# A/B Testing

## Overview

<figure><img src="https://246234927-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FkSyCQUyE0u7ya9ULaprd%2Fuploads%2Fng0g0hChBd39BtrLqBFu%2FScherm%C2%ADafbeelding%202025-02-19%20om%2014.34.55.png?alt=media&#x26;token=c86ef4cc-dd0f-4a65-9205-eafe6cb02c2e" alt=""><figcaption></figcaption></figure>

***

## Summary of test

<figure><img src="https://246234927-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FkSyCQUyE0u7ya9ULaprd%2Fuploads%2FmrahRsiORYqyd6csj3gQ%2FScherm%C2%ADafbeelding%202025-02-19%20om%2014.46.11.png?alt=media&#x26;token=13b7582c-5a69-433c-9de6-e3b1cbfb141c" alt=""><figcaption></figcaption></figure>

{% hint style="info" %}
This overview will show you the following:

* **Group** → when we do an a/b test we will be make different groups (emails, flows, etc) to find out what leads to the least amount of churners.&#x20;
* **Contacts** → the numbers revere to the amount of people that entered that group.&#x20;
* **Churn** → this will show you the amount of people that churned in their group.
* **Churn** **Rate** → this percentage will show you what percentage of people churned that entered the group.
* **Churn** **Loss** → the amount of money you lost based on the value of the people that have churned.
* **Avg.** **loss** → this is based on the following formula: *churn loss/contacts = avg. churn loss*
  {% endhint %}

$$
Churn Rate = Churners/Contacts\*100
$$

***

## Winner

<figure><img src="https://246234927-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FkSyCQUyE0u7ya9ULaprd%2Fuploads%2FYEe32rLXpfErrWFG0UGy%2FScherm%C2%ADafbeelding%202025-02-19%20om%2015.30.10.png?alt=media&#x26;token=80bcf7cb-d1b7-48af-838e-c7de6666b8fa" alt=""><figcaption></figcaption></figure>

{% hint style="info" %}
The **winner** is based on the group that had the lowest churn rate
{% endhint %}

***

## Churn Rate Graph

<div><figure><img src="https://246234927-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FkSyCQUyE0u7ya9ULaprd%2Fuploads%2F36pErULbDgXHWgAyrXYp%2FScherm%C2%ADafbeelding%202025-02-19%20om%2016.08.05.png?alt=media&#x26;token=27624dcc-bb19-4217-805f-4495962e306e" alt=""><figcaption></figcaption></figure> <figure><img src="https://246234927-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FkSyCQUyE0u7ya9ULaprd%2Fuploads%2FYpHPTqHyEnx8f7uF0Gcb%2FScherm%C2%ADafbeelding%202025-02-19%20om%2016.08.32.png?alt=media&#x26;token=0b69a9c0-ba77-4b1f-9ca6-690349792ff1" alt=""><figcaption></figcaption></figure></div>

{% hint style="info" %}
The **Churn Rate graph** is interactive. This way you can look at the development and the process of the a/b test.&#x20;
{% endhint %}

***

## Average Churn rate per segment

**You can view the chart in:**\
&#x20;    \- Churn Rate \
&#x20;    \- Avg. Churn Loss

### Churn Rate

<figure><img src="https://246234927-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FkSyCQUyE0u7ya9ULaprd%2Fuploads%2FZjU0dYOhSAzsGlqf0Goo%2FScherm%C2%ADafbeelding%202025-02-19%20om%2015.50.31.png?alt=media&#x26;token=82249d6a-9815-47ff-9e34-264d5eccfafc" alt=""><figcaption></figcaption></figure>

{% hint style="info" %}
The **Average Churn rate per segment** will display different percentages that are based on the customers that have churned. This can display a lot of different factors like:

* **Channels** → Mail, Inbound and Website, etc.
* **Gender** → Male, Female, Neutral, etc.
* **Product** → Online, Physical, etc.
* **Client** **Type** → Individual, Subscription, etc.
* **Engagement** → Daily user, Weekly user, Monthly user, No engagement, etc.
  {% endhint %}

<figure><img src="https://246234927-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FkSyCQUyE0u7ya9ULaprd%2Fuploads%2FsR1fbDgnIi7XApJQUHoN%2FScherm%C2%ADafbeelding%202025-02-19%20om%2016.11.37.png?alt=media&#x26;token=cfe4bff4-837f-4ddc-900b-2d3e5b4cbf3f" alt=""><figcaption></figcaption></figure>

> You can use these graphs to dive a little deeper into who the people are that churned. This will give you more insight on what there behavior was and what you should focus on to prevent churn.

### Avg. Churn Loss

<figure><img src="https://246234927-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FkSyCQUyE0u7ya9ULaprd%2Fuploads%2Flaup2nM90h4TtBAJzc2k%2FScherm%C2%ADafbeelding%202025-02-19%20om%2016.22.24.png?alt=media&#x26;token=dce358fd-e47c-4b0c-8d47-8c7404ff11db" alt=""><figcaption></figcaption></figure>

{% hint style="info" %}
The **average churn loss** is a way to show you were you will lose the most amount of money. This can help with understanding which customers are going to cost you less or more money.&#x20;
{% endhint %}

***

## Different tests

<div align="left"><figure><img src="https://246234927-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FkSyCQUyE0u7ya9ULaprd%2Fuploads%2FSCfzxAYflPgvXF2QEgF5%2FScherm%C2%ADafbeelding%202025-02-19%20om%2015.37.59.png?alt=media&#x26;token=215fc915-79ac-47fa-8909-3a4a6bfbb480" alt="" width="426"><figcaption></figcaption></figure></div>

{% hint style="info" %}
You can have **multiple a/b tests**, they don't have the be live at the same time but you will be able to look back and review their results.
{% endhint %}

***