Adventures in Visualand: Data visualization today (Episode 6)

The five most important types of visual charts (and their right uses)

Right dashboard. Right to the boardroom.

There is no right or good or appealing visualization for every purpose. Usually, BI solutions provide a basic set of 5 to 10 graphic models, from column and pie charts to bubble charts and treemaps that can be shaped by the number of applied attributes and measures, filters and benchmarks, and also color and size weighting. Each broader type of graphic model has proven its usefulness for specific pattern representations of data.

Figure 7: Financial dashboard

Source: www.bimeanalytics.com

In order for business users to use the right types of data visualizations at the right moment for the right audience, one must review the categories of graphic models and their intended use.

Figure 8: Types of charts

Source: www.bimeanalytics.com

Here are the 5 main categories into which 15 of the most used types of visualizations fall:

TIME COMPARISON

Figure 9: Time comparison charts

Source: www.bimeanalytics.com

By using column charts, line charts, area charts and sparklines, one can ensure a global view of a specific element or multiple elements over a time period. Line charts and sparklines outline evolution differences between concurrent elements, area charts focus on the weight discrepancies between these, while column charts can represent comparisons between different elements but also different intervals of time which can help business users understand seasonality or growth potential for specific products.

Multiple impacting factors can be compared by representing a dual axis in these charts. Moreover, these graphic models are suitable for various different calculation approaches (such as cumulative totals, difference, or % change), maintaining for each one the clear view of ups and downs in the dataset.

Thus, these models are one of the best ways to set up visual time alerts for a business user.

CATEGORY COMPARISON

Figure 11: Category comparison charts

Source: www.bimeanalytics.com

Figure 11: Category comparison charts

Source: www.bimeanalytics.com

Pie, bar and column charts are natural choices for comparisons between categories and outline immediately overall performers that have to be further developed while also revealing growing gaps that have to be filled or eliminated altogether.

Figure 12: Category comparison treemap

Source: www.bimeanalytics.com

More than these, treemaps offer additional ways to understand differences between elements in different subcategories. Funnels also offer the ability to view the evolution or involution of a certain category at a stage-level, which empowers micro-decision management (we do not have to cancel a whole production line, just have to repair a certain part of it to make it efficient).

These representations can help business users avoid the sin of drafting false stars that appear in the first stage of visual processing.

TARGET COMPARISON 

Figure 13: Target comparison charts

Source: www.bimeanalytics.com

Figure 14: Target comparison charts

Source: www.bimeanalytics.com

Everybody is dealing with KPIs and using them as static reference points in their visual dashboards. But it’s bullet and gauge representations that best serve the purpose of evaluating target reach. While gauges have the advantage of visual similarity with the personal speedometers, bullet models are more compact and also offer the possibility of setting intermediate targets and understanding how many resources have to be invested in order for a target to be accomplished or surpassed.

Column and bar charts can also be used for this purpose but people are using these models because they encompass multiple data points in the minimum amount of space while displaying a race-like visual approach that better delivers the importance of achievements or under-achievements. This is also extremely efficient when doing competitive analysis on similar features between companies in the same field.

MULTI-METRICS COMPARISON

Figure 15: Multi-metric comparison bubble chart

Source: www.bimeanalytics.com

There are so many aspects of a business that are mutually influencing each other that basic charts cannot cover all of them in a comprehensive manner. Bubble charts can encompass multiple measures and represent them by shape, color and size weighting as well as placement so that a visual model can be loaded with information while still delivering a view of the predominant pattern and its outliers.

Moreover, there is an increasing number of cases when 2-axis representations can no longer comprise all factors in a clear visual navigation mode. Therefore, business users started using radar multi-axis representations to evaluate different combinations of interdependent elements. This model is extremely useful for analyzing expanding business models or companies that are developing additional sales / communication channels and want to analyze how a change in one of them generates simultaneous changes within the others.

GEO COMPARISON

Figure 16: Geo comparison

Source: www.bimeanalytics.com

For global companies to SMBs looking to expand, geocoded maps have become an indispensable tool to identify opportunities among local communities and redistributing resources in real-time. This kind of visual model is suitable to cross internal data with national economic, social or cultural figures to understand target groups’ behavior and taking into consideration an increasing number of patterns.

Figure 17: Frequency Map

Source: FlowingData.com

Furthermore, on interactive geocoded maps, users can also see the evolution of patterns over time (one relevant example is the Hurricane History Map).

The major advantage of geocoded maps becomes clear in these interactive dashboards where the user can zoom in and out on specific areas, so CEOs are always in the know about the worldwide situation of the business, and local managers are able to benefit from niche trends within communities. Real-time maps are also used on both business and social emerging critical situations (the two examples here present the Digital Attack Map done by the Google Research Group and the Riots Map done by The Guardian during the upheaval in London of 2011).

Figure 19: England Riots Map

Source: DigitalAttackMap.com

Figure 18: Digital Attack Map

Source: TheGuardian.com