Funnel Analysis helps you understand how users are successfully navigating a defined path in your product and where there is a drop-off. A funnel is a series of events that a user progresses through within your app, such as successful onboarding. A user is considered converted through a step in the funnel if they perform the event in the specified order.
As an example. Funnel Analysis can be used to measure the conversion rate of regular users to power users, or free users versus paying users.
What You Will Learn in This Article
This article provides information on how to create a default funnel and the main functionalities of the Funnel Analysis chart. It specifically will cover:
- Functionalities of the Event Module (left module)
- Functionalities of the Segmentation Module (right module)
- Additional features that are available to set how users navigate through your funnels
Before you start reading this article, we recommend that you have a look at the prerequisites list below. The information presented in this list will help you out with both useful information and good to know practices but will also contain links to related topics and Help Centre articles:
- A/B Test View and A/B etc is a feature that is only available to the Growth and Enterprise plans
- There are 2 other Funnel Analysis articles to help you make the most out of this chart: Funnels Analysis - Conversion Drivers and Funnels Analysis - Advanced which contains info regarding some of the more advanced metrics:
- Conversion Over Time
- A/B Test - Improvement
- A/B Test - Significance
Table of Contents
- Creating a Funnel
- Chart Interpretation
- Default Configuration
- Holding a Property Constant
- Exclusion Steps
- Event Conversion / Drop-off
- Date Picker
- Compare to past
- Breakdown Data Table
Creating a Funnel
Events: Select which events you want to be included in a funnel. You can also specify event property criteria for each event by hovering over a step and clicking "+where" to the right of the event name. To rearrange the steps in your funnel, you can drag and drop each event by clicking on the number to the left of the event name. If you would like to combine multiple events together into a single step, then you can create a custom event out of your event types. This would function like an "OR" clause so that users can do any of the individual events inside of a custom event and will count as having completed that step in the funnel.
by...: Define whether you want the funnel to look at Active or New Users.
- Active Users: Includes all users who have performed the first step of your funnel.
- New Users: Includes users who are new who have performed the first step of your funnel (at any point in the time period selected, e.g. funnel entry is not restricted to the day the user is new). This is the default behavior for a new user funnel. If you wish to change it this, see the bottom module section.
You can also specify user property criteria for each event by clicking "+where" to the right of the user segment.
The right panel in the Funnel Analysis chart control panel functions analogous to the right panel in an Event Segmentation chart. You can read more about how to compare user segments here. To learn more about changing "Users" to account-level reporting such as "Country" or "Company Name," click here.
...completed within: Here you can specify how much time a user has to complete the funnel from the moment they enter it. "Completed within" is also referred to as the conversion window. You should adjust your conversion window when there is a sequence of events that you expect users to perform within hours/minutes/seconds of each other. The default conversion window is 30 days (in UTC), meaning users have 30 days from when they enter the funnel to complete the funnel and to count as converted. The minimum conversion window is 1 second and the maximum conversion window is 90 days.
any day: If you are looking at a new user funnel, you will see a dropdown appear here. If you select "any day", then the funnel will include users who are new who have performed the first step of the funnel at any point in the date range selected.
their first day: If "their first day" is selected, then this restricts the funnel to users who perform event 1 (and enter the funnel) on their first day they appear in Amplitude (their new user date).
grouped by: See how users with a property value at a specific step have converted through the funnel. Read more about conversion by property here.
holding constant: This allows you to hold a selected property constant through the entire funnel. If you wish to hold an event property constant, the dropdown will only allow you to select an event property that was present in every step of the funnel (since event properties are not global properties like user properties). Read more about this setup below.
shown as: Select "Conversion Funnel" to see a step-by-step funnel or "Conversion Over Time" to analyze conversion rates over time. You can choose to view conversion rates over time for the entire funnel or for a particular step.
AB Test View: If you click "AB Test Improvement Over Baseline" or "AB Test Chance to Outperform" in the bottom module, you can see two different visualization of an A/B test you have instrumented via Amplitude's Funnel Analysis chart. You can use the "..where baseline is set to" dropdown to select the property value that you want to use as the baseline. You also have the options to show the visualization as "Improvement Over Baseline" or "Chance To Outperform". Read more about the AB test view below. This is an Enterprise only feature.
You can set up and interpret any Funnel Analysis chart easily as the platform allows you to read the parameters like a sentence. The default configuration has "grouped by" selected in the bottom module configuration. For example, the following chart shows you Events performed in this order by Active Users completed within 30 days shown as Conversion Funnel for the last 30 days.
In the chart, we see that there were 165,241 users who fired the event 'Search Song or Video' in the last 30 days. Of the165,241 users, 161,696 of them fired 'Play Song or Video' within 30 days of searching for a song or video. And of those 165,241 users, 26,423 of them fired 'Share Song or Video' within 30 days of 'Search Song or Video.'
Holding a Property Constant
When not holding properties constant, the funnel chart will share the unique count of users who have gone through the funnel once or more. This means that if the user goes through the entire funnel multiple times, the user is only counted once. For example, given the following funnel, if a user were to perform 'SearchSong' -> 'PlaySong' -> 'ShareSong' ten times in the last 30 days, they would only show up once in this chart.
When holding properties constant, the funnels chart will share the unique count of user and user/event property pairs that have completed the funnel. If a user goes through the entire funnel N times with M distinct event property values, the user will be counted M times. For example, if a user does 'SearchSong' -> 'PlaySong' -> 'ShareSong' with ten different 'SongName' property values, then they will show up ten times in the following chart. In this example, 'SongName' is an event property that has been sent for all three events in the funnel. An event property can only be held constant if you have instrumented it for every event in the funnel.
Exclusion steps allow you to exclude users who perform selected events between steps in a funnel. This provides a deeper understanding of how selected behaviors impact your target conversion rates.
To exclude events, you can click on the "exclude users who performed" button, and use the dropdown to select the excluded event. You can apply the exclusion between all steps in the funnel, or between two specific steps. For "any order" funnels, users are excluded if they perform the exclusion event between any of the funnel steps.
Event Conversion / Drop-Off
The bar graph shows each step in the funnel and the number of users who converted at each step.
- Solid regions: These represent the users who have successfully converted to or reached that step.
- Striped regions: These represent the users who dropped off at that step or have not yet reached that step.
Like all of Amplitude's chart types, you can use the date picker to choose a more specific time range to analyze your data and can switch between "Last", "Between", and "Since". You also have the option to view data in daily, weekly, monthly, or quarterly units by toggling between the different options in the dropdown menu next to the date picker.
Compare to past
On the top, left display area of the chart display area, there is the Compare to past where you can add a comparison of your current analysis to a specific day in the past. This feature is currently available in Segmentation and Funnel chart types. The options available are: Previous day, Previous week, Previous month, Previous quarter, Previous year or by setting a specific, custom date.
If you are looking at multiple segments in your chart, you can manually select and deselect each segment by hovering over the segment name in the bar below the chart and removing it or by clicking the "+" button to add it back. Finally, you can click on any data point in the chart and inspect the users that make up that data point by using Microscope.
Breakdown Data Table
Underneath the chart is a table of the data displayed. If you create a Group By, you can select or deselect which segments you see in the graph by clicking on the segment name in the data table. Here are some helpful definitions:
- Conversion: This is the percentage of users who successfully completed the entire funnel. This is calculated by dividing the number of users who successfully exited the funnel by the number of users who entered the funnel.
- Event Name: The number of users who complete that step in the funnel. The first step will always be 100% because a funnel only considers users who performed that first event (not all users).
- Average Time: The average time it takes users to move from one event to another event in the funnel. The average time is a function of the time of users' first conversion. So, if one user did Event A and Event B multiple times, the average only uses the time of the user's first "Event A-> Event B" instance. This cell will also show the median time it takes users to move from one event to another event. If the funnel is "any order", then the average time reflects the absolute value of the difference since it could be a negative number.
You can also export the table as a CSV file by clicking the "Export CSV" button.