I've shown this slide before. Big believer in it. That what we do as consultants in presentations, it starts and ends with storytelling. Storytelling, putting things into buckets, breaking down the problem, doing the analysis, and then visualizing that data and analysis to round out the story. So what does this analysis part look like? What does data visualization look like? First take away, it's not about the data. Trust me, it's not. The data is just a vehicle for your message, right? So data visualization not about the data, it's about what the data is saying, okay? It seems like a nuance, very important. So before you even get started, I want you to think, what is the story that I'm trying to tell? I want to show that costs are going up. I want to show that there are three different segments of customers and we should treat them differently. I want to show that you shouldn't be concerned about this particular problem. How would you explain it to your cousin? How would you explain to somebody who knows nothing about this topic? Also consulting's about no drama. And one aspect of no drama is thinking about, from the customer's perspective, what is what is the toughest question that they might ask me? And one way to think about that is, what's the so what? If you show a graph and it doesn't say anything, that's a problem, right? Constantly ask yourself, what's the so what, what's the so what, what's the so what? So a couple of quick tips and then we're going to talk about a couple of examples of how to do this well. What I would say is, be very diligent about documenting your assumptions, your definitions, providing your sources, be credible. Once you have a crack in your credibility, once you create a visualization that is bending the truth a little bit because you took some data out or you said it was this and it was that or you mislabeled something, you're in trouble. Because after that, your client or whoever is in the audience will constantly be a little bit skeptical, a little bit cynical, a little bit suspicious of what you're doing. So be very cautious around misleading with your graphics. And then finally, since you're being credible, since you're being truthful, since you're thinking about the audience, persuade. Make it easy for them to get the point, make it easy for them to agree with what you're talking about. And the best way to do that and if you want to highlight anything on this page, it's right here at the bottom. Keep it simple. Simple, clear, direct. Don't make this an art project because it's not, right? Data visualization. What are you trying to say? Be truthful, be persuasive, be direct. One way to think about it in terms of analogy and having stuck with me during this specialization, I talk a lot about cooking and top chef and cooking shows and stuff like that. So using the analogy of cooking, with the data, you've washed the data, you've cleaned the data, you've sliced and diced the data, you've cooked the data, you've done everything. So what's left now is to put the food that you've cooked on a plate and show it to your diner, show it to the judges. What you don't want to do, you don't want to take the fry pan of food and just dump it on the plate. One, it's not going to be appetizing, number two, it's probably too much, and number three, you're not showing your best results, right? So first takeaway is don't forget to edit. You may have multiple graphs that you could show. You don't have to show all of them, show the one that you feel most tells the story that you want to tell. Always remembering, what's the purpose of this visualization and how much time do I have in this presentation? You may have ten things you can show, pick the one that's most representative your story, put that in the presentation and put the other ones in the back. Now, if you go to Microsoft PowerPoint, you will see that they give you a lot of suggestions. So one, thank you Microsoft for trying to make our life easy, but the number two, bad Microsoft because they gave us way too many. Do we need 18 of these things? No, there's so many of these types of graphs that I never use. [LAUGH] So one, practice and experiment, figure out what works for you, right? Data visualization is a very personal thing. You'll find that different managers that you work for have different preferences. You'll find that different consulting firms have different templates that they'd like for you to use, okay? So find out what works, what does your company like, what does your manager like, what do you like? A couple of main points are one, simpler the better. If they take a look at that graph, [SOUND] in five seconds, they should know exactly what you're saying, okay? Two, is it possible that you don't need a graphic? You can just put it into words. Sometimes words are actually better than a data visualization. And then finally, should we just put this in the appendix because it's a detail? Reverse engineer what you're trying to say, make it obvious. And would somebody who doesn't know this story understand it? One, line graphs are simple. I love it, right? So here in this graph, you'll see it goes from 2019 January to January 2021, and it is going up. It's not a controversial thing to say. It's very clear what it is. It has a title that you can read and very well labeled Y and X axes. Nothing wrong with that, it's an uptrend. We like that, right? If I think about one potential thing that could be better, maybe you should have more data. Not just two years, but why not ten years? In addition to line charts, you'll see a column charts. They're very effective also. They communicate a lot very simply. So here, this is health care related and these are the different sub-service lines of cardiovascular service area. You don't need to know the details, right? The main thing, there are 12 of them. They rank them in terms of highest growth to biggest decline. Super useful. It's also nice that the gray are positive and the red are negative. So it's in sequence. Good use of color. Well labeled. Nothing wrong with that, okay? One area of potential improvement, and you can see I wrote it there right at the bottom, is that this is percentage, right? So this is 27%, 4% minus 8% minus 29%. Potentially a little misleading in the sense that those things are probably not the same size, right? So how big is the opportunity, right? It's very possible that one in the middle is very big and it's a third of all of these procedures, right? So once again, that's a small tweak. Be careful to not only show percentages. Don't be afraid to mark up the slide or the data visualization and draw the audience's attention. Make it obvious like exactly what you want them to see, right? So here, they want you to see then and now, those are not the same. Now, first of all, there's a lot going on here. A lot of columns, blue and red and a lot of really long words here as well. It's a little problematic, right? It's difficult to read. What am I looking at? And so here, you'll see the gray, which is April. This was the tallest gray on the left hand side, that's the point. And then several months later in November, that wasn't the main one. The main was something different here on the right. This doesn't pass the five-second rule. It took me way more than five seconds to explain this to you. It took me more than five seconds to figure it out. Not great. So it brings up the question, is a table even better? If I'm doing all this explaining and it's not really working, you need to take a pause, take a step back and think, is there a better way I should be telling this story? This is the exact same information. The point, and I'm going to make it very clear. Tables are great and we don't use tables enough, right? So I mean, which is better? This one or this one? I'm going to make an argument that the table is better. One, the date's very clear that we're comparing the difference over time. Also very clear that it was ordered, the rows, from the most important in April down to the least important in April, from 76 to 19, got it. Therefore, this was super important in April, got it. Now, even though that was super important in April, in November, same survey, same row, it was only 28% important. Well, what's so important now, it turns out it's this thing here at the bottom at 68%. There are times where a table or a different way to tell the story can be more useful. Key takeaways. Think about the story. What are you trying to say? Who's the audience? How much time do we have and how much of the context do they have? It's okay for your data to tell the story and just really punch them, right, make it obvious. Don't make them struggle to figure out what you're trying to say, okay? The second point is, simple, clean, clear, direct is beautiful, all right? A simple line or a column chart, love it. Also tables. Let's not forget, sometimes you do the analysis in Excel a lot of times anyways, and you rank it from biggest to lowest and you only have ten, and it's helps to attract the audience's attention to what you want. So have fun with it, experiment a little bit, don't be afraid to one, keep it simple, number two, if it doesn't work, put it in the appendix or throw it away.
