Multiple-Item and Multiple-Value Charts

Most Brightmetrics users are familiar with single-item charts, which is to say charts that display a single value, potentially broken out by a dimension. For example, count of all calls taken by a workgroup, either as a single total value or perhaps broken out by exit reason, answering agent, or wait time interval.

But what if you wanted to see the count of all calls taken by a workgroup, as well as the 28-day moving average of the same value to see which days are above or below trend? Or average queue time for a workgroup vs average talk time? Or the count of calls coming in to the main customer service number vs count of calls handled by the customer service workgroup? For these cases Brightmetrics has an advanced option known as multiple-item charts.

When you are in edit mode, you can not only drag and drop a data source onto the “new chart” area, but also on to any existing chart to add an additional item to that chart.

micss2-small

However, there are some important considerations when creating multiple item charts. First, it only applies to time-series charts — bar charts only support a single item and will ignore any additional items that are added.

Second, the selected values from the items have to match in scale — they have to all be counts, minutes, percentages, etc. It would not make sense to chart a count of calls along with a duration value because the units are not comparable.

Lastly, the “Category” selector to break charts out by a dimension no longer has any effect. When you have a single item broken out via Category, the values of the category dimension become the series that are shown in the chart legend. When you create a multiple-item chart, each item is its own series as shown in the chart legend. The name you give to the chart item in the advanced item options is the name shown for it in the legend.

Other than those considerations, the restrictions are few. The values could be from the same data set (e.g. queue data), different data sets (e.g. user activity and queue data), or even complete separate data sources (e.g. call data from two separate ShoreTel systems). There is no limit on the number of items that can be included.

There is an optimization for adding multiple values from the same data set with the same filters, which we call multiple-value charts. When you are on the chart options page there is a little green “plus” icon next to the value selector, which will add an additional value selector. This can be repeated multiple times. Now you can select multiple values from the same data set — for example average queue time and average talk time.

micss3-small

The end result is the same as adding each as separate items, but you do not have to re-select all your filters to match each other, the single filter selection applies to all values. Additionally, bar charts do support this mode, so each selected value will be displayed as a bar on the chart. The other restrictions still apply: the selected values have to match in scale and any category/subcategory selections will be ignored.

Multiple-item charts and multiple-value charts are mutually exclusive — once you add multiple values you can not drag and drop an additional item on to the chart. Likewise, if you have put additional items on the chart you will not be able to click the icon to add additional values from any one of them.

The majority of the time, single-value charts, with or without category and subcategory breakdowns, provide the kind of information that needs to be displayed. Occasionally however, more complicated data relationships need to be displayed, and while the multiple-item/value feature itself can be complicated to use, it exists to meet that need. I hope this article provides some insight in to when and how it might be used.

You may also like

Leave a comment