Conversion Drivers surfaces the user actions most correlated with a specific outcome or conversion event that occur between two steps in a funnel. With this feature, all you have to do is set a starting event and an outcome or conversion event by creating a two-step funnel. Amplitude will then automatically sort through, aggregate, and analyze all the events that occur for each user between those two steps.
To help you identify the ones that are most relevant, Amplitude provides a correlation score, the frequency with which the behavior is performed, the percentage of users taking that behavior, and the overall time to convert when that behavior is one of the steps. This helps you understand the frequency of different actions being performed, and if they serve as an accelerator or decelerator to conversion.
Use this feature to understand which behaviors are driving key outcomes in your customer journey. Build conviction on where to prioritize your team's efforts:
- Instantly find behaviors that lead users’ to sign-up or drop-off
- Uncover friction points in onboarding and critical conversion funnels
- Find the actions most likely to cause cart abandonment
- Discover common experiences that lead to repeat consumers
What You Will Learn in This Article
This article is an in-depth guide of how to use Conversion Drivers within the Funnel Analysis chart. It is part three of our Funnel Analysis help center articles. We will cover where to find and how to use the Conversion Drivers feature and what the correlation score means.
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 useful information and good to know practices:
- This feature is only available to Enterprise, Growth, and Scholarship plans
- An understanding of the Funnel Analysis chart is needed. Please see our Funnels Analysis - Getting Started article
- To make sure you properly use and understand this feature, please keep in mind that: correlation does not equal causation
- This feature will only work for Funnel charts that are setup for the Conversion metric, and the order of the events is to to: In This Order
- Funnel charts using: In Any Order or This Exact Order do not support this feature
- We have another article for Funnel Analysis' other features for reference: Funnels Analysis - Advanced
Table of Contents
- Analyze Events Performed Between Funnel Steps
- Understanding the Correlation Score
- Sharing the Report
Analyze Events Performed Between Funnel Steps
Within each Funnels chart, click into any of the steps after the initial event to open up the Conversion Drivers option. From there, you can look at the actions performed by users between steps in the funnel. This helps to clearly identify potential drivers of conversion, or drop-off.
Within Conversion Drivers, the step controller (see screenshot below) allows the option to choose the steps of the funnel to look between. The conversion and drop-off numbers show the rates between the two selected steps.
Below the step controller, there is a table of the events users have performed between the two selected steps with a number of metrics shown. At the top of the table, you can choose to look at the event list for users who converted, or users who dropped-off in the funnel.
The event table shows a series of metrics along with the ability to hide an event from the visualization if determined that it is not relevant to your analysis:
- Correlation Score: Correlation means that that there's a relationship between two things. In this context, the correlation column helps to understand if there is a relationship between the event in question and conversion (or drop-off), depending on which tab you selected (Converted vs. Dropped Off). For more information, read the next section on "Understanding Correlation."
- Frequency: The average number of times a given event was performed between the two selected funnel steps.
- % Who Did Event: The percentage (and absolute numbers) of users in the selected cohort that performed a given event.
- Time Between Steps: The amount of time it took users who performed a given event to convert between the two selected funnel steps. This metric is useful to help understand if performing a given event serves as an accelerator, or decelerator to conversion.
For users who are considered converted, we look at the events performed between the timestamps of the two selected funnel steps. For users who are considered dropped-off, we look at the timestamps of the first selected funnel step, and their entry into the funnel plus the conversion window.
Imagine a funnel defined as A --> B --> C, and you wanted to investigate drivers of conversion at step C. The time periods analyzed each set of users is shown below, where t() represents the timestamp of the event performed:
|t(b), t(c)||t(b), t(a)+ conversion window|
Understanding the Correlation Score
Clicking the “correlation data” button will expand a panel that will see a detailed confusion matrix that shows the count of users in your base cohort who constitute your True Positives (TP), False Positives (FP), False Negatives (FN), and True Negatives (TN).
Correlation is a measure (ranging from -1 to 1) of how two statistical variables relate to each other. In Conversion Drivers, the variables for each user are whether or not the user performed the selected event and whether or not the user was in the selected cohort (converted or dropped off). You may have heard of different variations and definitions of correlation, including Matthews correlation, Pearson correlation, phi coefficient, and R-value. In this case, all of these definitions are equivalent because Conversion Drivers looks at pairs of binary random variables.
For example, in the screenshot above we have selected "Converted" and see a "strong positive" correlation for the "Add Content to Basket" event. This means that there's a strong relationship for those users who perform "Add Content to Basket" and conversion. Therefore, "Add Content to Basket" could be a conversion driver.
Remember, correlation is not causation so hypotheses generated by Conversion Drivers still must be tested and verified in the real world. Here are some more technical intuitive definitions of correlation:
- Correlation of X and Y is the covariance of X and Y divided by the geometric mean of their variances;
- If X is modeled as an affine function of Y and Y is modeled as an affine function of X, each with minimal root mean squared error, then the correlation of X and Y is the geometric mean of the predictive coefficients of these two functions.
Sharing the Report
When you find a valuable insight using Conversion Drivers, it can be valuable to share that with your colleagues and talk through various strategies of improving conversion. To share Conversion Drivers with a teammate, click the “Share” button in the top right. Clicking “Copy Chart Link” will copy a unique URL for the chart to your clipboard and allow you to send your analysis with others.