Funnels

Overview

Mixpanel’s Funnels allows you to examine how end-users perform events in a series. Funnels calculate and display the amount of users who convert from one event to another within a particular time window. This allows you to identify where your users drop off, what segments convert the most, and other important facets of the user journey.

Use Cases

Imagine your product is a B2B messaging application. You might use Funnels to answer these questions:

• What percent of users converted through my signup funnel within 7 days?
• At what step of the signup funnel did most users drop off?
• How did my A/B test impact conversions in the signup funnel?
• How has the payment funnel conversion rate in the US changed over time?
• How long does it take most users to complete my payment funnel?
• What departments complete the payment funnel most often?
• What flows do users take between opening an app and making a purchase?
• Why did the successful users purchase?
• What flows do users take that don’t lead to a purchase?
• How do these two paths differ? What actions should I nudge towards or against?
• What did the users that dropped-off do instead?

Quick Start

Step 1: Define your Funnel Behavior

A funnel is the basic building block of the Funnels report. A funnel is sequence of events done within a certain period of time. Using the example above, imagine we know the signup flow contains five steps. Within the funnel definition, add one event for each step, in the following order: "Landing Page," "Download Page," "App Install," "App Open," "Sign Up." At this point, your query should look like this.

Step 2: Choose your Measurement

After your funnel is defined, you can choose how you want to measure the funnel. By default, Funnels report will select conversion rate of the funnel as the measurement, but you can also measure the number of users converting, total conversions, and more. You can change the measurement by selecting "Conversion Rate" in the bottom left of the funnel metric definition. In this case, we don't want to change the measurement, so it can be left alone.

Step 3: Choose Filters

Filters exclude unwanted data. In this case, we only care about events performed on the iOS platform. Therefore, add a "Platform" filter, where Platform equals "iOS Native". At this point, your query should look like this:

Step 4: Choose Breakdowns

Breakdowns segment data into groups. In this case, we want to break our funnel down by marketing medium, tracked via UTM tags. Therefore, add a "UTM Medium" breakdown. At this point, your query should look like this:

Step 5: Choose Visualization

You can visualize your metric in a variety of ways. Mixpanel allows you to choose how you would like to visualize your data. Select "Funnel Steps" in the top right to change your visualization. In Funnels, you can choose:

• Funnel Steps - see how users are progressing through the funnel
• Line - see how your metric is trending over time
• Bar - see a summary view of your metric
• Metric - see a summary view of your metric

FAQ

Why is my user shown as dropped off even though they've just completed the full funnel today?

Assuming that the user performed all actions in the funnel, on their first try, in the right sequence, within the conversion window, the user should be counted as converted. However, if you are looking at the Funnel report for today, they might be shown as dropped off. This is because there may be up to a 24 hour delay on our reports (except for Activity Feed), as we work towards linking anonymous user activity to known profiles. This would explain why the funnel displays the user as not converted, soon after ingestion. Once we account for the delay, the funnel should display the conversion, especially if the user performs the steps in the funnel in the specified sequence.

This can also happen due to identity management issues. Mixpanel uses distinct_id to keep track of user events throughout their conversion journeys, and if the distinct ID on events tracked by a single user is not consistent, they are misattributed and appear as a drop off.

Can users do other actions during the funnel?

Users can complete the steps you designate in your funnel in loose order. Loose order means that a customer can engage in other actions in between funnel steps, as long as they complete all the funnel steps in order. Let's start with an example where the funnel has steps: A, B, C, D, E and go through a few cases:

1. The customer does steps A -> B -> C -> D -> E in exact order. Mixpanel counts this as a conversion.
2. The customer does steps A -> B -> F -> C -> D -> E. Mixpanel counts this as a conversion. This is an example of loose ordering.
3. The customer does steps A -> B -> C -> E. Mixpanel will not count this as a full conversion, and the customer will not appear in the funnel after step C. The customer's completion of step E is excluded from the funnel because step D did not occur.

Why do I get differing numbers when comparing Funnels and Insights?

Funnels and Insights are fundamentally different when it comes to counting events. In the Funnels report, users will need to perform Step 1 event, in order for their Step 2 event to be counted. Insights, on the other hand, will count the Step 2 event, even if the user does not perform the Step 1 event. This will create some discrepancy between the Insights and Funnels report, as users are expected to trigger events in sequence in the Funnels report, whereas, they are not expected to do so in the Insights report.

Can multiple events count towards the same Funnel step?

Custom events were designed to specifically target this exact issue. With custom events, you can essentially "combine" events into a single event so that you can create Funnels that measure the following:

Step 1: Event A OR Event B

Step 2: Event C

This allows users to make multiple paths to the same end goal. Learn more about creating custom events here.

How does Mixpanel handle simultaneous events?

Mixpanel Funnels include a built-in two second grace period. This means any consecutive steps that have timestamps within two seconds of one another are interchangeable. This window allows for inconsistencies that arise with cross-platform tracking, batch event sending, mobile time tracking, or variations in connection speed.

What does this mean in terms of your funnel? In a funnel with step A -> B, if a user completes event B first and then performs event A within two seconds, Mixpanel counts this as a conversion from A -> B or B -> A.

How does Mixpanel calculate statistical significance?

Note

This section is intended for users who want to understand the mathematics involved in statistical significance in depth.

To determine whether a particular segment’s conversion rate is significantly different from the overall conversion rate, we use a hypergeometric distribution to calculate statistical significance. The hypergeometric distribution is used to model the probability of picking k items of a particular type in n attempts without replacement from a population of size N having K items of the same type.

For example, let’s say we have a sock drawer with 20 socks, 10 blue and 10 red. If we randomly picked 10 socks one at time from the drawer without putting them back between picks, and we wanted to know the probability of 9 of those socks being red and 1 of them being blue, we would use a hypergeometric distribution to calculate that.

This is applied to funnels by considering the total number of users who enter the funnel to be a finite population of size N, out of which a subset of users convert (K). We then estimate the probability of getting k conversions in a particular segment (given that there were n users who entered the funnel in that segment) if users in that particular segment were picked at random from the overall user set. The higher the probability, the higher likelihood that variations we see in conversion rate are due to random chance.

To calculate the actual p-value, we estimate the hypergeometric cumulative distribution function (CDF) for N, K, n.

In the CDF, the value of any point (X) represents the probability that a random draw would result in fewer conversions P(k < X). 1 - P(k < X) represents the probability that a random draw would result in more conversions P(k >= X).

These two probabilities are used to represent the probability that the selected segment will either outperform (P(k < X)) or underperform (P(k > X)) the overall set of users. Mixpanel takes the higher probability of the two, and calculate the p-value as 1 - max(P(outperform), P(underperform)).