Choosing metric aggregators

Metric aggregators define how the data queried by a metric is calculated. The default aggregator for many metrics is SUM which adds all results together, but you can select a different aggregator to change how your results are calculated. Use this article to learn about the different aggregators you can use, and how to use them.

Note: If you want to set a new default aggregator to be automatically used when you add a metric to a query or would like to prevent specified metric aggregators from being selected, see Setting a metric's default and visible aggregators.

This article contains the following sections:

• Changing the metric aggregator
• Metric aggregators reference
• Choosing the right metric aggregator

Changing the metric aggregator

To change the metric aggregator

1. In Query Builder, add a metric to the Metrics panel.
2. Click the metric you added.
3. From the list, select a new aggregator.

   

   The available aggregators will differ depending on the metric you choose. Use the Metric aggregators reference to find out more about each aggregator.

4. To save the new metric aggregator, click Apply.

Tip: You can also select multiple aggregators to use with a metric. When you add the metric to the Metrics panel, click the down arrow next to the metric, and select the aggregators you want to use.

Metric aggregators reference

Use this table to learn about the different aggregators and when to use them.

Aggregator	Description	When to use
COUNT	Counts all values of an attribute. If the same value is recorded more than once, each occurrence is counted.	Used for counting objects like tickets, calls, chats, users, organizations, etc.
D_COUNT (distinct count)	Counts the number of different (unique) values. Values recorded more than once are counted only once.	Used for distinct counting objects like tickets, calls, chats, users, organizations, etc.
SUM	Adds up all numeric values.	Used for database counted, arithmetical operation and time duration metrics like Chat Messages, % Satisfaction Score, Resolution Time.
AVG (average)	Calculates the arithmetic mean of numeric values.	Used for time duration and database counted metrics like Resolution Time and Chat Messages.
MED (median)	Identifies the numeric value that separates the higher half of the values from the lower half.	Used for time duration and database counted metrics like Resolution Time and Chat Messages.
MIN (minimum)	Identifies the lowest numeric value.	Used for time duration and database counted metrics like Resolution Time and Chat Messages.
MAX (maximum)	Identifies the highest numeric value.	Used for time duration and database counted metrics like Resolution Time and Chat Messages.
VARIANCE STD_DEV (standard deviation)	Measures how far numeric values are spread out from their average value. If it is low the numeric values are close to the average. If it is high the numeric values are spread out.	Used for time duration and database counted metrics like Resolution Time and Chat Messages.

Choosing the right metric aggregator

By default, metrics use the most relevant aggregator. In some situations, a different aggregator might better suit your reporting needs or you might need to decide which aggregator to use on a new calculated metric. Use the following sections to help you to decide which metric aggregator to use.

COUNT and D_COUNT

One of the most common reporting tasks you will use is to count specific objects (tickets, calls, chats, users) or events that relate to these objects. The metric aggregators you use to count are COUNT and D_COUNT.

The difference between these two aggregators is that one allows counting all possible values and another only the unique values. In many cases, both aggregators will return the same results because the values they are counting are not repetitive. For example, if you count Ticket IDs in the Tickets dataset, the results for both aggregators are equal, because the ID of each ticket appears only once in this dataset. If you do the same in the Ticket Updates dataset, the COUNT aggregator returns a much higher result, because Ticket ID is recorded for each update. In this situation, the COUNT of Ticket IDs is equal to the number of updates but D_COUNT is the actual number of tickets on which this updates occurred.

SUM

The SUM aggregator is mostly used for database counted and arithmetical operation metrics, but it also can be used for duration metrics, for example, to calculate the total time spent by agents on calls or chats.

The database counted metrics are computed outside of Explore and appear in Explore as ready to use metrics. They normally measure the number of specific events or objects like Agent Replies, Assignee Stations, Reopens, Chat Messages, etc.

The arithmetical operation metrics are calculations created in Explore that perform some numerical operations with existing metrics. For example, SUM should be used for % Satisfaction Score metric, which is a division of Good Satisfaction Tickets by Rated Satisfaction Tickets.

AVG, MED, MIN, MAX, VARIANCE and STD_DEV

Team efficiency and performance is analysed in Zendesk through time duration and database counted metrics aggregated via AVG, MED, MIN, MAX, VARIANCE and STD_DEV.

AVG and MED play a similar role in understanding the tendency of a metric. If the data you are comparing is mostly uniform then the AVG aggregator can be used. However, in most cases, Zendesk data has a high number of outliers, meaning that MED is the safest option. See Average vs Median for a closer look at these aggregators.

MIN and MAX aggregators can be used to identify the highest and lowest metric outliers.

VARIANCE and STD_DEV are more advanced analytical aggregators and are rarely used, but they are helpful in measuring how uniform or spread out the metric values are. If you never used these aggregators start with STD_DEV. Because the standard deviation is expressed in the same units as average and median aggregators and can be easily compared with them.