178: The One Skill That Makes You More Valuable Than Senior Analysts (ft. Mike Cisneros)

Hey, it's Avery, host of the DataCruel Podcast. First off, thank you so much for listening. I so appreciate it. And I hope you're having an awesome day. I have some sad or exciting news, I guess, depending on the way that you look at it. But I started this podcast five years ago and I started my bootcamp shortly there after the Data Analytics accelerator. We've been running this bootcamp for five years and it's been absolutely amazing. We've helped over 500 people land awesome data jobs. And we're gonna keep doing that. However, ever since I started the program, we've had one price, $2,000 for lifetime access to the program. Meaning that you never get kicked out, you always have access to all the upgrades, the new lessons, you can come back and get more mentoring if you'd like. As long as this product exists, you guys have access to it. And that's going to change. So starting September 1st, we were going to move to a one year membership, which is more than plenty time to land your first data job. This will allow us to cater to our students who are in the program a little bit more right now, as well as continue to bring in mentors and lunch and learns in different ways that we can offer mentoring and new things to you guys as well. So the good news is it's not too late to get lifetime access. If you join this July cohort, this cohort is going to be starting Monday, July 13th. If you join before then, you will get grandfathered in and have lifetime access to the program. Meaning you never get rid of me. I never take you out and you can always come back and ask me questions and get advice, revisit the modules, whatever, all that good stuff for absolutely free. So if you want lifetime access, consider joining this July cohort on starting on the 13th. I have a little bit of a discount on the bonus for you as well when you join. So go ahead and check it out. I data careerjumpsstart.com/da or we'll have a link in the description down below. And hope to see you guys in before those lifetime access goes away. All right, see you soon. Let's get in the episode. (upbeat music) What if I told you there's one skill that can make you more valuable than analysts with twice your experience? And no, it's not Python, it's not SQL either. And anyone can learn it. I couldn't figure out why all these interesting things I found weren't having an impact, why weren't people acting on it. That's Mike Cisneros, a Harvard trained data storyteller who discovered what separates invisible analysts from the promoted ones. And it's this, the best analysis in the world is worthless if no one understands it. I realized it's because I wasn't ever communicating things in a way that was gonna resonate with them. He's led data visualization teams, coach executives, and now he's breaking it all down for you so that you are not in-- - A data analysis cog, you want to be that human being that people go, oh, I know when we go to that person, they are going to be able to explain this to me. He's opening up his data story playbook and breaking down the exact chart fixes that turn people like you to the go to data person in your organization. This is the Data Career Podcast. Let's dive into this data storytelling masterclass. All right, Mike, so if you could go back to your junior analyst days, what's something that you wish you knew then that you know now to make you a better data storyteller and a better data analyst in general? - One of the things that I wish I had known was how much of the job involved talking to people. And I got into data analysis largely for the same reason that a lot of other people got into it is because they were very comfortable working with numbers and they liked the objective truth that comes with looking through data as doing some detective work, figuring out interesting things that you wanted to highlight. But then I couldn't always figure out why is it that all these interesting things I found weren't having an impact, why weren't people acting on it? And I realized it's because I wasn't ever communicating things in a way that was gonna resonate with them. I was putting myself first, I was building dashboards that were largely suited for me and they were entertaining me. I was coming up with all these different crazy graph types that was interesting to build, but they weren't really useful for people to actually look at and utilize on a regular basis. So all of the people who are at the early part of their career, I would say, if you are the person who learns how to explain all these interesting things you found to management, to leadership, to decision makers, then you end up being the person that they come to because you are the one who speaks their language. And that gets you noticed much more than an incrementally better job of finding those little things that might need to have action taken on them. Because the data isn't always the point. The data is usually the raw material that we work with. The data and the analysis is going to highlight, we need to make some decisions about this. What do we think those decisions are? And we have to communicate this in a way that is clear and quick to people who are responsible for a lot of other things in their daily lives and they need to understand things right away. You are always going to understand your data and your analysis way more than anybody else, but you also are more interested in it than anybody else. So it's figuring out how much of that you need to share with somebody else so that they get the point, they feel thoroughly informed, but they are also ready to take action on it. - Okay, that is very interesting. So many things resonated there from my experience working in corporate. The idea of you become a go-to person in an organization for explaining what the numbers actually mean is that ability, that title, that role isn't actually given to you. It's kind of earned based upon your perception of how people perceive you, how good you are at that, basically, how good you are as a data storyteller. And I think that happens a lot to me at Exxon. It's actually one of the reasons why I ended up leaving Exxon was I was kind of a nobody in a 70,000 person company, but I was doing decent work that was getting attention and some bigger people in Exxon mobile started coming to me instead of my boss or my boss's boss for data information and my boss didn't really like that. So you can kind of use this as your superpower in an organization and if you don't have a data job yet, I think also it can be really impactful to recruiters and hiring managers that, wow, Avery maybe hasn't worked as a data analyst before, but I really understand this analysis they did because of this data visualization and how they communicated and how they presented it. And so you are part of the data storytelling with data team and you are one of the co-authors, a new book coming out called Storytelling with Data, Before and After, Practical Makeovers for Powerful Data Stories. And that's what this book is all about. It's basically how to have impact with your data visualizations, is that right? - Yeah, it's exactly right. What we do with Storytelling with Data and what we work with clients to do is take the things that they create on a regular basis, their communications that they build, using the data analysis that they've done and work through how we can make them more impactful, how we make them stronger so that the people who are receiving those communications are able to take that action that needs to happen because too often I think, and I was guilty of this too when I was working as an analyst at yes, a 100,000 person company I was working in Northrop Grumman for a long time in my career prior to Storytelling with Data, is we spend so much time on the exploration of our data and we find things, but then we run out of time when it comes time to share those things with somebody else. And so the communications that our audience sees, which by the way is the only part of your work that they will ever see ends up being geared not towards that specific communication, that specific insight you're trying to share, but really they're just the output of the explorations that you did. And ideally you would create something, spend just a little bit of time tailoring those communications to the people you are talking to so that they can grasp what it is they need to do, what decisions need to be made as easily as possible. And this isn't to say that the communications that we see are bad, they are not, but they can be strengthened. And so you don't want to be every to your experiences, you don't want to be an FTE, you don't want to be a data analysis cog, you want to be that human being that people go, oh, I know when we go to that person, they are going to be able to explain this to me and not just verbally, but in the way that they put their presentations together, this always makes sense to me the way that they communicate. - So one of the things that you mentioned there is just spending a little bit more time on, I guess the final product of our analysis, like instead of just exporting the graph, five minutes before the meeting, bringing in, or maybe you didn't even export it, maybe you're just like in like a Jupyter notebook or you're just like in tablo showing them the data visualization, you're saying to spend more time beforehand. What are some things that we can do as data professionals to try to make our graphs and our analysis more easily understood? - Well, you hit on one thing right away, which is that if you are just taking screenshots out of tablo or how many times that people said, can you just export the data table from this dashboard you built like, well, what was the point of all that? Why did I build all these graphs? Why did I do this analysis? It's making sure that what you are using in order to explore your data isn't the exact same thing that you were sharing to somebody else. But the stuff that you build to explore shouldn't be directing your attention in any one way or another. You shouldn't be predisposed to see one data series or one outlier or one thing just by the way that you laid out your graphs or used color or what have you. It should be as neutral as possible so that you can explore without prejudice, let's say. But when you found those interesting things, you have already made those discoveries and you want to show things to your audience so that those discoveries, already found. Hop, for instance, using color specifically, using color sparingly to highlight something, because it's such a strong way to grab someone's attention. Using, I would say, more words than you might think you would normally use in a dashboard, because the easiest and clearest way for people to see or to understand the message you're trying to deliver is to say it in words. Human beings have communicated with words for quite some time. This is a good way for us to get our messages across to one another, and we know that without words, that's why charades is a game. It's because it's actually harder when we don't use words. We need to make sure that we are clear and as clear as we can possibly be. So a lot of times it's removing the things that don't need to be there for your specific communication. Did you work with 15 data series? Fine. You probably don't need all 15 for this communication you're sending out. Did you look through three years of data? Of course you did, but it's only this two month period that actually matters, so let's limit ourselves to that, or at least make sure we're only highlighting that explicitly for our audience. Getting rid of what doesn't need to be there, that is the critical thing, so that all that's left is just the meat on the bone. None of that extra stuff. It's such a good point that the idea of using words on graphs and charts. I was reading through the book a little bit this morning, and one of the, I don't know, the headers in the book was like make sure that you have whatever the point of the graph is, make sure you have the result or the main takeaway in words. And I was like, oh, that's a good reminder. Sometimes I like to think that my graph is so good that it's like self-evident that you can't miss the point. Like this is obviously the point and maybe I've, you know, I've drawn it out with colors. I've drawn it out with different geometric shapes and I've chosen the right chart and everything. And I'm like, it's so obvious you can't miss it. But it reminds me of what my 11th grade American history teacher, Mr. Ochoa used to tell me. He worked at Disneyland as a teenager. And I don't know if this story is true or not. But he worked at Disneyland as a teenager. And I think he played like EOR, like he'd wear like a giant EOR costume. And his boss would tell him like every week or something, never underestimate the average stupidity or the the stupidity of the average American. And it's something that I've kind of taken that like, yeah, we all are kind of dumb sometimes. And we as our jobs as, you know, data storytellers as translators of numbers to actionable business impact, like why leave it to chance? Yes, you can make the most beautiful chart on planet earth. But like the result, the main takeaway you want that person to think when they see the chart, just write it in words. You can just use it as the title. You can write a little blurb on it. Like why do we like to try to make people guess like what the main takeaway from the chart is? And the reason why it's obvious to us is because we made the chart. Of course, it's obvious to us. We made every choice that went into that particular visual. And it's as a result of the investigation that we did into the data. But somebody else is seeing it for the very first time. And they could make a completely different interpretation than what we intended. And then we've got that confusion to deal with. So it's also that we care a lot about it. We know it. It becomes obvious. Everything is obvious when you have been working with it forever. So yeah, absolutely. It's something that we don't want to underestimate people's intelligence. We don't want to, you know, the explain it like I'm five, you don't have to explain it like I'm five. You can explain it like I'm 50, but I'm just uninformed on this subject. I'm pretty smart. But I just don't wait no, don't yet know what you're talking about. So inform me. And that I think is the difference is making sure that you are being clear because what's the worst thing that can happen if you overexplain something? Everybody in the room gets on the same page. What a tragedy that everybody now knows what we're talking about. But if you skip over it, you lose people in the room and they can't keep up. This is the same thing if you're using lots of terms of art or acronyms or things like in all of your communications, assuming that everybody knows what they mean. But then you're losing people left and right. And you wonder why am I not getting engagement on the things that I'm asking people to do? It's because people in the room aren't listening to you anymore because they feel excluded. Being more inclusive, especially making sure your key audience member is included. Often means starting from a slightly simpler base, but not assuming that they're unintelligent just that they aren't quite informed yet. I like that. It's almost like you've spent hours with this data. You spent hours making the chart. So it's really hard to take yourself, take all that out of your brain and try to be like, if I saw this for the first time, what context would I be missing? It's like if you watched like all all the trilogy of the Lord of the Rings and you had to like explain in like, I don't know, five sentences to someone like the plot of Lord of the Rings. It's like, it's really hard because I just watched all of it and I kind of know exactly what happened. How do I simplify that down to you and give you enough information that you could understand what's going on and kind of like the gist of it, but not also like spends five hours just explaining things that maybe aren't necessary. So you as the author with your co-authors, we should also mention your co-authors are Cole Nathwick, who is the founder of Storytelling with Data in Alex Vela's. Is that how I say the name properly? Okay, perfect. And you guys are all on this Storytelling with Data team. So we've had Cole on the podcast before. And you guys have written this book with like 20 real world examples that you guys have actually seen in industry. Like you've gone to companies, maybe some big, maybe some small, maybe some successful, maybe some struggling. And they all have this problem that they sometimes make bad data visualizations that we're human. We all do, right? And you've taken some of these data visualizations and gone from like maybe a difficult to understand visualization to something that's more clean and easy to understand and can have better business impact because of its clarity. And I think if I'm not mistaken, you've brought an example that we can actually take a look at together. I do. And all of the examples that we use on the book, they are anonymized for everyone's protection, but they are real world examples from clients we have worked with. And some of them, I would say, are good visualizations themselves, you know, taken out of context, but maybe they didn't have the impact exactly that the companies wanted to have or maybe they make sense to a certain audience, but not to the audience that they were trying to communicate with. So what we do is we talk with them and we show them different ways to modify their visualizations for those specific scenarios where what if you're trying to accomplish a, what if you're trying to accomplish b, what if your audience member is x or y instead of z and all of these things inform the decisions that we make as data storytellers. So the book is going to help readers see how we go through the process of not just assessing a visual, but assessing what is the visual intending to accomplish. And then how do we provide that kind of feedback to our clients so that they can strengthen those communications and have the impact that they were hoping to have at the outset. So I'm sharing with you now one of the examples from our book. And as you look at it, if you are just listening, I can describe it. You're seeing a slide that shows the title is hcc volume weekdays ytd et which when I read it out loud sounds goofy. But really what this is is hcc is headquarters call center. So what we are looking at are two graphs two very long line graphs that are showing the volume of calls that come into a call center on the course of an average day. On the top, we have an orange line graph. It's actually an area graph because the area under the line is shaded. And it shows for every minute of the day on an average day, how many calls come into the call center or how many calls are initiated. And then the graph below it, it takes that and breaks it out into a day by day. So it's Mondays Tuesdays Wednesdays looks more like an EKG when you kind of put it all together. There's different bumps in the road. It's honestly, it looks like they took five graphs and sort of stuck them together side by side. So it looks to me like a dashboard that got cut and pasted into a slide presentation. Let me know if that's the impression that you get here as well. Yeah, it's, I mean, the the top graph, like you said, x-axis is time y-axis is is number of calls. But it's like it's almost like one of those things called when you get like a lie detector and it's like, oh yeah, it's really noisy. And it almost the bottom graph, which splits it up by the days, like you said, honestly, kind of looks like my two-year-old scribble a little bit with a little bit more structure. And I'm noticing lots of grid lines. There's like lots of grid lines all over the place. So yeah, I think that's a fair, a fair summary of it. It's a default setting masterpiece. There's grid lines, there's text that's like six point type or smaller. And the older I get, the larger my minimum acceptable font size is looks like just output from a large data set. But yeah, and the other thing I'll just say, sorry, didn't I, but it it looks like the output of like of like an industrial software that is, it's that's not made for data visualization. It's not like it's a tablo graph. It's not like it's an excel graph. It looks like, like for instance, when I worked at Xon, there's like very specific, like refinery specific equipment that makes graphs that kind of look like this or one of my wife was giving birth in the hospital. There's like the thing that's tracking like the baby's heartbeat. And it kind of looks like this. So yeah, it could use some more for sure. So I want to focus just on a top graph for now, which is the in any given day, what's the call volume that comes in? So here's minute by minute. So we want to simplify this to start with. So the first part is getting rid of the stuff that's not necessary. And you already identified stuff like this grid lines, maybe extra decimal places, things like that. But we also want to change the stuff that's going to make it hard for our audience to get a sense of what they're looking at. And things that are making them do too much work. And that, to me, includes, let's make sure these axes are titled. Let's make sure the actual title of the graph is something that a human can read and understand right away. Let's make those font sizes a little bit larger. So when I apply all of those changes to the top graph, we can end up with something that looks a little bit cleaner. Take in a way the orange, which was the original color. I kind of like blue as a color here, it's a little less aggressive here. Getting rid of the grid lines so that the data comes to the foreground, aligning titles in the top left of the graph. So it forms a nice border around. It suggests a border around what the graph is. And then we're labeling all of our axes, getting rid of decimal points that don't need to be there. The original axes went to like 20.0 instead of just 20. Like, what do you need a decimal point there? What I have on the screen looks like it could be the after example of the previous slide being the before, this is the after. What I'll tell you now is that in the book, this is actually the before image, the before to the after that we will create the reason being, even though we've made this better looking aesthetically, we haven't actually made this an effective visual for purpose because what we want to do is think about what is the right data to show, how much data should we show and what was the purpose of this communication in the first place. And this is where talking to people is going to make a big difference because we need to know what was the client trying to achieve at this point for this specific case, they wanted to look at a strategic level. How did they, how should they strategically staff and schedule their employees within the call center environment? So given that looking at the average day makes perfect sense, but maybe looking at it minute by minute is too granular because yeah, the data is still very sketchy, very hairy. There's a ton of noise here. So we said, why don't we start experimenting with different aggregations? Maybe we go into a five minute block of time and do what was the average within each five minute block? And that really smooths out the noise quite a bit. It looks like there's a bimodal curve desperately trying to emerge here, a nice smooth curve, but it's still got a little bit of hair on it. And we said, well, what if we go to an hour aggregation? And now it looks way too jaggy. It looks like we've run out of, I don't know, memory. And so we've started losing polygons and it's just drawing straight lines. And if we're going to aggregate to an hour anyway, I probably wouldn't keep it as a line graph because I would assume we were no longer trying to look at things as changes over time, but maybe compare individual hours, individual day parts to one another. So I might switch to a bar graph, which we have upon the screen now because that would make it easier to compare any one of these hours to any one of the other hours. This is not what the client wanted to do. They did want to look at the trend over time. So what we settled on was a 10 minute aggregation, which smoothed out most of those noisy parts along the day. You can now see what the call volume actually looks like. And this is at Eastern time. So ramps up, starts to ramp up around 8 a.m. Eastern time, peaks for the first time at lunchtime East Coast time, then it dips down again, rises up to its second peak at the west coast lunchtime and then starts to taper off again as the East Coasters start to go home. Then there's a shallower decline over the last bit of the day and all the calls eventually peter out somewhere between 9 and 10 p.m. Eastern. Now I can talk through this and I can draw your attention to those different parts of the graph. Ideally, now that we've gotten rid of all of that noise, we've got a lot of space available to add annotations in. So that's what we suggested they do. Put some reference line in reference lines in to identify specific numbers of calls per minute that they thought were important to call out. They felt like once you get to five calls per minute, that's when our work day really begins. Once you get to nine calls per minute, that's peak time. So we put those reference lines in. We're using color now a little bit differently. We made most of the graph instead of a blue line. We made it all a gray line except for those periods of time that were the peak hours where there were more than nine calls per minute. We made sure that we labeled those 10 minute day parts as well added in how many calls per minute they were getting in those moments. So here we have a graph that does a little bit better job of telling the actual story of what an average day in the call center looks like. But if we're going to present this information to our discussion earlier, Avery, it's using a few more words, not having this graph stand on its own, but rather make this graph part of an actual slide where we use words as the very first thing people see at the top. Make it having a slide title to say when call volume is the heaviest lunch hour and in the mid afternoon. And then all of the narration that I would give you verbally. If I were in the room, I'll put that in text along the right hand side of the slide. And I'll also make sure that I have a very explicit recommendation for what I expect our discussion to be after we've seen this data because as a rule, when you are sharing data, you should have a reason why you're doing it. There should be an action that you expect people to take, even if it's let's discuss what this means. So we want to make sure that we have a recommendation for what to do next and have that included in the deliverable that we give to our audience. So we could do that for again, this day, any given day view, which was the top graph in our before and after or our before graph, I should say. We also had that, that, you know, EKG sort of graph the Monday, Tuesday, Wednesday graph at the bottom, where if you look at it now, looking now at the before graph again, if you look closely, you can probably see that Monday tends to have the most call volume and Friday tends to have the least call volume. And as I draw your attention to it, you see it, but maybe that didn't jump out at you right away when you first saw the graph. So if we applied the same tactics, we just applied to the any given week day graph to the week day by week day graph that we're seeing here, we could point out to our leadership team that not only do Fridays have the lowest call volume, it's by a significant margin that it is the lowest call volume so that this is probably the best time to let people schedule their PTO or if we were going to do team building or have meetings, this would be the best day of the week to do it. Maybe that's not a ground breaking, you know, in sight to have, but it's nice to see that the data can back it up and have it obviously presented in a way that you don't have to do work to figure out that that is what it is that we discovered in our investigation. So that's a good example, I think of the strategic look. There is a second way to look at this that we can talk about in a minute. Yeah, I really like this chart. This is your guys's more finished chart of that bottom one earlier showing, you know, the difference in call times versus day of the week. And yeah, it's a lot easier to see that Friday is by far the lowest volume of calls. It also lets you see, you know, which ones are kind of like the highest. The thing I really like that you did here is you used color very strategically for the blue where, you know, Friday's blue and it sticks out, but also the line thickness is a lot more thick. It's a lot wider and almost feels bolder. So it's like, okay, the story here is Friday and it's just so amazing what color and what a little like thicker line can do there. And I also like that for these that you're showing, you kind of have the graph on the left hand side, a slide title and then some, I don't know, like slide bullets on the right hand side. They're not bullets. It's like a, it makes it very complete. It makes it feel like I could almost not that you'd want to do this, but you could almost take away any one of those three parts and kind of come to the same conclusion. It's, I wouldn't say it's redundant, but I feel like you're catering to people's preferences. Like, for example, I, I've obviously done data analytics for a while. So like I might just look at the chart, but maybe you're talking to a VP or some sort of non-technical person. And it's like, okay, the chart might feel a little overwhelming. Not that this is super complex, but like they might just do it from the title at the time. Oh, Fridays have the lowest daily call line. Okay. And then maybe that helps them, you know, okay, now I see that on the chart that makes sense. I just like we're catering to all different like learning types and technical ranges. Yeah, absolutely. And we're trying to use them in combination, in smart combination with one another because not only do we have the chart title or the slide title with a big takeaway as our title of the slide, Fridays have the lowest call volume, Fridays. The word Fridays is in that blue. And so we assume that everything that's the same color is related to one another. This is a gistual principle, the similarity principle. So if there are light blue or bold blue, whatever you would call that color blue, baby blue, I don't know. You look for that color in the graph because that is what tells you here is where you will find evidence in the visual of what I've said in words. And you see, we use that same color in the narrative annotations on the right hand side of the slide as well. So color ties those things together. And if you have people in your audience who don't have perfect color vision, and I think it's about 5% of people in the Western world do have some form of color vision deficiency. We're not just relying on color. It is something that is helping us to see, but we are using annotations as well where we position our annotations close to the data. All of that makes it easier for our audience to see what it is we want them to notice in the communication we're sharing. We're not leaving it up to chance. And yeah, to your point, we're not using bullet points because we tend not to talk in bulleted phrases. We tend to talk in full-ish sentences. So why not do that in our slides as well? If this is meant to tell the story of what we found, let's tell the story in sentences rather than in bullet points. So this was a good example of doing the strategic look at this call center data. But there are going to be times when going back to this was the looking just at the view of what the average weekday call volume was. There are going to be times when you want to look at a very granular level as well. And what this client found was that on one particular week in February of this year was the sixth week of the year. So I'm going to show you the week six only customer service calls per minute graph. The analyst noted noticed that there were these weird spikes in the morning. There were two or three spikes between 10 and 11, 30 a.m. Eastern time. So they looked into the data a little bit more. What they found was that those spikes were only on a single day. They were only on the Monday of that week. So we're going to zoom in a little bit. We have to make our vertical scale a lot bigger because we now have just a single blue data series here that is showing that Monday's data on its own. And there are huge spikes in the data, but they're all preceded by zeroed out call volume. So this is very strange and it was strange in reality because the people in the call center said, yeah, we were here Monday. Nothing happened. Nothing was weird. It was very much a normal day. So the temptation for the people putting this in their weekly report was just to say, I don't know the data was bad. I don't know what to say about this. So they just put a modified version of the standard report into that week's update. So instead of using a blue line, they used a red line to call attention to the fact that this isn't what we're showing. Normally we would show Monday through Friday. This is just Tuesday through Friday. And they included a dotted line that was, well, if you included Monday, which was weird, and had weird data, it would have looked like this. But we think the red line is more authentically representing what the call volume was. But what happened a few weeks later is that somebody on the analysis team was talking to an engineer who was like, Oh, I have a total explanation for that. I don't know why nobody asked me. This wasn't a data error. This was a system failure. They were like, what do you mean it was a system failure? And it turns out there is a software piece of the system that does the call tracking and routing. It's just the scheduling app. And that kept crashing and rebooting. And what would happen is every time it would crash, when it rebooted, it would re-log all of the calls that had happened since it had last crashed and throw them all into the same minute of time. Like they all started. So that's why you'd have these zeroed out times when the system had crashed. And then these big spikes when the system came back up, then it would crash again. So it didn't affect the calls. It just affected the logging of the times of the calls. So from one perspective, you could say, well, that software glitch was immaterial. It didn't affect any of the calls, didn't affect any of the customers. But it would be irresponsible for us not to report that because just because this was a graceful failure this time doesn't mean it won't be a catastrophic failure next time. So we said, well, here's what you would do is you would create an incident report. And instead of doing your normal visual, you would show that graph with the big blue spikes, but you would also annotate each one of those spikes. So you might put like little numbers next to each one of those spikes, explain what happened at each period of time when those spikes occurred. And then your big takeaway is the slide title, which is the call tracking system, suffered that brief but major outage in week six and explain why it didn't seem to affect the calls. So the choices that you would make for what you would do to improve, let's say, your visual to make over your visual are going to change, depending not just on, well, going from a screenshot of the dashboard to something that's more aesthetically pleasing because that's a good first step. But you've also got to think about who you're trying to reach and what they need to know. And if you're trying to highlight big picture trends, you might aggregate in a certain way, but if you're going to investigate anomalies in detail, you might choose to do something slightly too completely different. Either way, you just have to think about giving your message the depth that it deserves without drowning your audience in the specifics when they don't need to know every single piece of information that you have at your fingertips. Well, that was super awesome. I love how we saw all these different graphs and they're all from the same data set. Like we literally are just showing showing the same data a little bit differently every time and based off of how we show it based off of where we use color or how much granularity we give it. The story is different. The takeaways are different. And so I totally get how it's it's important to make sure that what you're showing matches the story you want to tell. And this was basically from the book and I'm guessing the rest of the book is pretty similar to this. Is that what we can kind of expect? When I can tell you about each one of these 20 makers, this is one of the 20 makeovers that we go through in the book, one of the before and afters. Each chapter we focus on a makeover that helps us to highlight one aspect of data storytelling or one thing you might want to consider or one common business challenge that we always find ourselves trying to overcome or one specific type of communication because this isn't limited just to graphs. We also talk about reports. We talk about emails. We talk about infographics. There are all different kinds of ways to visually communicate information and each one of them can benefit from a data storytelling perspective, especially if it's something that you are then going to share to somebody with the hope that what you are sharing is going to give them what they need to make changes to the world that are going to improve the lives of actual human beings. And I think it's important that we always remember that we are not doing data analysis for the love of the game. We are doing it because we want to make people's lives better, almost all data that we see in some way or another represents an aspect of somebody's life. So keeping that in mind and keeping in mind that what it is doing is helping us to understand people's stories, being able to tell their stories in a way that makes our decision makers able to do things to help them is important for all of us to remember. I love that and absolutely agree with that. So Mike, thank you so much for this explanation and for all of your golden advice. Super excited to dig into this book a little bit more. If any of you guys enjoyed this, you guys will enjoy the book as well. It's available to purchase. So we'll have a link to the show notes down below and I think it'll be well worth the investment. Mike, congratulations on the book and thank you so much for joining us. Thank you for having me, Avery, it was a pleasure. I hope you enjoyed that episode and if you did, be sure to hit subscribe so you don't miss out on any of our future episodes. We have some absolute bangers coming out soon that you definitely do not want to miss. If you already are subscribed, would you mind leaving a short review wherever you listen with your help we recently became the biggest data podcast in the world, which is pretty dang cool if you ask me so thank you so much for leaving a review reviews like yours help others find the show and it also helps me convince other awesome guests to come on the show and share their knowledge with us for absolutely free. So once again, thank you so much for listening for hitting subscribe and for leaving a review. Y'all are truly the best and make this podcast what it is. Thanks so much again and I'll see you in a future episode.