How to go about visualizing a single KPI – Part 4

Last and final post in this series. Previous links are as follows Part1, Part 2 and Part 3.

Style #6: Stack Up’s
Stacked representations are similar to spreads discussed earlier, but slightly vary in terms of interpretation. Stacking up is typically used to show the single KPI with reference to two dimensions and how the KPI adds up to the total value against the dimension sets. One key item to be considered here is that both dimensions are discrete in nature and not time time based. In many instances visualization styles can be swapped seamlessly without diluBlog_SingleKPI_8ting the content and including noise. For example, the chart above can be easily adopted as a column based representation.

What if one dimension happens to be time in our data set, then the best representation again boils down to trend based charts where the lines are stacked on one top of the other instead of intersecting within themselves. In the example to the right, both line and area representations are shown and user can easily infer volatility over time based on the lines. Area chart on other Blog_SingleKPI_9hand has one advantage over the line chart when the user wants to visually analyze the change in each spread/contribution over period of time.

Style #7: Trickle Down’s and Up’s
This is a special scenario, which is related to a special dimension that I would refer to as “workflow stages”. The passage of a single value KPI over and acBlog_SingleKPI_10ross multiple stages of the workflow is represented visually. A good and common example of this is “Sales Opportunity Pipeline” analysis, which is typically represented as a funnel chart. The base data is similar to spreads, but the quantum and type of information conveyed is different. It is easy to spot a bottleneck by looking at the size of a data point and immediately delve into it.

Note: This representation was adopted for performance monitoring of transaction volumes processed by different teams in an organization.