Dashboards 101

Good morning everyone. Thanks for joining our Dashboard 101 webinar. My name is Johnny Girardi. I'm the Data cell founder and CEO. We've been doing analytics and enterprise BI for about 16 years now as a team and quite often people ask us how can we make our dashboards better? So today we will focus on some tips and some techniques we learned over the years from our own experience and from others. And hopefully you will find this very informative. This is our agenda. We'll talk about some concepts and several other topics that have proven to be very valuable when you are building your dashboards. Let's get started. So first of all, how do people build dashboards overall? The same story. Someone asks for a dashboard, someone else, usually a more technical person, a consultant, an IT person, will build it, send it back to the user and what happens next? The user says, well, great, but can you please make these changes? The publisher does it. What happens next? More changes back and forth. This is the regular process to build dashboards. It's completely normal. Building dashboards, bios overall is a journey. It's a process and it takes interactions. Now what I'm going to show you next is how ideally this process should go. Again, ideally, I know that's not going to happen most of the time, but at least I want to tell you if you can go through these steps, you will get to the final version of the dashboard much more quickly. First of all, do a needs analysis since most of the time the person building the dashboard is not the person actually who's going to use the dashboard. So what you need to do as a publisher is start discussing with the people who need the dashboard the business goals. What is the purpose of the business of this dashboard is like a lead or lag kind of dashboard. Meaning is this dashboard to monitor your current goals versus actuals or is more informative to see the past. What are the metrics you need on this dashboard? What are the dimensions? What are these metrics used for? Do the users know the context? So try to understand as much as possible the business goals of this dashboard. So when you build it more quickly, you find or you put together what users will really appreciate. Make a mock up, use some samples. When you're talking with other people that are asking for the dashboards, sometimes they don't know exactly what they want and they have a hard idea putting together in words or in charts and graphs. If you are the talented person that knows this stuff, show them. Design something on a piece of paper. Maybe if you have A library of dashboards. Show them some ideas to get the conversation going that can be very helpful to get you and your users on the same page more quickly. Well discuss, refine the markup and then design it. Now, in the design phase, typically I tell should you show to all users? Should you show to only one or two users? Sometimes this can be a very lengthy discussion. You don't have time. You need to get it out quickly. So sometimes what you learn is some people in your users will be the most useful to give you feedback. Focus on those first. So get it out. So talk to them, show them, refine it and test security. I've seen people get fired because they didn't test security. They push out dashboards and information went to the wrong people. And sometimes it's critical information. So be sure that before you push the dashboards to production to test security, if that's something of a concern. The idea is every user will only receive the information they're entitled to. So that can be very critical. And finally push to production. You know, send the reports to people. And when you send to everyone, most likely there's still many adjustments required. You know, check with people, send emails right after, a week later. Usually the next few days or weeks is when you'll get everything correct, speed and whatnot. So monitor the usage, ask people's feedback, and I also recommend to use dashboard platforms that give you usage over time. So when you see the most used dashboards, go back, see if you can make them better. If some dashboards are not being used at all, either scratch them or see what's missing from making those dashboards more popular. Having a dashboard platform that helps you with all of these things is great. If you have a platform that can let users build even their dashboards, even better. The more closely to the business, the people building the dashboards, the better. So this is the ideal scenario. Can you do all the steps? Probably not, but try and you'll see that you'll be able to get the job done much more quickly, much more efficiently, and everyone will be happier and you'll have more time to sync to other things. All right, so these are some tips about building dashboards. Now, when building dashboards, you want to be sure they are informative that users will get it will get what the dashboard is saying. What I'm going to talk about now is what visualization types have proven to be helpful. I kind of make this question here. If you use a scattered plot to show some kind of visualization, we will you have time to teach Users how to interpret that scattered plot. If you use sophisticated visualization techniques, maybe you like it, maybe you love it, but some users don't understand it all, so it's missing the point. This page is talking about ways to show information in a very informative and easy to understand way. Bars. No brainer. Pretty much everyone can read bars and how they represent lines. Same thing. Areas. Very easy to understand. Labels. Putting labels on your visualizations can be tricky. You cannot put too much but sometimes at the beginning and at the end of the curves or some points where you see some important data sets. So use labels and put them in a clean, simple format so people will be able to quickly get the point that the labels are conveying. Absolutely very valuable and easy to understand. State indicators. Up, down, flat, traffic light kind of indicators. Very popular, very easy to understand. Tables. Some people say you should not put tables on dashboards. I think that in many cases is absolutely necessary. Tables can convey precision down to the penny if needed. So use tables as needed. So don't shy away if people are demanding some and actually some people have a hard time reading charts and they will need tables. And we'll talk more about after in a few slides. Maps. Very easy, very popular to show. Geography analysis. Pie charts. Pie charts are a little tricky. If you're showing up to two or three different slices, they are fairly different. They're absolutely very informative. They can show you trends over time. So up to two to three slices. I recommend as great visualization types more than three slices. It gets tricky and it's very hard to tell patterns and we'll talk more about that later. But pie charts, if well used, they can also be easy to understand and easy to convey messages or insights. The last one is what is called bullet charts. They have bars, they have colors, they have indicators to show kind of targets. Very useful, very efficient way to show actual version, budget or target. Very popular also and easy to understood. Once people get it at the beginning, sometimes they don't. You'll have to teach them. But boy, once the users get the concept, they are very insightful visualization techniques. Alrighty, let's move on. Number formatting. This is something that is interesting. I've seen so many situations where reports and dashboards were very hard to read. And after users get after the people get the idea of how to make it better, how much easier to read dashboards get. Let me show an extreme example. How about that? If I show this table quantity sold sales and percentage gross profit. Try to understand these in actual numbers. It's freaking hard. There's no formatting whatsoever. They're all big numbers. What exactly is going on here? Well, I'm just showing an example of an extreme. I'm going to try to go to the other extreme to show how much better the right formatting can convey the content of this visualization. How about that? Now much more quickly and a user can have an idea of sizes. There is the thousand separator. We're using K for 1000, M for millions. So you don't see a lot of numbers. They actually don't matter once you see the first two, three digits from a size standpoint, that's enough percentage, no brainer formatting, you have a much better perspective. So very easy to understand these numbers to get a hold of what's going on. The colors also can be very impressive. In this case, for this visualization we have a threshold where below certain percentage it's red, it's bad. Profitability is not good enough. So right now I can tell this line profitability is way below the other folks. So red in green or different colors can quickly get people attention to the right area. The other thing is overall, you have little real estate in a dashboard and you have to make good use of that to show data, to show insights. Well here, if you have noticed, I made this header smaller using smaller fonts. Why? Well, at the beginning, absolutely. I need to see quantity sold for this column. I need to see this is the sales and this is the percentage. But if I'm seeing this dashboard often, what happens is eventually I don't need to look at the headers. By looking at the numbers, they're telling me what they are. I can tell this is quantity. I can tell this is dollar amount. I can do this percentage. So. So I can kind of push to the background or kind of fade away or make them smaller the header and put more real state around the actual data. So those techniques can make something hard to read into something much easier to get to the point to get to the insight. So number formatting can be very impactful in conveying your messages or your data or your insights. So overall, you know, as you work with number formatting, you get better readability. Using relevant units, like in this case thousands, millions, makes it easier and making much better use of the space to convey the insight. How about date formatting? Very popular. Many reports and dashboards using dates. This example, just using a vanilla approach. This dashboard was an initial approach. A client of ours was trying to see their trends and how they're growing. Kind of Year over year in a three year period. So very simple dashboard, it kind of shows some of that. The dates, format kind of, you know, just what people usually start working with. Now again, when you have limited space, you might want to trim down information to focus more on your data. So one thing is just using abbreviated months. I love it. It's very easy to know what months we're talking about and it takes a lot less space than these. So I think this is usually my default approach for any date formatting. As a starting point that I need, you know, month names, you can go even further. Just use only one letter to represent the month. If this dashboard is seen very frequently, this one saves a lot of space and sometimes is quite enough. Or you can also use numbers instead. Maybe your months are not calendar and you have to go with the numbers instead. Or maybe on a calendar basis people like better numbers than than the first letter. Just personal preferences. You actually have to see what your users like. Overall it's not about you, it's about people reading the dashboard. So try different approaches and see which ones works the most. These two work great for non recurring, you know, if people are not seeing these dashboards quite often or not used to these other two standards, these two work really well. This one I think is the best. Otherwise these two will save a lot of space. So these are the four main types for exactly months, let's say. And overall the main frequent. The main factors contributing your choice of which one to use is how often people see this dashboard and will they be remembering these different two standards and the available space you have in the dashboard to use one or the other. My preference is this one, but again sometimes these are a better use. I'm just showing you here the evolution of that particular chart that I just showed you. So this client was trying to get a hold of trends over time in a three week, three year window, see what's going on year over year. So eventually this was a dashboard that they got to. The top part shows the same as before but using different colors to tell the trends over time. We also added a trend line showing how things are going up. So this top part gives a three year perspective how things are evolving over time. The next two are linked together. They have a single month legend here. So using the bands to tell exactly how each month is happening year over year, the same color theme. So this chart here, very clean, very simple and it shows overall how they're growing from a trend analysis in the course of a year. Very visual, very insightful. To compare Year over year. And the bottom one is very detailed for every month. It has three bars, same colors and showing numbers. So this dashboard, you can give a big picture overall three years, this one, year over year and this one detail information showing how much we're growing in those three years. So just showing you how something simple like this, hard to get some of the insights, can over time become a clean, simple dashboard. I can confess that first when you look at this dashboard, it can be a little bit overwhelming. Oh my God, so many stuff here. But once you see and you understand the pieces and if you see this every week or every month, you'll get it. You'll get all the insight that you want from making this kind of, you know, using this kind of techniques. Let's talk a little bit about interactivity of, of dashboards. Interactivity can make a big change for the right type of dashboards. Typically dashboards that you need people to change things like dashboards for data analysis. If you have dashboards for storytelling, sometimes you want no interaction whatsoever, no interactivity whatsoever. So. So use it wisely. It can be confusing. People click things and they change and people don't know what's going on. But for those dashboards that people need to change things, need to drill down, slice and dice and whatnot. Absolutely. It's very important, very insightful and very helpful. So what kind of interactivity you can give on a dashboard? Several options. Let's say pop up windows. You know, sometimes you want to just give the user some context behind a certain data point. So a little pop up window could, you know, put the mouse over pop up window shows. Now really keep in mind that we have a very short term memory, meaning, you know, usually we cannot remember things for too long and we cannot remember too much, much information. People say that we can hold in our memory three to nine chunks of information, not for so long. So overall, if the information is important, do not put on pop up windows. Actually put on the main dashboard, put more complementary auxiliary information in pop up windows. Let's see an example of pop up windows here. So where's my mouse? Oopsie, sorry about that. Not on this slide yet. I'm going to a data self analytics portal just to show an example. It could be any analytics tool. Let me show this particular dashboard. It's a financial report and this dashboard here has information about different locations, HR overhead, revenues, bars and whatnot. And maybe what you want to do is when the user puts the mouse over a bar, it has this pop up window that shows more details for this bar, in this example it has all the metrics, all the KPIs or metrics associated to Houston GL, segment account 2 and whatnot. Now if I move to the other bar and I see for the other bars, if you want to compare between those two bars, those metrics, either you use the actual bar size, but if you look at the pop up, it's very hard to remember, okay, revenues, how much on this one, that one is hard. We don't remember things really well. So the pop up, once they move out, you probably cannot remember the numbers you saw before. And in this example, this dashboard, that information was very critical. So the initial dashboard only has this big chart with all these numbers in the pop up. Eventually for this client we added at the bottom of the dashboard a table that has the same information. You see the pop ups is also down on the table with a more granular information. So still many users, they're just looking at the bars and look at the numbers to the pop up. Some users get the actual information at the bottom and the two together did the job for all users. So keep in mind pop up, Very, very cool, very insightful. To give some extra context, sometimes you have to bring it out because that information can be critical. You need to be sure that people will retain it in their memory as they do their jobs. So that's an example of pop up windows and how you can address them in a better way. Filters only add filters that are required in a given dashboard. If you put too many filters, it can get cluttered. Sometimes filters can help performance, sometimes they can deteriorate performance. So you got to be also using filters wisely. One example of filter that was causing problems for a client of ours was when we started the system, they didn't have a lot of data yet in the system and one of the dashboards was looking at, I think was invoice number. So looking some trends, the user could add an invoice number to filter the whole dashboard specifically for that transaction. Well, as the system start to capture millions and millions and millions of invoice numbers, that particular filter was taking a long time just to load to the user interface. So eventually we have to redesign the dashboard split into two in a way that people could drill down from a big picture dashboard into a more detailed dashboard, like for one specific customer or salesforce whatnot. And then the number of invoice numbers for that particular more smaller data set was much smaller and the whole report and dashboard would run quickly. So filters very useful. Let me show an Example of using filters in a good way. Here let me open another dashboard, this dashboard which shows pipeline. So in this case very simple. Dashboard has pipelined from the CRM system by account manager. I have the account managers, each one has a different bar with how much is in pipelines and the details of the opportunities. And we have two filters here showing opportunity status and opportunity stage. Very easy to interact in this dashboard, very clean. Let's say I only want to see open opportunities, so I remove closed loss in one. Now I only see open opportunities, let's say I don't want to see them all. I only want to see the ones that are in the decision process because as the manager of my account managers I want to be sure we close them all. Then I have here only the account managers doing open opportunities. And I see, hey, ad has a lot. Let's click on add. And now the dashboard at the bottom only shows there's only one opportunity for ad. Big a deal. I want to be sure to be involved to push this forward. If I should check the other guys, you know, some of them have more opportunities. So one example where filters give very good interaction allow people to go through this decision making process in a very efficient and focused way. Now one thing that is also critical is when you set up filters in your dashboard, it must be clear that the dashboard is being filtered by those parameters. Sometimes if it's not clearly visible what you're filtering by your dashboard, people might not know that a dashboard has been filtered and make incorrect decisions. So like just this example that we were looking before, as you see the filters are clearly posted on the right. So if the user is looking at the whole dashboard they can easily tell that. They can easily tell that this information is only showing for open and just for decision. And if you look at the dashboard more detail, we're going to see that here only is open in the only decision. So those two informations are redundant. Being sure the users are considering all those components in their assessment. Now I also want to see some other interaction possibilities. I'm going to go back to this dashboard here, financials, KPIs and sometimes when I allow users to only view a dashboard like this one, there's no filters. People cannot change. It's completely kind of static refresh, but it's static. But sometimes you want to allow people to to do some kind of drill down, slice and dice of the data. So I'm going to use this example to show some of that. When this dashboard was built, it was designed so people could use Here to this segment two, click the expand button and they drill down to segment three. So by just clicking on that expand the dashboard allowed a specific drill down a pre configured drill down direction. So I can expand into the segment three. So you can design dashboards to have pre configured drill down capabilities. Another option that we have in this example is when a user puts the mouse over these bars, they can click on a bar and you might design the dashboard. When you click on a bar to go one level down, one drill down option automatically click and go. Or in this case if I click on a bar I have options to drill down into. So these are also pre configured drill down options that I could go into. This case Balance details, monthly details, we're all looking at financial information and even in this example go into the CRM opportunities. Let's say for Atlanta when I click here I could go into a completely different report from a completely different data set. So dashboard sometimes is very useful to provide that kind of flexibility. Let me choose one of those options. I'm going to go to balance details now I see a completely different report still from the Atlanta office. And then users could continue to change the information. Now as I showed before, you can design things to be preconfigured to get drill down and non sophisticated users just go into it Easy to use. Dashboard tools should also give you ways to give power users more control so they can completely change the dashboard. They can completely create new ones from scratch. I'm just going to go and show an example very quickly. So in this case if I'm a power user, not only I navigate in the pre configured drill down, but I can click the edit button, I can go into an edit mode, select any of these dashboard areas and then just drag and drop. Let's say if I want to add segment three to my dashboard, I just drag it and put let's say between segment one and segment two. So right now I put segment three right there. Maybe I'm going to put after segment one. So now it's put on the bottom here at the end of here. So giving users the ability to extensively change dashboards can be very valu. And having the right tool will let your users become very self sufficient and very independent in their decision making process. All right, that's pretty much a little bit of interactivity in dashboards. Let's talk about data ink ratio. This is a very interesting concept. Edward Tufte, I think that's how you say his last name. He came up with the idea where most of the ink In a dashboard should be focused on showing your data, showing your insights that you want to get out of it. If you use ink to anything else, it becomes distracting. It's a very interesting concept. Let me show what I mean by that, what Edward means by that. If you look at this chart here, I found it's beautiful. It has nice colors. It is 3D things at an angle. I have the legend that I can associate with bar personally. I think it's designed by an artist. It's beautiful, it's visually appealing. Now, is it really easy to get insights from this dashboard, this chart? Well, according to Edward, the data ink ratio, it's not good. There's a lot of colors, not necessarily for the insight. Why do we have black? Why do we have 3Ds, all distracting? Maybe you're just wondering all the beauty. It's not about the beauty, it's about the insight. Let me show an example of a very good imbalanced data ink ratio for the same information. Trying to show sales by salesperson. How about that? Very clean, very simple. Every bar has its own salesperson name. So you don't have to be relating to which one is what. The blue color, very neutral. Most people will find it. Okay, each bar has already the dollar amount. So I know where they are. I don't need to have an axis to tell that there's no 3D, there's no additional color. Pretty much the only ink used in this dashboard is to tell how much each salesperson is doing from a sales standpoint. Now, one funny story that I have from this kind of data ink ratio was one of our clients. The person who built the chart used colors like this one. Different people, different colors. And one of the bars was pink. It was associated with a male salesperson. Guess what happened? That particular guy did not like to have a pink bar. And actually there was some confusion and some discussion around it. It's funny, it's silly, but hey, it's human nature. You want to be sure that doesn't happen. Why would you use colors if it's in a chart, if they're not representing, if they have no value? You know, it's kind of funny, but it happens. So if you don't use color like in this example, it's less distraction, less confusion, less waste of time with unrelated topics. So there's an example where data integration really well used. So in a nutshell, you want to have a non distracting dashboard, you want to focus on insight and that's it. Now I'm going to show A little bit of an opposite situation here. This other example here is a client that want to show for their SalesPeople2 charts. One of them showing sales by salesperson alphabetical order and one from a performance order, meaning top to bottom. So salespeople could, you know, see how much they're making and where is their, you know, total amount. Now they did a great job getting the data. I mean, a lot of, you know, very little data, I'm sorry, very little ink to represent this data. But actually it's not enough because let's say if I'm Henry, right, which is right here, it's very easy to find Henry in this list because you go alphabetically. Where is Henry here? You know, you gotta go until you find. I mean, it takes several seconds to find Henry on the bottom. So this doesn't have enough ink. It's not helping users to really find their own bars in both charts. So when you have that kind of situation, sometimes you have to use colors or shapes to help users more quickly find their trends. Here's an example of what this new dashboard became eventually. So they did add colors. And now if you're Henry and once you see this chart once, you'll know next time that your color is green. Very quick. It takes a fraction of a second for you to find where your bar is on both charts. And let's say in this case Rick didn't like pink. Well, it's worth the fight. In this example, Rick comes, didn't like the color, you go to your dashboard, you change Rick to a different color, and that's it. The point is, adding colors and shapes in situations where you have more than one chart is usually really helpful to let users quickly find their data points, make a decision more quickly and move on. Cool, huh? Another example where sometimes you got to put more ink. Also a very clean chart. For every. I'm not sure what this is, customer or product, you have a bar. Oh, it's pipeline. So there are reps as well, sorted in ascending order. Now, how much is this one? You know, you can get lost here and not find exactly which one of these three is for this one. Well, you can use bands or lines, so you can very quickly and truly know. Okay, for this particular salesperson, this is the amount. This is my bar. So the data ink ratio is the balanced amount. Not the least ink is just the balanced amount. So visualization, far from labels and data points, using bands and lines can more quickly help you get people have the right insights. How about text and charts? As I mentioned Some people don't like to put texts in tables on charts on dashboards. I say absolutely do it when they make sense. This is an example where the initial chart was very insightful but not so helpful in a way. For this client, they're looking at the top customers by sales growth and sales decline in a given period. These are growing the most and this is how much they're growing since last year to date. These ones are declined the most since last year to date. And the bar is the amount of decline. At the bottom they have the customer names. Great idea, Great insight. Actually we have out of the box in our reports and dashboards, but kind of hard to read this dashboard or this insight. Eventually we actually went into this approach. So horizontal instead of vertical. Once you know that blue is growth and red is declined, it doesn't matter if it's vertical or horizontal, people will get it. Bigger blue better, bigger red no good. And we also added the labels for the amount of sales for each bar. So not only you have an idea of how much it is, you have the precision if you want to see the numbers. Now in this example, we also found that many of their users, they didn't like much of the numbers to be in this kind of. You got to go. And the format was not very helpful for some of them. So eventually this dashboard became this. So it's an actual table. Straight lines just go down the list to see and see the numbers. So people who are more numbers driven. This table was really helpful. Blue and red, the same colors. And on the right for people who are more chart driven, they like. So you get both of the crowds happy and anyone in between found this dashboard a lot more insightful. So main factors here, readability and user preferences. Trying to get as many of your users happy with the way that you're presenting the data. When to use dual axis charts. This is an example. So we have sales on this bottom and at the top we have website visitors. And the client was trying to see how website campaigns were affecting overall sales trends. Because both of the charts are not really together, it's a little hard to see. Okay, these two campaigns eventually derive some sales. It takes a lot of effort to relate the bars and the trends. But when you put on a dual axis. So you're using the same real estate, exactly the same. But now we have website visitors on the left with the orange bars and sales on the right with the blue bars. It's so much easier to make the correlation between the trends. So dual X can be very helpful for that now notice that in this particular example, not only we have the legend here with the colors, but even the color of the axis use the same standard. So more quickly your eyes, you want to see sales, you go to the blue which is here, you want to see the axis for the website visitors, you go to the left. So using colors on the axis can also be very helpful for more quickly help users to get the right insights from any dashboard. Specifically even more from dual axis charts. So dual axis is very useful for interrelated metrics and make readability better role focused dashboards. This one was an interesting story. The CEO, she asked, she doesn't like charts so she want numbers simple, big on her mobile device. So this dashboard was designed looking four main areas of the business. Sales, cost on hand, inventory, profitability and AR past due with actual target and variances of course as being bi people. This dashboard was designed with colors. So red, not good. Yellow, kind of, you know, pay attention. Green, you're doing good. The dashboard, you know, I mean the CEO loved having these KPIs on her cell phone. But the funny story is that accidentally the person who maintains the dashboard changed the color themes of the dashboards in this red, yellow and green became a different set of colors. And a few weeks later the seal said wow guys, that was awesome, I love the new colors. And then was like what happened? And make a long story short, she's actually colorblind and the red, yellow and green were not connecting with her. And after this color was accidentally changed, she actually said that the colors really help her get a much more quickly insight into what she need to give attention. And there was a gold, the red, yellow and green. If you don't know, about 10% of the male population is colorblind and if I'm not mistaken, 5% of the female population is also colorblind. So when you have your users and you're rolling out dashboards, you might want to consider that be sure user preferences and needs. And sometimes you should not use green and red yellow because they might not work well with some of your users. Another way of making dashboards really helpful is by putting them in the workflow of people. Let's say salespeople are using CRM all the time. Well, build the right dashboard in the right page of your CRM system. So as people go through this CRM or erp, whatever this is they're using, they'll find dashboards that are contextual to help them make better decisions for that particular step of their workflow. So make it Easy the access so they don't have to go to a different application and make them get better insights very quickly. Easy access and speed is the idea when you put dashboards in the workflow. Now what some of the visualization that you should avoid. Overall, I mentioned earlier that pie charts are usually very insightful. If you have two to three slices. Well, if you have more than two to three slices, sometimes you cannot really get any story from it. Like this example. Very good article from the Business Insider. If you look at this three pie charts, they were sampled in three times in the course of a year to see the trends between these different, you know, 1, 2, 3, 4 and 5 segments. Colors well used, easy you know. But overall, can you tell any trend here what happened? You know, in the course of the year? You pretty much can tell they're about all the same. Nothing happened much. You cannot read much. Now how about this? It's exactly the same information, but you can tell at the beginning of the year this was the trend going up between these five different entities. What happened at the end of the year it completely reversed. Try to get the same insight from the pie charts. Unless you are gifted, it's just impossible. So when you have again more than two to three slices, pie charts can become almost useless. So useless for comparison. Similar data points, especially over time. 3D charts also not very useful. Confusing, sometimes even deceiving. In 3D pie charts can be very deceiving. They can tell you a lie. Actually, like in this example also from the same article Business Insider we have here S and D and Epp. They're both the biggest chunks from the pie chart. So if you look at this, these two slices, which one is bigger? They look about the same size. Don't tape. Now how about this? It's exactly the same information, but in this case if you look, The EPP is almost 55, 0 50% bigger than SMD. Try to get that insight from the pie chart. So 3D pie charts can actually tell you lies. They can be deceiving. So this is a business, you don't want to do that. Or if you want to do that for some reason, you know, be sure you know that 3D pie charts can give you such an incorrect insight. Overall, we tell no 3D charts, especially no 3D pie charts in your visualizations. Here's some suggested reading. Very good content for dashboard design. Feel free to assess them and that's it. If you have any questions, comments, feedback about this webinar, please give me a call. You have my cell phone here, my email and my Twitter handle. Thank you very much and enjoy your day.