As we covered in a separate article, Getting Started with Event Segmentation, Event Segmentation is the core Amplitude chart that uncovers behaviors of your users and what they're doing in your product. Event Segmentation allows you to create charts from events and event properties. There are 3 main components to the Event Segmentation chart: the Events module (left module), the Segmentation module (right module), and the Metrics module (bottom module). Together these 3 make up what we call the Chart Control Panel.
This article will cover more details of the Metrics Module (Bottom Module) and how to use additional aspects of this feature not covered in the Getting Started with Event Segmentation article. We also have an article covering the Segmentation Module (Right Module) in more detail!
Table of Contents
- Real-time Segmentation (Enterprise)
- Period-over-period Comparison (Enterprise)
- Group By Visualization
- Custom Formulas
- Rolling Averages
- Rolling Windows
- Cumulative Sum
Real-time Segmentation (Enterprise)
Please note the following limitations when you view data on a real-time basis.
- You can only segment two days worth of data for real-time
- The event times are rounded down.
- Charts are cached every five minutes for everyone.
Period-Over-Period Comparison (Enterprise)
The "Compare" option allows you to compare the results of the current time range with the previous day, or the same day from previous week, month, quarter or year.
For example, let's say you want to compare the daily active users for the current week compared to last week.
The blue segment shows you the current period and the green segment shows you your data for last week. Microscope is currently shown on the data point for Sep 19, and the corresponding green data point would be for the day before, Sep 18:
Since the period-over-period comparison interval is configurable, you can choose what dates you actually want to compare.
Period-Over-Period for Custom Formula (Enterprise)
Period over Period comparison can also be applied when using the Custom Formula metric. For example you can compare your current rolling average with the previous months rolling average:
Microscope is a feature that allows you to dig deeper into a specific data point and is accessible in all of Amplitude's charts except Pathfinder and the Custom Formulas in the Event Segmentation chart. When you hover over a data point of some charts, a black pop-up will prompt you to inspect the users who make up that particular data point. Depending on your Amplitude plan, Microscope will display up to five options, detailed below.
A full outline of the Microscope feature can be found in this document.
Group By Visualization
If you have multiple "group bys" applied, then often times the default visualizations can become confusing and hard to interpret data. The group by visualization offers a clearer way to analyze your data. For example, the following chart control panel groups 'Play Song or Video' by 'Genre_Type' and also by country and platform.
The group by visualization will generate a table view that displays the values in separate columns. This makes it easier to digest the data. For example, we can see that for users in the United States who played Pop songs, most of them were on a Web platform.
Please note: If you are doing a single group by, these results may be limited to 100 group by results. If you are doing a double group by, these results may be limited to 500 group by results in which each group by value pair is counted as 1 result.
# Absolute vs. % Relative
# Absolute will provide you with the user volume whereas % Relative is the series value divided by the sum of all the series values. This reflects the total contribution value of all series within the group bys.
In an Event Segmentation chart, you can write formulas to perform or calculate specific analyses and metrics on events. You can plot up to four formulas on the same chart, separated by semicolons (;). You can also perform the following arithmetic operations: parenthesis, addition (+), subtraction (-), multiplication (*), and division (/). There are several functions you can use to create custom formulas to apply to your Event Segmentation chart. The text box will autocomplete with suggestions for formulas to use, and you can also press ctrl+space to display all formulas. To read more about each formula and see some examples of use cases, see our Custom Formulas article here.
Rolling averages will display the unweighted mean and therefore "smooth out" a chart. This functionality is useful if you have cyclical users such as people who use your product Monday-Friday, but not on Saturday-Sunday. To apply a rolling average to your chart, select the "Advanced" button in the right-hand side of the bottom module.
Example 1: Measured by Event Totals
The below line chart displays the daily Event Totals between February 5th and March 7th, without a rolling average.
The below line chart displays the daily Event Totals, with a rolling average of 7 days. In this chart, the February 28th data point is an average of the numbers between February 22nd and February 28th from the first chart shown above.
You can read further data points as follows:
- March 7th will be the average of March 1st to March 7th. If March 7th is the present day, a dotted line will be shown to indicate that the 7th day (March 7th) is incomplete. To hide this, use "Offset" in the data picker.
- February 6th will fetch data from outside the time period selected and would average data over January 31st to February 6th.
Example 2: Measured by Uniques
Similar to Example 1, suppose we create a line chart of daily Uniques as a rolling average of 30 days between February 9th and May 9th.
You can then read certain data points as follows:
- May 9th will be the average of unique counts over the last 30 days, including May 9th. If May 9th is the present day, a dotted line will be shown to indicate that the 30th day (May 9th) is incomplete. To hide this, use "Offset" in the data picker.
- February 13 will fetch data from outside the time period selected and would average data over January 13th to February 13th.
Microscope with Rolling Average
When Rolling Average is active, Microscope's "View Users" functionality will continue to show users in the current data point chosen and not over the whole interval being averaged on.
Rolling windows will display the aggregate last N days of information in a single data point. This functionality is useful if you want to see metrics such as your 7-day active user count on a daily basis. To apply a rolling window to your chart, select the "Advanced" button in the right-hand side of the bottom module. Rolling windows are not supported for the Property Histogram and Custom Formula metrics.
The following chart displays daily Uniques between April 5th and May 5th without a rolling window. With Microscope, we can see that on April 21st, we had 173,144 users.
The below line chart displays the daily Uniques with a rolling window of 7 days. In this chart, the April 21st data point is the number of unique deduplicated users between April 15th and April 21st. So, while the above chart is showing you the unique users on each day, the rolling window chart will allow you to look at unique users over 7 days on a daily basis.
You can then read certain data points as follows:
- April 21st is the number of unique users from April 15th (7 days before April 21st) to April 21st.
- April 6th will look at aggregate data from outside of the time range selected and show you the unique users from March 31st to April 6th.
The same logic applies for all other metrics in the bottom module that allow rolling windows.
Microscope with Rolling Window
When Rolling Window is active, Microscope's "View Users" functionality will continue to show users in the current data point chosen and not over the whole interval being aggregated on.
Cumulative sum will display a running total of information in a single data point. This functionality is useful if you want to see a total stated at any point in your selected timeframe without having to sum up prior data points.
To apply a cumulative sum on your chart, select the "Advanced" button in the right-hand side of the bottom module. Cumulative Sum is not supported for the Active% and Average metrics. If you would like to use cumulative sum within Formula, you'll need to go directly to the Formula tab and type out CUMSUM.
The chart below shows a running total of revenue generated by "Complete Purchase" events. The April 19th data point is a sum of revenue generated on April 17th, April 18th, and April 19th.
Below is a screenshot of a breakdown table that does NOT have cumulative sum applied. As you can see, the April 19th datapoint (1,495,858) is a sum of data points from April 17th to April 19th (= 499,576.21 + 493,117.31 + 503,164.83) rounded to the nearest integer.
For cumulative sum of UNIQUES, a deduplicated count of unique users will be returned in each data point.
- On 4/17, User A fired "Complete Purchase".
- On 4/18, User A and User B fired "Complete Purchase".
- On 4/19, User C and User D fired"Complete Purchase".
On the data point for 4/19, a total count of 4 will be returned because 4 unique users fired this event from 4/17 - 4/19.