 And look, my name is Shannon Kemp and I'm the Chief Digital Manager of DataVercity. We'd like to thank you for joining this DataVercity webinar, Building Effective Data Visualizations for Business Intelligence, sponsored by Adira. Just a couple of points to get us started. Due to the large number of people that attend these sessions, you will be muted during the webinar. For questions, we will be collecting them via the Q&A in the bottom right-hand corner of your screen. Or if you'd like to tweet, we encourage you to share your questions via Twitter using hashtag DataVercity. As always, we will send a follow-up email within two business days, containing links to the slides, the recording of the session, and additional information requested throughout the webinar. Now, let me introduce to you our speaker for today, Stan Geiger. Stan is the Director of Product Management at Adira with over 25 years using Microsoft SQL Server. Stan has worked in various industries from fraud detection to healthcare. He has held several positions, including data-based developer, DBA and BI architect, and has experienced building data warehouse and ETL platforms, BI analytics, and OLTP systems. And with that, I will give the floor to Stan to get today's webinar started. Stan, hello, and welcome. Hello, Shannon. Thanks for the introduction. So I guess we'll get started. And since Shannon already talked about me, we'll skip this slide, but just kind of, well, we'll still in this slide, just to kind of give you a little bit of a background. I actually started my career a long time ago, probably 30-some-odd years ago, and I originally started doing what we used to call decision support systems. And those things have morphed into... And those things have morphed into things like business intelligence, data analytics, you know, business analytics, the things we hear today. So, you know, a lot of the same concepts have been going on for a long time. So the first thing I want to talk about is let's talk about business intelligence, because obviously the title of the presentation was Building Effective Data Visualizations for Business Intelligence. And I kind of like this quote that I found. It says, business intelligence is a technology-driven process for analyzing data and presenting actionable information to help executives, managers, and other corporate end-users make informed business decisions. The thing I liked about it was it's a technology-driven process and, you know, it's... And presenting actionable information, not just any information, but actionable information that helps people make decisions, helps people make decisions on how the business is doing, where the business is going, you know, where the business needs to be going based on corporate and, you know, initiatives. So just to expound on that a little bit, you know, what is business intelligence, you know, it provides historical, you know, current and predictive use of business operations. So oftentimes the business intelligence, we look at what happened in the past, what's happening now, and then try and predict trends and things like... Tends around information gathered through the business on where the business is going, and then effectively make decisions based on the objectives within the company. Helps identify, develop, and create new strategic business opportunities, provide businesses with competitive market advantage. You know, we've heard for a long time that data now is the strategic, you know, asset of the business, you know. We collect so much data and information now, how we can cognizantly present that information to people who need to make decisions about the business has become a huge competitive advantage for companies. You know, common functions of business intelligence technologies, you know, include things like reporting, online analytical processing or OLAP, data mining, process mining, event processing, business performance management, benchmarking, text mining, all of the buzzwords that we hear now. And the thing about business intelligence, it can be used by enterprises to support a wide range of business decisions. And those decisions range from the operational to the strategic. You know, you can think of operational decisions as things like, you know, product positioning, pricing, you know, whether to ramp up manufacturing, things like that, where strategic decisions involve priorities and goals and directions at a broad level in an organization. But the key thing is when we combine the external and internal data, it can provide a complete picture of the business, you know, and that's where the intelligence gets created. So second thing, I want to talk about business analytics because we often hear these things talk to interchangeably, but they're not really interchangeable. Business analytics uses more statistical and quantitative tools for predictive modeling. Things like, you know, you hear things like principal component analysis and things like that, machine learning, AI, those are all part of analytics. But that's kind of a predictive. So we think of business analytics as more of the predictive side of business intelligence. You know, as I say here, business analytics is a subset of BI. You know, it focuses on the statistics prediction and optimization rather than reporting functionality. Meaning that, you know, BI in a broad term reports on what the business, you know, history, where they are currently, and then we can take that information and try and predict using business analytics where the company may be going in the future. So I kind of like this comparison here. The side-by-side comparison is the orientation. We can think of BI as kind of backward looking, looking in the review mirror, and analytics as looking at the future. And so what types of questions with, you know, are we trying to, do we typically answer what with BI, you know, what happened, when, who, how many, things like that. Advanced analytics, you know, as you can imagine, what will happen, and what will happen if we change this, you know, what's next. So if we stop doing this or increase doing this, what do we, where can we go from here? And you know, I'm not going to go into a lot of detail, you can look at some of the methods. Typically in BI we see reporting with KPIs and metrics, dashboards, scorecards. You see OLAP, you know, you know, analytical processes, things like cubes, slice and dice, stuff like that. But then the buzzwords you typically hear with analytics are predictive modeling, data mining, text mining, multimedia mining, lock mining, descriptive modeling, statistical quantitative analysis, things like that. You know, both BI and analytics support big data, of course. Data types, both structured and unstructured. Knowledge generation, this is kind of interesting, is that, you know, business intelligence is a more manual process. Whereas now with machine learning and artificial intelligence and algorithmic programs and things like that, you get a more advanced analytics really covers, there's a more automated scope to that. So the last thing is users, you know, you tend to see a lot of business users, business analysts in BI, management, things like that. With analytics you start hearing, you know, about data scientists, business analysts, you know, you have people in IT that are doing algorithms and machine learning and AI and, you know, business users form a component of that. Typically, you know, your data scientists and those types, the more mathematical types, work on the predictive modeling side. So, you know, let's talk about applying business intelligence. You know, when we talk about applying business intelligence, we see things like performance metrics and benchmarking. You know, we use those, you know, to inform business leaders of progress towards whatever the goals of the organization are. Analytics quantify processes for the business to arrive at optimal decisions. Analytics, you know, involve all of those buzzwords that I mentioned before. Business reporting can use BI to inform strategy, meaning business reporting often involves data visualization. And that's kind of what we're going to talk about here. So, we see a lot of data visualizations and we'll talk about why as opposed to, you know, these number crunching reports. You know, and that information is often presented in things like executive information systems, which that term has been around, decision support. We see this information also presented through online, you know, OLAP type, things like cubes and things like that with some kind of a visual or a reporting front end. But BI can facilitate collaboration both inside and outside the business. So, the way business intelligence can do that is by enabling data sharing. So, basically, both internal and external, we can see, you know, through visualization with partners, business partners and things like that, kind of where we're going and what we need to do and how the partnerships are doing. So, let's talk about data visualization now. So, let's talk about data visualization in the context of business intelligence. Now that I've kind of set the groundwork for business intelligence, how does data visualization fit into that? And, you know, simply you can think of it at a 50,000 foot level is that a picture is worth a thousand words. And that's key here because our brains are wired for visuals. And I found some of these stats, which I thought were interesting. It says, half our brain is dedicated to visual functions. So, you know, if you really think about it, humans have been visualizing data for hundreds of years, from maps to graphs to charts. We've been taking data and arranging it, so it tells a story. And we arrange it in such a way that it tells a story better and in more depth than the data could alone by itself. So, you know, if we think of it that way and we think of it in that context, it kind of helps us to put that context around why data visualization is important and why proper data visualization is key. So, you know, 90% of the information transmitted to the brain is visual. We process images 60,000 times faster than text. And that's important too because if you sit down and put a report in front of somebody with a whole bunch of numbers, they have to spend a, you know, considerable or a good amount of time trying to get the, you've ever heard of the term, I'm trying to get my head around this. Well, that's because the brain is trying to process the numbers and crunch and figure out what the context around all of that is. And because of that, you know, it's important to get key concepts across to people you're trying to present to by using visualization. Perception involves subconscious processing of visuals, meaning perception or context is what our brain sees when it looks at a visual. So, things like position, color, alignment, size and shape can be processed subconsciously, excuse me. And that's important because that's why building effective data visualizations is a key. Just putting a visualization out there or a chart or a graph or a pie chart or anything like that in front of somebody doesn't mean that it's necessarily understandable. You know, things like we mentioned here, position, color, alignment, size and shape. Because they're processed subconsciously, it's important to do it in such a way that you convey the meaning that you're trying to get across to the audience that you're presenting to. You know, it is important to create the right perception. You know, in plain English, perception is the process by which the brain sees the information in a dashboard or a set of visuals or something like that. So, let's move on to the next slide. So, let's talk about BI and data visualization in the context of perception. So, basically what you're trying to answer, the question you're trying to answer is, how do I communicate this data in a way that can be understood by everyone who looks at this information? And it depends on the audience. You know, we're going to talk about that later. So, once data is made interpretable, in other words, once I have a context around that through the visualization, you know, the get value spark action moment occurs in your audience. So, they either get it and it spurs them to act on the information or at least understand what's going on, you know. So, you're really looking at how do I communicate this data in a way that can be understood and business, you know, your audience is typically business leaders and they need the ability to easily get that context. Now, that doesn't necessarily mean you present everything in front of them. You know, we'll talk about this a little bit later, but you know, you may do things where you give them an overall context and provide drill-down capability, for example. You know, oftentimes the business needs the ability to easily drill down into the data to see where they can improve operational processes and grow their business. So, I can present charts in such a way that, oh, yeah, I can see the trend going on here, but if I'm an operational manager, I may want to drill into that information. So, oftentimes we create interactive visualizations where I can click, drill down, and then get the detail behind that. But I don't want to present everything on, you know, in this one big complex visualization or dashboard because, you know, while two people out of 100 may get value out of that, it may confuse the other 98% or 98 people out of 100 and then you kind of defeat the purpose. So, you know, when we talk about visualizations, they can be everything from pie charts to area maps to bar graphs to gauge charts, you know, whatever. But the goal is to make the content attractive and informative. So, a fitting chart speaks better than anything. So, using the right chart or graphic, et cetera, is key. Because the different ways that you present the same information can answer different questions. You know, the other thing to keep in mind is the comprehensive and interactive view is rarely attainable. So, quickly by approaching raw data without using visualization software, it kind of defeats the entire purpose of what you're trying to accomplish. So, let's get into an example. I put this example up here and I apologize if it's kind of, the text is kind of small, but visually it kind of gets across kind of what I'm talking about just in a simple example. I presented the same information in two ways. This is nothing more than sales by territory and I presented it in a table view and I presented it in a chart view, basically a bar chart. So, if I'm presented this information and say my perspective is I want to know the territory with the highest sales amount, it takes me a few seconds to read that chart. I've got to go through there and then my brain has to stack right the sales and then associate it with the territory. So, I have to figure out in my head which number is the highest in that chart and then my brain has to look at the territory. But if you look at the bar chart, it readily displays the desired information because my brain immediately goes to the highest bar and then I can go down to the bottom and see the territory. And I don't know if that was 60,000 times faster as the stat said than I did with the chart. But you can see from the simple example why visualizations become important here depending on what questions you're trying to answer. For me, the bar chart immediately drew my eye to the highest bar and gave me the answer to what I needed or maybe I was looking for the lowest. But my brain immediately go to the smallest bar and immediately to the sales territory whereas with the chart and numbers, your brain has to process those and figure out how to rank those and then get the information out of there. So, let's talk about best practices which is kind of the key to this whole presentation. And, you know, the key thing about best practices is you need to have a methodology, a method to the madness if you want to call it. And the methodology that I highlighted here is based on the fact that the more pre-intensive processing, you know, we talked about how the brain works we can bring to our visualizations the more intuitive they become. Most good data visualizations are built. All of this is just to say that there are best practices that if followed will aid in building intuitive dashboards and visualizations. So, I look at this as kind of a three-step process is conceptualize, visualize, finalize. In the conceptual phase, you identify your audience. You know, what's the main questions you're trying to answer for that audience? What's the theme for that audience? Identify themes. You might have various ones, test them out, figure out, you know, which ones fit for the audience. And then the next thing is to visualize. You know, what are your primary views to support that audience? How, you know, what views support whatever themes that you've decided on for the audience? And then identify layouts. You know, start spitballing layouts. How do I, how can I lay out, you know, if we use dashboard, if we're doing a dashboard, for example, how do I lay out the charts and graphics on this dashboard so that they fit with the theme and fill the primary view that I'm trying to accomplish? And the last stage is to finalize that. And we do this through prototyping. And we'll talk about prototyping later. But, you know, then we revise as needed, you know, maybe testing with our audience and then come up with a final revision and then we present that. So this method, you know, basically the methodology is very simple. And that's why I like it. You conceptualize what you're trying to do. Then you create the visual, then you start visualizing how you want to present it. And then you finalize that and then present after that. And then basically, because it's a simple methodology and it follows the old adage without a plan, we have a plan to fail. You know, and I've been involved in and I have seen way too many projects kind of go over budget or fail because an adequate methodology was not followed. In other words, you know, I got involved in rescuing a couple of projects, you know, for dashboards and executive reports and things like that where another area had worked on it and they had put these visualizations together. And it was obvious that they created these visualizations because they thought they looked good from not because they presented the concept or the answered questions, but they looked good. You know, things like 3D bar charts with, you know, 25 bars on there and 3D pie charts and some things that didn't really fit what was trying to be conveyed. So it was obviously that, you know, there were people's biases that came into, you know, these visualizations and not really understanding what the purpose was. You know, and I liken it to basically just kind of slinging mud at the wall and then you go, wow, this looks really cool. I mean, how many times have you guys seen, you know, dashboards that had like five or six charts on there and it looked really cool when you first looked at it but then when you started looking at the information you weren't really quite sure what was trying to be conveyed and this goes back to the perception. So that's why it's important to have a methodology because if you have a methodology and you work through it then at least you have a plan not to fail. It doesn't mean you're going to have some, you know, roadblocks that you need to overcome but at least you've got a plan. So let's start out with the concept of knowing your audience. So knowing your audience means, you know, understand who the primary uses for your data visualizations, charts, et cetera are going to be. You know, what are they interested in, what problems are they looking to solve, things like that and then figure out how they were perceived the data and what we mean by that is not everyone sees the same visualization in the same way. For example, you know, a product line manager in a company, for example, is going to see profitability, for example, in a different way than a chief financial officer. So you need to know your audience. If your audience is the chief financial officer then you're going to show profitability in a way that he perceives proper profitability and at a level that is complementary to the decisions that the chief financial officer typically tends to make. Whereas a product line manager has a much narrower focus on the data so you're probably going to present the data in a much more granular fashion. So, you know, there's an example of how you need to identify and know your audience and how they perceive the data. You know, their perception based on the audience also determines what types of charting, not just how but what types of charts. And like I've got here, your audience should influence how you visualize your data, understanding how your audience best digest data will influence which chart types you will use, when pie charts are appropriate, when they're not appropriate for a different audience, when bar charts, things like that. Let's talk about defining resulting actions as part of this. You know, identify what actions your audience might take after viewing your dashboards. I find this is one of the things that people leave out. It's like now you've given your audience this information. You need to think about what actions they might take based on the data that you've presented them. Because, you know, for example, let me give you an example. What action would you expect a sales manager to take after viewing information that shows sales in a particular region was down based on projections? Well, you start going through your mind. He might want to see, okay, sales is data in this region. And let me look at sales by sales rep. I'd like to know who the low performing sales reps are. Or I might want to know what marketing programs. How much did we spend on marketing in this region and by what? And that's where, you know, drill down capability comes in. So you may want to present a drill down capability. You can see here, if I hover over, I can see that sales number. And I can see sales was a certain amount, but I can see profit was a certain amount. And then it says click to see sales per city. So I can look and see, you know, in the European region, for example, let me see what sales were by city. And I can see that, you know, a couple of cities in that region maybe are really low. And I need to go and figure out why that is compared to the other cities. But you can see, you know, whoever's viewing this data based on, you know, what their position and their perception is, the questions in their mind that they would like to answer. And that's why you don't present all the information up front. Because you need to present the capability to drill into that data based on what questions that the audience would come up. And like I said, this is one of the things that people often don't think about. It's like, I give you this information and then I'm done. And then I move on. But you really need to think about what questions is your audience going to come up with, what actions might they want to see when they're presented this information. And then, you know, that might lead to follow-on charts and visualizations based on those questions. So another thing to talk about, you know, when you're talking about themes and primary views and things like that is to classify your dashboards. So if you're doing dashboards, you can think of dashboards in three different ways. There are three common dashboards. Typically you've got your operational, your strategic slash executive, and then your analytical dashboard. Now we can think about operational dashboards as focusing on things like performance monitoring and measurement. And they're typically easier to deploy than analytical dashboards because they require less data to create and less training to use. This is basically a snapshot of the performance of the business at an operational level. When we talk about executive dashboards, you know, these are, this information, you know, is typically used by managers to get a big picture view of the organization. So it's a higher level view of the organization, you know, using critical metrics that you've identified, you know. And it's often used to identify opportunities for expansion and where improvements are needed. So it's a much higher level. We typically arrange numbers and KPIs and performance scorecards on a single screen. So you've got a single pane of glass that typically is tailored for C-level executives and management. So it's a much higher level presentation of information and visualizations. Now analytic dashboards, you know, tend to allow you, give you more insight in your historical present and your predictive data. So basically past, present, and future from an analytic standpoint, you know. And things on analytical dashboards are typically used to kind of back up and validate your current strategy and determine what adjustments are needed to be made in the future. So they tend to be a lot more number oriented, number crunching, lying graphs, things like that. Sometimes you'll see bubble charts and things on there. But, you know, a lot different from the other two types of dashboards. They're also not as common because the data featured on analytical dashboards tends to be more complex. And therefore they typically require more training to use. They're generally used by business analysts instead of widely deployed, you know, across departments or management levels. But the key thing is if you know your audience, you know which classification your visualizations fall into. And then you can design accordingly. I mean, obviously you don't want to give executives an analytical dashboard because they'll probably revolt on you. But they'll have no clue on how to interpret that data because that's not what they're looking for. And that's in the perception. They won't get the perception that you're looking for either. So let's talk about another area that's important when you're designing your visualizations. You know, you need to profile your data. You know, what data do you have access to? And this is important because it's used to determine which types of charts you'll use for that data. And you can classify your data in three categories, basically categorical, ordinal, and quantitative. With categorical data, it's data that logically belongs together. So, you know, you might have sales regions like North America, Europe, and Asia, you know, states, things like that. That's what we refer to as categorical data. When we talk about ordinal data, ordinal data is data that logically belongs together in a logical sequence. First place, second place, third, months, January, February, March, you know, years, things like that. But ordinal data is typically in a sequence. Then we have quantitative data. Quantitative data defines how much of something. So, you know, when we look at sales charts, we'll see, you know, one million in sales. You know, I've got examples here, you know, 150 defects if we're looking at defects. Say we have a chart that's, we're trying to do a chart that's defects by, you know, product line. Well, the quantitative data is the defect data. Product line is, you know, categorical data. But the key thing about a chart that you're building there is that you're really looking at quantitative data. And the reason we do this is because different visual components work better for certain kinds of data. And if you don't learn anything else, realize that one thing. Because the most common, kind of the most common error that people make is using the wrong chart type for a particular type of data. You know, you know, some examples are scatter plots work well for two pieces of quantitative data, whereas line graphs work best with ordinal data. So, you know, ordinal data, you know, I use my defects per month. You know, I can have month, January, February, March, April, May, and then I can have a line graph that graphs the defects per month. So line charts work great for that. However, line charts are horrible for non-ordinal categorical data because they imply continuity. Meaning, if I take categorical data and I plot it on a line chart, it connects it and implies that there's a continuous, but it's a false premise. So you can see how using the wrong chart type would convey the wrong information. So that's kind of the second thing that I often see is that by using the wrong chart type, you're actually creating a perception that is not there. And you don't want to do that because you're not doing your audience any favors by doing that. So let's talk about some best practices. And this chart here that I put up here may not be as intuitive as you might think, but if you think about it, it kind of makes sense. You know, it tells you to use the proper visualizations. And across the top, we've got the categorical ordinance of quantitative data. And then going down from lowest to highest effectiveness, or actually highest effective, I'm sorry, from highest effectiveness to lowest, if we're using categorical data, position is one of the highest factors on that. And you can imagine that because we use sales territory again. So we position sales territory across the bottom with a bar chart. That's probably the highest, you know, impact for perception. And as we go down using different color use, you know, for different categories, shape, clusters, boundaries, you can see going down from highest effectiveness to lowest. For ordinal data, position is important, again, because it's very similar to categorical. With ordinal data, position is important. And then size, color, intensity, color use, and shapes as we go down. For quantitative data, you can see position is high. Length is also high for quantitative data, because length implies quantity in charts. Size implies quantity. But as we go down to, you know, color intensity and color use, for quantitative data, it's not as effective. So you just kind of use this chart as kind of a guideline on what's the most effective. So when we talk about iterative design, this falls into that last category that I had up there at Finalize. And this is kind of where agile comes in. If you're familiar with agile, which everybody should be at least at a cursory level, you know, what agile is. Basically, we're saying don't wait until your requirements are 100% complete to start putting your visualizations together. You know, visualizations require an understanding of the end user's desired needs. So it's difficult to achieve without starting with some sort of proof of concept and building off of it through feedback. So that's where the agile approach comes in. You build it, you prototype it, you go out there, you test it with your audience, and you do iterative design. You go back to the drawing board, you get your feedback, you go back to the drawing board, you tweak it, you come back and you say, you know, what information can this convey for you and just meet your needs? You know, and you just go through this until you get to a point where you get an effective set of visualizations or effective set of visualizations. So I find the agile methodology comes in handy for doing iterative design. So you plan, you design, then you build it, then you test it with your audience, test slash release it, you know, with agile, and then you get feedback, and then you go back and you plan again, meaning you go back and you do your iterative design, you go back and you tweak things. You know, if you wait till the very end and you go, okay, what do you think? Then, you know, a lot of times your audience comes back and you go, well, it's nice, but I'm not sure what I'm looking at, or, well, I really need to get this information here out of this. So it's very important to not get analysis paralysis, you know. And that often occurs, you know, when you're starting out. So just start. If you don't learn anything from this slide, just start somewhere and build off of that, or, you know, scrap it and start again. But at least start somewhere without fleshing out, you know, this huge monolithic, you know, application with visualizations. So I'm going to look at, we're going to look at some examples here, and then I'll leave some time for questions. All the examples, you know, here we have a product called Aquadata Studio from a company that we own called Aquafold, and you can go to www.aquafold.com and get a two-week trial that's fully functional. But I built all the examples in that, and the Aquadata Studio product is kind of unique in that it is not only a database development tool, but it has a query and analysis tool built into it with visual query builder, visual analytics, and data modeling as part of it. Which, you know, all of those fit into the visualization because you have to pull the data in order to build visualizations. So one of the things you need to do is you need to have data, right? So, you know, you can visually build queries in the tool, pull the data, then click on a button and go to the charting and start dragging and then dropping fields into the charting tool and building these visualizations. So anyway, just kind of a plug there, but all these visualizations were built there. This one I'm not going to go into again because I put it in here twice, but I put it in here at the beginning. But again, it just shows you the difference in the perception of how your brain works. But you can see here, you know, I built these visualizations within the product. So let's take a look at this one. Here's an example of what I call a confusing chart. It's basically profit by product category by state, but I've chosen to use a gauge here. And you can see that I might be able to determine the profit amount by category, but there's no way of knowing what state that it goes with. And I happen to know from the data that there's three states. One of them is in purple, one of them is in green, one of them is in orange on this chart. But it's counterintuitive. It really doesn't tell you that. And gauge charts are poor visual representation of this type of data. You know, you really should use something like a stacked bar chart, for example. So, you know, if this was a product demo, I would go back in here and I would take this chart out and I would go to like a stacked bar chart with the states in color. And then I could see, you know, the product categories, you know, as you can see listed over here on the left-hand side. But that would intuitively be more representative of what we're trying to convey here. You know, this is an example of somebody, you know, using, they may like gauge charts. I don't know. And they say, oh, I'm going to use a gauge chart because I'm familiar with those and I like them. But you can see here it doesn't convey what you're trying to get across. So this would be an example of a poor choice of a visualization here. Now, this is another example of a confusing chart. This is what's called a tree map. You know, it's kind of like a heat map, but it's not. It represents profit by shipping method by territory. And it has a very similar problem to the gauge chart. You can see I've got one, two, three, four territories, Australia, United Kingdom, and Germany. And if I'm trying to compare, you know, a shipping method by territory, it's not, you know, I can see overnight is in this, overnight shipping is in a particular color. But this is kind of like reading a table chart. My brain has to get wrapped around the way this is all split up here, for example. Again, this is kind of categorical data. So a bar chart would probably, some type of a bar chart, whether it's stacked or a bar chart split by maybe, I would split it out by shipping method by territory so I could see them side by side the bars. That would be much better as far as if I was trying to see which territory. Say I was trying to answer the question, which territory had the most overnight shipping. I can kind of get that here, but it would be so much more intuitive with a different chart type. So again, you know, be careful. Don't use chart types just because you think, you know, they're cool. So here's an example of an executive type dashboard. And what this shows is it shows trends over time. Executives are always, they tend to like to see trends over time. So you can see this is average profit by year. And then I've got another one that's raw profit by year. But what I can see here is I can see trends. So it immediately draws my eyes to where I can see the dips and valleys. And if I really want to know what the numbers are, they're on the chart. But from an executive standpoint, I can see from a trend that 2015, for example, I had a declining trend in average profit until, you know, month 10, which would be October. And then it shut up. And then I might want to ask a question, well, why did it shoot up all of a sudden? So, you know, this is just an example of where perceptually, you know, if I'm an executive, I'm concerned with trends. And my brain is immediately drawn to the slope of the chart, for example, so I can see where the ups and downs are. So this is an example of more of an operational dashboard where I've got, you know, sales by state and then I've got yearly sales by calendar month and then I've got units sold by customer. And this type of information is usually used by management. Basically, operational management. You know, in other words, people who make decisions on the day-to-day or month-to-month business. You can see sales by state, yearly sales totals. We went through that. This type of information is best displayed by using things like bar charts. And the reason being is because, like we saw in the other examples, it immediately draws your eyes to, you know, highs and lows. What you don't see in this chart was if you were to hover over these, it would give you the exact numbers that made up those values. So, you know, if I wanted to see a particular value, I could just hover over the bar and get more information out of that. And, you know, a lot of people, some people like maps. Other people don't. I mean, I like them depending on the information because this map shows sales by state. So immediately I know, you know, the darker the color, your brain tends to perceive the higher the value, which is typically the case. And the lighter the color, the lower the value. So you can see there's a graduated chart, kind of legend here on the right-hand side, which shows that these things are, that the higher values are the darker colors. But I like them just from a conceptual standpoint that you can look at them and immediately it draws your eye to specific areas that you're concerned with. So you can see Texas, California, and I think that's Pennsylvania and there somewhere, or whatever's below. I guess that's Massachusetts or New York. Anyway, but immediately you can draw your eye to that. And if you hover, I know when I built these, if you hover over a particular state, a box will come up and show you more details about that particular state. So again, it's a good way of presenting more data, but not making it confusing, so that I can visually see the states that are high. And if I want the extra information, I can hover over it. Thanks, New York, Pennsylvania, and Texas. So in other words, though, but this is something to keep in mind, you know, hover overs are great because you can design it to present more information to your audience if they want to see more information. They may not want to see more information in which case, you know, that's perfectly fine. So anyway, that's kind of the gist of the presentation there. In an overview, you know, some of the things, I'm not a particular fan of pie charts. Some people love pie charts. I think pie charts get overused, especially 3D pie charts. And the reason I say they get overused is because I've seen more bad implementations of pie charts than good. Bar charts are good. Line graphs are good as far as conveying information, just because your brain gets wrapped around those things pretty quickly. You just make sure you use those types of charts with the most pertinent, you know, with the right data, you know, ordinal quantitative categorical. So that's why it's important to categorize your data. So, you know, just kind of summarize, you know, you know, have a plan. You know, if I can't stress that enough is to have a plan around your visualizations and do an iterative development process on there. You know, put some things together, go to your audience, show them and say, hey, what do you see here? You know, maybe that's all you do. Put it in front of them and say, what information do you see here? You know, don't lead your audience, right? Your perception is reality. So, you know, what information do you see right here? And if they don't say the things that you're expecting them to say, then you probably need to go back and rework those visualizations. You might want to present the data in a different way. You might want to clarify things for them. So, you know, don't build them in a vacuum. So I think we have about 10 minutes left, Shannon. So, let's see. I've got some questions here. Yeah, so jumping right in. The rest is one that I see here. How do you get users to switch from data-heavy reports to charting? So, it depends on the end user. If it's probably somebody in finance and accounting, it's probably going to be difficult. But, you know, one of the ways to get users to switch is just to present the information in such a way that you answer their questions. So... Oh, so, Stan, can you hear me? Let's see. Somebody asked where I got my source for. We process images 60 times faster than tax. I Googled it, and I don't remember what the source is. I'm real bad about annotating my sources. But I would go and just Google that and see that because I just found that when I was looking up basically some topics around how your brain processes visuals. And that's where I found a lot of that information out there. How do you know which type of visualization is fit for each purpose? So, that goes back to part of... So, to decide which visualization is fit for each purpose, there's two things you need to keep in mind is what is your audience and what type of data are you trying to present? So, the type of data you are trying to present will determine what visualizations you'll use. You know, some examples I used before, you know, talking about where line charts are better than for certain data and don't use them for categorical data, for example. Plus, depending on who your audience is, obviously, if your audience is upper management, you can't...you don't want to use highly complex charts. Like, you know, I see people will put, you know, five dual-axis line charts, for example. You know, where you've got two Y-axes and one X-axis where you're charting two things at once. Sometimes, that works fine, but for other, you know, types of audiences, they get confused because they can't relate the different Y-axes with the two different lines on the chart. So, you may want to use bar charts there. So, anyway, so basically, so wrap that up. There's two things, you know, know what type of data you're trying to present, what category it fits in, and then what your audience is and what information you're trying to present. So, somebody's asking me, by the way, can you hear Shannon? I'm not sure Shannon. Are you talking, Shannon? I'm talking, but you can't hear me. So, um... Oh, well, I can't hear you for some reason. Let's see. Shannon, try it again. All right, can you hear me? I can hear you now, somehow. Oh, there we go. Yeah. Somehow, I looked down and saw my speakers were off. Oh, no. No worries at all. I was like, Shannon's leaving me out to dry. No worries at all. No, you did a great job. Do you want to continue? I see you've got to a couple questions here, but I got a couple more in the pipeline. I think I've answered the ones that are on here, so if you've got some more. I do. So, coming in first, it says you seem to be concentrating on BI focusing on data within the business, but surely a major focus at BI is presenting external information such as competitive and market information? Yep. No, no, I agree. I probably... I probably didn't emphasize it as much, but there's external and internal data that goes into BI. And the combination of the two, you know, is really where the value is. It's really where your competitive advantage is. And whoever asked this question, I guess this is Ray. He's exactly right. You have internal focus and external focus, so competitive and market information is just as important in business intelligence as it is in how the business or what the business is doing. It's the combination of the internal and external that creates the most value in business intelligence. How do you know which type of visualization is fit for each purpose? Did I miss you answering that already? Oh, that one I was just answering. Just sum that up. Two things, know what data, what type of data, where it falls, what category, ordinal, you know, categorical, and know who your audience is. And the combination of those two should give you a pretty good idea of where to start with your visualizations. Because obviously for C-level executives, you don't want to put complex visualizations out there, like I used the example of dual y-axis line grout charts, for example. You know, where you're plotting two different things on the same line chart. That may be too complex for executives. But for an operational manager or a business analyst, that may be fine. Sure. And there was a comment here that says, I agree with page slide number 8 and find the statistics fascinating. Where could I source the, quote, we process images 60,000 times faster than text? Yeah, I was saying, I Google that. I don't have the exact source. I was doing some research on, I had come across somehow, something about how your brain processes stuff. And I just went out there and started Googling, you know, brain and visualization. And I found some of these nuggets out there, different research. So I would just use Google, you know, Google as your friend. Indeed. And it's funny, there's a couple of comments about pie charts. We had at one of our enterprise data world conferences, we have what we call lightning talks, five minute presentations, and one of the presentations that won one year was pie charts are evil. And why not use pie charts? You know what, and people who have been in this business a long time, they don't use pie charts. They know not to use pie charts, but you still see them everywhere, right? You know, I think it's just, that must be the first one that comes up in Excel charts or something. I'd have to go back and look. Yeah, it just depends on what you're looking at. Sometimes it's the easiest to understand. The other thing you see, and I talked about this in the presentation is, is some people become or are becoming enamored with all of these really fancy 3D looking charts, like a map with bars coming out of it, you know, and they can get really confusing if you don't use them in the right way. Like if you have a lot of data and you have 53D bars coming out of a US map, it can get kind of confusing depending on what information you're conveying. So, you know, keep that in mind when you see, you know, like really cool looking charts you want to use on whether they're appropriate. Indeed. We'll say that is all the time we have for today. Just a reminder to everybody, I will send a follow-up email by end of date Thursday for this presentation with links to the slides and links to the recording. And thanks all of our attendees for being so engaged in everything we do, and thanks for the assist in the sound and the attendees. Love it, just love it. And Stan, thank you so much. Thanks, Adira, for sponsoring. And I am pleased to announce that we've got a series coming up next year with Adira. We're going to have a webinar a month. I'm very excited. So, we will get that going and get that all set up for you all. So, thanks very much. Happy holidays, everybody. Thanks, Stan. Thanks, Shannon.