1. Grouping call center data
Now let's see how groupings and aggregations can help Oakmark's team analyze monthly escalation calls.
First, from our column menu we'll drag Call Date into the Groupings menu.
Sigma creates this box to help us understand how our data is organized.
Use the Call Date dropdown to truncate to Month.
Next, to add a call count calculation, we click the plus sign then row count to automatically add the count. Let's rename that to Total Calls.
Now, to add a count of escalated calls, we'll drag the Was Escalated column into that same grouped calculation box. Because this column is a True/False data type, Sigma automatically counts only True values.
Let's rename this column to Calls Escalated
Take a look at the new table structure. The toggles on each grouped column collapse or expand to drill further into individual months.
Call escalations at Oakmark Bank are supported by three separate teams: Card Services, Fraud Detection, and Account Support.
Let's add a team breakout for each month by dragging the Agent Team column into the same Group By box as Call Date.
Notice that our counts are now based on the combination of these 2 columns. Where we previously showed 246 total escalations, we now see a breakout by team and month.
To see separate monthly totals and totals by team, we could separate our Groupings into 2 Group By boxes.
With the Agent Team box, below the Month box, we can add a row count in the agent box to answer in this month, how many calls did this team receive?
If we change the order of the group by boxes so Agent Team is on top, we can answer a different question of How many calls has this team ever received and break it out by month.
Groupings give us lots of flexibility to answer different kinds of questions.
Let's say we want to see the share of calls handled by each team for each month.
In the Agent Team box, we can click the plus sign to add a new calculation that references these grouped calculations.
We can select our Agent Team Call Count and divide it by the Total Calls for that month.
Let's format it as a percentage and rename it to avoid confusion.
Remember that under the hood, this is an aggregate calculation - its operating not only on the row of data we can see, but because this table is grouped, our calculation is operating on the entire subset of rows making up this group.
In addition to the grouped totals, we might also want to keep the total number of calls in the table within view. To do that, we can use the Summary Bar at the bottom of the table.
Click the arrow to expand, then click the plus sign to add a table-level summary, like Row Count, which shows the total number of calls across the entire dataset.
These table summaries can be referenced in calculations just like our other grouped calculations.
In the next exercises, you'll practice setting up groupings, calculations, and summary statistics on Oakmark's call log data.
2. Let's practice!