Funnels

Follow

Funnels in Countly are an intuitive way of tracking and analyzing user behavior. This guide provides a quick overview of their setup and capabilities.

Availability

The Funnels plugin is available only in the Enterprise Edition.

Funnels are used to track the completion rates of a step-by-step journey inside your application. These steps or goals are defined as events in Countly and no additional or extra API calls are necessary.

Understanding Funnels

Funnels are an ordered grouping of events that helps you see how users flow through your mobile or web application based on the app events you determine.

Countly Funnels are retroactive, meaning that you can create your funnel in the present and still be able to see the data for a past period. However, funnels depend entirely on events, so first, you need to collect events as they are the basic building elements for a funnel.

In order for a user to complete a funnel successfully, they should have progressed through all the previous steps in the given order. For example, if a funnel is defined as Register > Share media > Proceed to Checkout, then any series of events that contain this sequence (e.g. first event before the second one, and second event before the third one) leads to a funnel conversion. Note that any number of unrelated events may occur between any two funnel steps.

Funnels are not bound to sessions or time limitations. This means that if a user completes one event in one session and other events of the same funnel in a different session, the funnel would still be considered as complete, regardless of the time that passed between sessions.

Creating a Funnel

First of all, you must enable Funnels in the Countly Dashboard by going to Management > Plugins in the top-right corner. You will then find Funnels under the Behavior category of the Explore section of your Dashboard.

empty_m.png If you have not created a funnel yet, you will see an empty funnel page with a description. You need to click on the Create a Funnel button on the top-right corner of the Funnels overview page. 

If you have already created a funnel before, they will be visible on this overview page and can continue working with them.

A Simple Funnel

When you click on the Create a Funnel button, the Create a Funnel drawer will prompt.

Funnels require a minimum of 2 steps and up to 8 steps which is configurable from Management > Configurations. You can increase the number of steps as you like, and you can discover at which step users are stuck or getting problems.step_q.pngFunnel Name: You must give your funnel a name to create it, while funnel description is optional. 

Funnel Type: This helps you configure the session dependence of your funnel.

Funnel Steps: You can define the steps of your funnel by either inputting the event key you plan to send in the future or selecting one from your existing events. step2_crop.png

When you select an existing event as the first step, the add property link button will appear, letting you add a segmentation filter for your funnel step. For example, if you add the segmentation filter count is 2, you will see funnel users who completed this particular event two times. Similarly, you can select the City user property to filter the completion of the event by location. You can keep adding segmentation filters to each step of the funnel for more precise targeting.

The order of steps does matter.

For example, say you have an e-commerce website. When your users enter a product page ("Detail View Level 1") you want to know the selling rate of this product for users from a specific country.

drawer_crop.png

The steps you need to set are:

  1. Define an event for the buttons to this product page
  2. Define an event for the "add to cart" button on this page
  3. Define an event to track when it goes to the checkout page
  4. Define an event for when the order is completed

Session Dependence 

By default, funnels capture all users who complete a series of events, regardless of the session to which those events belong. In other words, events taking place in the exact order you determined is enough for considering that a user completed that funnel. These type of funnels are called "Session Independent".

In some cases, however, you might want all events of the funnel to occur during the same session. For those cases, funnels can be set as "Same Session". Here, funnel users must complete all steps through the funnel within the same session they entered the funnel to be counted as completed. If the session ends before the last step, the user will be dropped out of the funnel right after the last step they reached. 

You can specify the strategy you prefer when creating a funnel. In addition, this option can be changed after the funnel has been created. So you are allowed to change it for existing funnels.

Step Variations

A user must do all steps of the funnel to be regarded as having completed the funnel. In some cases, however, this can be too restrictive.

For instance, you may need a 3-steps funnel, where the second step can be either adding an item to cart or redeeming a promotion code. In other words, you may need to define a funnel that accepts both of these paths:  

  1. View > Add to Cart > Checkout
  2. View > Promo Code > Checkout

multistep.png

In order to accomplish that, you simply change AND, which is between Add To Cart and Promo Code, to OR. This will merge both steps and convert them into a group of steps. Consequently, users who have done at least one of the above paths (1 or 2) will complete the same funnel. 

You can add multiple steps to a group, and create more groups as well. Regardless of the number of groups you are still bounded by 8 steps limit (which again, can be changed from Management > Configurations).

Funnels in Action

Overview lets you take a look at your existing funnels to learn about how they are doing. You can see how many users entered that day, how many of them completed it that day, and a 30-day trend completion rate graph for each funnel. You can edit or delete a funnel by using the ellipsis menu on the right of each row. By clicking on a row, you can navigate to the funnel's detailed view.overview.pngIn this detailed view, you will see the total number of users, with a toggle to filter them down if necessary, how many users of your application entered this funnel by performing the first step event, and the success rate and number of users who have completed (or didn't complete) this funnel.

Each step has its own rates of completion as well as the number of users related to each rate. The average time spent between each step is shown on the vertical timeline (e.g. ~50 minutes) as well.detail_m.pngBy default, the number of completed users is displayed relative to the number of entered users. In addition to that, you can make it relative to all the app users, by simply selecting all users as demonstrated below. all_users_m.pngWhen you click on the number of users in each step, you will be redirected to their User Profiles. There you can see all of those users who have or not completed that particular step and analyze what went wrong to cause them to fall out of the funnel at the given step.

You can also observe the evolution of the steps you have set between specific dates.

period_m.png

For example, you want to set the purchase rates for a specific date range that you have campaigned for Product_1. All you have to do is pick these dates from the top right corner.

You will be able to determine your users who were interested in this product but did not buy it and reach out to them with a push message before the end of the campaign. You can also use the toggle to filter the first step by the segmentation properties and user properties that apply to the event.

query_m.png
If your funnel has a grouped step (i.e. containing an OR), it will be displayed as below:multid_m.png

Funnel Analytics

The most common use cases of funnel analytics are:

  • Tracking revenue conversions, understanding the paths that lead to buying a product, and optimizing them.
  • Understanding in which level or part of your application or game users tend to leave.
  • Performing A/B Test analysis to see which versions perform better.
Was this article helpful?
0 out of 0 found this helpful

Looking for help?