon this and they’ll find success. I made this thing a while ago, it was like the Emoji States of America or something like turn off Emoji faces. The guy, a turnoff, the guy who created the chart type, like he did it like late 60s, late 70s or something. And like nobody really took him seriously, everybody was like, wow, this is not good. But I found it in 2018 –
JS: My son loves it.
LG: Great! You know, that’s my target audience.
JS: Five-year-old that’s it.
LG: So I think it’s hard to see the value of like experimenting and like doing weird things because maybe there’s no payoff. Maybe like nobody will read it, but like, maybe it’ll influence somebody in a profound way, much further on and you’ll never know about it, but you will have contributed to the craft.
JS: To the world.
JS: Cool. Awesome. Thank you.
LG: Yeah, my pleasure.
Thanks everyone for tuning into this week’s episode of the PolicyViz podcast. I hope you enjoyed it. I’ve got some great episodes coming up in the next few weeks. I’m really excited to share with you some great discussions with folks from around the world doing some really neat work with data visualization and presentation skills. So again, if you would like to support the show, please do consider leaving a review on iTunes or your favorite podcast provider or also consider becoming a Patreon supporter. For just a couple bucks a month, you can help me out, support the show and not force me to go find advertisers. So until next week, this has been the PolicyViz podcast. Thanks so much for listening.