On Clarisights, a visualization is called a 'widget', and each of the different types of widgets available is useful for a specific purpose. Here, we explain them with examples, sample use-cases and the constrains they carry. Data exploration features that almost all these visualizations support are filters, sorting and date override.
Table is one of the most commonly used widgets. You can use the table widget to display data in rows and columns. It supports any number of metrics, dimensions and breakdowns. A number of data enhancements such as breakdowns, trends, aggregate, share and comparison are also supported in tables. Note that if you have added more than one dimension, then you wouldn't be able to add breakdowns in a table. Tables are best to see a snapshot of a large volume of data at once.
A column chart displays data in the form of vertical bars. The dimensions you have chosen are represented on the horizontal axis while metrics are on vertical axes. Multiple metrics are supported, and you can visualize them with different colors and move them into separate axes. Note that if you have added more then one dimension, then each combination of the multiple dimensions will be a separate value on the column chart. Breakdowns are not supported in column charts. Column charts are great when you want to visualize overall trends in aggregate metrics.
Area charts help you visualize data as the area covered below lines. Dimensions and metrics work the same way as a column chart. Breakdowns are not supported in area charts. Area charts are recommended when you want to compare the values of one metric with another over time - for example, cost v/s revenue.
A horizontal chart is similar to a column chart with the difference that positions of dimensions and metrics swap - i.e. dimensions are on the vertical axis, while metrics are on the horizontal. Like a column chart, you can style metrics into different colors and/or axes. Breakdowns are not supported here either.
Combo charts enable you to visualize lines, areas, and columns in the same widget. Dimensions and metrics work the same way as a column chart and breakdowns are not supported. Combo charts are best when you want to include dissimilar metrics in the same graph and look for correlations, for e.g. a ratio metric like CPC and a cumulative metric like Spend or Impressions.
If you'd like to visualize metrics over time, then timeline charts are recommended. It removes the hassle of adding/removing dimensions, and allows you to quickly switch between daily, weekly, monthly views etc. You can add metrics, filters, and style the metrics in the same way as a combo chart. Note that you can't change the sorting of a timeline chart, it's always sorted by ascending order of the chosen time dimension.
Breakdown charts are used to break-down a metric for different values of a dimension, and visualize these values over another dimension. You need to add only one dimension, one metric and one breakdown for this visualization. Filters are supported, but styling and sorting are not. A common example of a breakdown chart is CTR breakdown by Channel, over Weeks, Countries, Product Categories etc.
Stacked charts are very similar to breakdown charts, in that they help you visualize the value of a metric broken down, across a dimension. You can also add a second metric to visualize it as a line graph across the dimension. Each breakdown value can be distinguished by the color of its section in the stack. Like with breakdown charts, filters are supported but styling and sorting are not. Stacked charts are useful in cases like breaking down orders/leads by channel across time.
Tiles help you look at aggregate values of metrics. You can add filters and comparisons, but dimensions and breakdowns are not supported. They work best when you want to look at the overview performance of a channel or campaign type (web v/s app, brand v/s non-brand etc) quickly.
Pie charts are the classic way to show split of a metric across multiple dimension values. You can add one metric and one dimension in a pie chart, and filter out data that you don't deem useful. The slices are sorted by the metric by default. These are recommended as a quick visual way to show distribution of spend, conversions, revenue etc. across segments like channel, country, device etc.
A heat map visualizes a metric with differing color intensities to represent its values. You can create a heat map by adding two dimensions and one metric. The scale adjacent to the chart helps understand the correlation between the color darkness and the metric's value. Filtering data is supported, but sorting and breakdowns aren't. Heatmaps are most commonly used to visualize data when you want to figure out the highs and lows of a metric across two dimensions.
A bubble chart is used to represent 3 elements in a chart. Here one bubble represents one dimension which is plotted corresponding to the values of selected two or three metrics (represented by two axes and the bubble radius).
To learn more about it's uses and application Click here
The Funnel Chart is used to visualise the progressive reduction of data as it passes from one phase to another. In addition to metric values, you can introduce breakdowns to analyse the contribution of other dimensions in that metric.
The drop of vertical values here represent the reduction in a breakdown value as it passes to the next phase.
This type is used to enter a text as a widget in a report.