Can data visualizations be easily automated?

Modern data visualization Are you using the full potential of your data?

If you are concerned with the visualization of data, visualization methods such as line or bar charts will probably first come to mind. These chart types can be found in many data visualizations. However, you should be aware that diagrams are just one of many visualization methods. The application of the right visualization techniques for the respective objective ensures an effective and at the same time interesting representation of the data. Typical visualization methods are diagrams, maps, infographics, tables and interactive dashboards. You may be surprised that tables are included in this list, as they are pure figures. By arranging data in rows and columns, data can also be visualized with tables. There are numerous subspecies for each of the visualization methods. For example, there are line charts, area charts, column charts and many more. We will go into more detail on the selection of the appropriate diagram types in one of the next chapters. The very modern visualization techniques include interactive dashboards. They allow the user to filter data and adapt the visualizations. Many of these dashboards update themselves and provide data in almost real time. You can see an example of such a dashboard here:

As far as the presentation of the data visualizations is concerned, there are various visualization options. Relics from the analog era are overhead projectors, flipcharts and pin boards. Even if these are not the most modern visualization options, flipcharts and pin boards are still in use in many companies today. Pin boards are suitable for small groups and can be used spontaneously for data visualization, for example by pinning colored cards and symbols with pin needles. Flipcharts are also intended for smaller groups and make it possible to explain simple relationships with sketches. For data visualizations such as dashboards and complex diagrams, however, flipcharts and pin boards cannot be used. As a rule, the presentation today takes place with digital visualization options on a screen or projector.

5) With 7 questions about the visualization concept

Before you start with data visualization, you need a visualization concept. In order to develop the visualization concept, it is helpful to deal with the following seven questions in more detail. By finding the right answers to these questions, you will be able to use the right visualizations for your business goal.

Question 1: What story would you like to tell?

The task of data visualization is to translate the numbers and other information into graphical representations in such a way that they tell the desired story. Messages should be conveyed to the audience that they can understand quickly and easily. In order for you to succeed, you have to be clear about what story you are actually telling or what you want to convince your target group of. Only then can statistics and metrics be brought into the correct visual form.

Question 2: What is your target audience?

Determining the audience you want to get your message across is another key element in choosing the right visualization techniques. For example, the target group can be a specific department of your company, management or an external investor. Depending on the audience, the data must be appropriately prepared and visualized. Take the necessary time to pinpoint the target audience.

Question 3: Do you want to analyze certain trends?

After the target group and story have been determined, you should ask yourself whether certain trends should be analyzed and visualized. Depending on the respective data, the time periods to be observed and the trend, certain chart types such as line charts, bar charts or area charts must be used.

Question 4: Should the composition of data be presented?

Data visualization is often about presenting the composition of data. It should be shown how the complete picture is composed of individual data. In these cases special types of diagrams such as pie diagrams, waterfall diagrams, stacked diagrams or maps (in the case of geographic information) should be used.

Question 5: Should data be compared with one another?

If your primary goal is to compare certain data or data trends with one another, the following chart types are the right ones: network charts, bar charts or column charts.

Question 6: Is a certain time frame to be considered?

Data can be viewed completely independent of time or in a specific time frame. If the messages to be conveyed are time-dependent, use diagram types that can display specific time periods and courses. Line charts or bar charts, for example, are suitable.

Question 7: Would you like to present key performance indicators (KPIs)?

Another important question for your visualization concept is whether KPIs should be presented. Think about what information you want to gain from the KPIs and experiment with different formats. This is how you can find suitable charts such as speedometer diagrams for displaying KPIs.

6) Types of Diagrams - Tips for making the right choice

Now that you've answered these seven questions, let's dig deeper into the top ten chart types and how to use them. For each of these chart types, we explain to you for what purpose it makes sense to use it and what you should definitely avoid. The basic categories for choosing a chart correctly are relationships, distributions, compositions, and comparisons.

Before we start, the following graphic gives you a good overview of which type of diagram should be selected for which visualization goal and which type of data to be displayed.

1) Numbers chart

Sensible use of number charts

A real-time number chart behaves like a live ticker, which shows the current value of an important KPI, for example. Values ​​such as sales figures, percentages or visitor numbers can be recognized at first glance. The number chart is one of the simplest types of visualization. Its purpose is to represent a value at a specific point in time. You just have to consider what time that should be. In order for the target group to understand the message, it is important to provide the number chart with a time label. By adding a trend indicator, the value shown can be compared to a target or a previous time period.

What to Avoid

Number charts are often the first piece of information that viewers grasp. Under no circumstances should you include too many of these charts in your data visualization. That confuses the audience. In addition, a dashboard with too many number charts can quickly become superficial. In order to go deeper into the information, other types of diagrams must also be used. If a trend indicator is used, it must be designed or labeled in such a way that the viewer can immediately see what the value shown is being compared with.

2) line graph

Sensible use of line charts

The purpose of line charts is to show history and trends over time. Line charts graphically show how data has changed over a period of time. The example above shows the sales trend over time in 2014 sorted by payment method. You can see immediately how the sales with payments by credit card have developed in comparison to the other payment methods or that there was a general slump in sales in September.

A great feature of line charts is that they can be combined with other types of charts. For example, you can incorporate a column chart by adding a second Y-axis for line and column values. Due to the common representation, relationships can be better represented.

What to Avoid

Too many lines on a single diagram make it difficult for the viewer to understand. You have to jump back and forth between the legend and the lines in order to know what you are looking at. This reduces attention and makes you tired. Better to use different line charts with a few lines each.

As far as the scaling of the Y-axis is concerned, it should be selected so that the highest data value is in the upper range. For example, if you scale the Y-axis to show values ​​up to 1000, but the highest data value is only 500, you are wasting a lot of space. Gradients are more difficult to see because they are compressed. Give your data room to breathe!

3) Map

Sensible use of cards

Maps are great for displaying data based on geographic locations. As in the example above, it makes sense to use different colors. Thanks to the geographical representation, the data can be embedded in stories or actions can be derived. The viewer immediately recognizes where measures are to be initiated based on the data situation.

What to Avoid

Cards are very popular. However, that doesn't mean you have to incorporate maps into every data visualization. If the location isn't part of your story or message, you'd better leave the card out. Since maps take up a lot of space in your presentation, you should only include what is really necessary. It is also not a good idea to overwhelm the map with many individual data points. That overwhelms your audience.

4) Waterfall chart

Sensible use of waterfall diagrams

Waterfall charts are extremely helpful in understanding how positive or negative values ​​affect an overall result. They show the composition of data and visualize the increase or decrease based on an initial value. Waterfall charts are widely used in finance departments. There they illustrate changes, for example in income or profit. The above example shows such a data visualization.

What to Avoid

Waterfall charts are intended for the presentation of static data. If you want to show dynamic data, a stacked bar chart is definitely a better choice. Dependencies between several variables are also difficult to illustrate with waterfall diagrams.

5) Bar and Column Chart

There are three types of bar charts, horizontal bar charts, column charts, and stacked variants of them. Although they belong to the same type of diagram, they can be used for different purposes.

a) Horizontal bar graph

Sensible use of bar charts

Horizontal bar charts are ideal for visualizing comparative rankings. For example, top five or top ten lists can be created. Even if data labels are relatively long, this type of diagram makes sense. The bars should follow a certain order, such as sorting by value or by alphabet.

What to Avoid

Since it is better to display temporal progressions from right to left rather than from top to bottom, the horizontal bar chart is better not to be used for such purposes. If the charts contain too many values, horizontal bar charts can quickly appear chaotic.

b) Bar chart

Sensible use of bar charts

Column charts are the standard for comparing data from different categories and showing chronological data such as growth over time. Our example shows the sales figures per country and per channel in the last year. It becomes clear at first glance that six regions and five channels are being compared. The color assignment informs about the region and the grouping with the generous free spaces in between over the canal. The viewer quickly recognizes the channel and the region with the greatest sales figures.

c) Stacked column and bar charts

Sensible use of stacked column and bar charts

With the stacked column and bar charts, data can be compared in itself in addition to the visualization options described above. Often in the form of percentages, as in the example above. The example diagram is not about showing the total number of customers, but rather their age distribution. The percentages relate to the total number. In comparison to a pie chart, it is not only possible to show the percentage breakdown at a single point in time, but also the development over a period of time, in this example quarterly. Thanks to the color coding, it quickly becomes clear that the weakest customer group is the over 55s. The grouping by quarter shows that the proportion of people over 55 is not increasing.

What to Avoid

If you try to put too much data into a single stacked chart, each bar or column will become very thin. Unfortunately, this leaves little space for lettering. Therefore limit yourself to the essential data. To make the diagram easy to understand, use appropriate color mappings, enough space between the bars, and a balanced layout.

6) Pie Chart (Pie Chart)

Sensible use of pie charts

Pie charts are useful if you want to show the relative composition of a variable at a fixed point in time. Pie charts are perfect at this type of visualization, but you shouldn't use them for other purposes. These diagrams are ideal when the individual parts add up to 100 percent. Each individual piece of cake shows at first glance how large the respective portion is. The proportion can be estimated well even without labeling. Labels with the exact percentages can be added for detailing.

What to Avoid

If too many shares are packed into a pie chart, it can quickly become confusing for the viewer. The percentage of individual circle segments is difficult to distinguish from one another and precise comparisons are no longer possible. It shouldn't be more than six different pieces of cake. This type of diagram is completely unsuitable as soon as a temporal development component comes into play. You cannot depict trends or time courses with pie charts. In order for your audience to understand what point in time the pie chart represents, appropriate labeling should be provided.

7) Dial indicator or tacho diagram (gauge chart)

Sensible use of dial gauge or tacho diagrams

This type of diagram shows data in a dial gauge or tachometer display. Indicator needles and colors ensure that the values ​​are visualized. People understand this type of representation very quickly. Tachograph charts are ideal for showing a single value in relation to a total quantity or a reference value such as a target. The diagrams are often used in management reporting and show individual values ​​of important KPIs, for example in relation to a target corridor to be adhered to. Thanks to the color design of the measuring scale, the displayed KPI value can be correctly classified immediately.

What to Avoid

A dial indicator or tachograph chart has difficulty comparing multiple variables. Chronological courses cannot be mapped either. The diagrams also take up a lot of space. You should therefore not include too many in a report or dashboard.

8) Cobweb diagram

Sensible use of spider web diagram

Spider webs are the right choice for visualizing multidimensional data with three or more quantitative aspects. They are suitable for ranking lists, evaluations and assessments. In the example shown, we are comparing different aspects of the three regions Europe, North America and Australia. The aspects are cameras, televisions, cell phones, games, and computers sold. The diagram quickly shows that Europe is ahead in all aspects. The Australia region is particularly weak in terms of camera and cell phone sales.

What to Avoid

Cobweb diagrams are not easy to put together. But if you do it right, they impress. If you use this type of diagram for too many quantitative aspects and dimensions, it can quickly become confusing and difficult to interpret.

9) table

Sensible use of tables

Tables are actually pure figures, but they can still be used as a visualization method. Tables present large amounts of exact values ​​in their raw format. Individual values ​​can be precisely compared while at the same time showing the big picture. This is useful, for example, if you want to go into the details of your data visualization in certain areas. Tables are also useful if individual people in your target group want to see different values ​​in a table. By using suitable colors, frame lines, fonts, number formats and symbols, individual areas or values ​​of the tables can be visually highlighted.

What to Avoid

Tables are useful for some purposes, but there are often better ways to visualize data. People take more time to gather information from tables than they do to information from charts. The larger the tables get and the more data you put in, the more serious these time differences are.In addition, you should not overdo it with the color and visual design of the tables, this makes the visualization restless and, in extreme cases, chaotic.

10) Area chart

Sensible use of area charts

Area charts are closely related to line charts. Both types of diagrams depict chronological progressions and are well suited for recognizing trends. However, there are differences. These make the use of area charts useful for special visualizations. Line charts connect the data points with straight lines. Area charts do this in a similar way, except that the area under the connecting line is filled with color. If you stack an area chart, various data aspects become visible. The height of a data point represents the total value, the individual heights of the colored areas show the individual proportions of the total value. For example, the overall trend of all sales together with the individual trends of the various regions can be displayed in a stacked area chart. Percentages can also be stacked in area diagrams. These representations show how the relationships between the various parts change over time.

What to Avoid

Stay away from unstacked area charts. They show no more than a line graph. Area charts are good for two to three different variables, but try not to fit more than seven. These stacked area charts are difficult to read.