---
metadata:
- name: generator
content: Diplodoc Platform v5.52.0
alternate:
- https://yandex.com/dev/games/doc/en/console/ab-builds.md
- https://yandex.com/dev/games/doc/hi/console/ab-builds.md
- https://yandex.com/dev/games/doc/ko/console/ab-builds.md
- https://yandex.com/dev/games/doc/ru/console/ab-builds.md
- https://yandex.com/dev/games/doc/tr/console/ab-builds.md
- https://yandex.com/dev/games/doc/vi/console/ab-builds.md
- https://yandex.com/dev/games/doc/zh/console/ab-builds.md
- href: en/console/ab-builds.md
type: text/markdown
title: Markdown version
- href: ../llms.txt
type: text/markdown
title: llms.txt
---
> **Documentation Index:** Fetch the complete configuration index at https://yandex.com/dev/games/doc/en/llms.txt
# Builds A/B tests
To see how players react to game updates, run an A/B experiment. Through the Developer Console, [add](#preparation) a new game build. Specify the audience percentage and experiment duration. When testing concludes, compare the [metrics](#launch) of the experimental build with the main version’s performance to decide whether to update your app.
{% note warning %}
You can only launch an experiment if the game is already published.
If something else affects the testing, the results will be [unreliable](*experiment). To avoid this, don't publish new game versions in parallel with the experiment and don't run other A/B tests.
{% endnote %}
To study how smaller changes affect game metrics, run an [A/B experiment with flags](https://yandex.com/dev/games/doc/en/console/vq-test.md).
## Step 1. Preparation {#preparation}
Prepare a new, experimental build. It must comply with the [platform requirements](https://yandex.com/dev/games/doc/en/concepts/requirements.md) and [draft materials](*draft).
Upload it via the [Yandex Games Console](https://games.yandex.com/console){.external}:
1. Select your game.
1. Navigate to **Builds A/B Tests** tab.
1. Click **Start A/B Test**.
1. Complete **Experiment name** and **What’s new** fields. Describe differences from the main version — this speeds up moderation.
1. Select the desired values in the [Audience](*audience) and [Duration](*duration) fields.
1. Click **Choose file** to upload the game [archive](*archive).
1. Click **Submit request**. Use **Edit submission** if needed.
1. Wait for the build check. Refresh the page in approximately 5 minutes:
- If passed: **Build passed the check** appears.
- If failed: Click **Back to submission**, fix issues, and reupload.
1. Click **Submit for moderation**.
{% note warning %}
Only one moderation can be in progress at a time: [draft](https://yandex.com/dev/games/doc/en/console/add-new-game.md#step-moderation-game), [creative A/B test](https://yandex.com/dev/games/doc/en/concepts/ab-test.md#submit-for-moderation), or build A/B test. While the build for an A/B test is being reviewed, other types of moderation will be unavailable.
{% endnote %}
## Step 2. Moderation {#moderation}
{% note info %}
You can withdraw a submission within the first 2 hours.
{% endnote %}
Results appear in **Builds A/B Tests** within 3–5 days:
{% list tabs %}
- A/B test launched
The experiment started automatically: the selected percentage of users began seeing the new, experimental version of the game, while the rest continued seeing the regular version.
- Version rejected
Moderation rejected the new build because the changes made to the game violate the [requirements](https://yandex.com/dev/games/doc/en/concepts/requirements.md). An email from moderation will be sent to the address specified in your Yandex ID, indicating the specific reasons for rejection with links to the violated Game requirements items. Moderation will attach files with recordings or screenshots of the violations.
Before resubmitting the game for review:
- If there are clarifications or testing methodologies for the violated requirements, familiarize yourself with them.
- Make changes to the game.
- Carefully check the game again for compliance with the requirements that were violated last time.
- If possible, test the game on devices similar to those used for moderation. The models of these devices will be specified in the notifications.
If you have questions about the release comments, contact [support](https://yandex.com/chat/#/user/a4fa5c06-75db-9b38-6eea-b1673785f7d5){.external}.
The number of attempts is unlimited, but the waiting time before resubmitting for moderation doubles after each rejection. The maximum waiting time is 384 hours (16 days).
You can withdraw the draft within 2 hours after submitting it for moderation. Waiting time before resubmitting for moderation:
- After the 1st rejection — 24 hours.
- After the 2nd rejection — 2 days.
- After the 3rd rejection — 4 days.
- After the 4th rejection — 8 days.
- After the 5th rejection — 16 days.
{% endlist %}
## Step 3. Launch {#launch}
After the experiment launches, testing data will appear on the **Builds A/B Tests** tab: the experiment start date and the date when it will automatically end. The first test results will appear in a few days and will be displayed in graphs and tables.
Comparison metrics:
#|
|| **Metric** | **Description** ||
|| **Timespent per player** | Average time in minutes a player spends in the game per day. ||
|| **Conversion To Play** | Percentage of game sessions lasting more than 60 seconds. ||
|| **Interstitial shows per player** | Average number of [interstitial ads](https://yandex.com/dev/games/doc/en/console/adv-monetization.md#interstitial) shown per player per day. ||
|| **Rewarded shows per player** | Average number of [rewarded ads](https://yandex.com/dev/games/doc/en/console/adv-monetization.md#rewarded) shown per player per day. ||
|| **Ad revenue delta*** | Difference in Yandex Advertising Network revenue between test and control groups, as a percentage of control group revenue. ||
|| **In-app purchases per player** | Average number of [in-app purchases](https://yandex.com/dev/games/doc/en/console/purchases.md) per player per day. ||
|| **Ratio of players with in-app purchase*** | Percentage of paying users from the app's daily audience. ||
|| **In-App revenue per player*** | In-app purchase revenue per player per day (in rubles). ||
|#
* Metric visibility is restricted to game owners and developers with **View income** permissions.
Color coding:
- Green: positive [statistically significant](*positive) result;
- Red: negative [statistically significant](*negative) result.
If metrics are not colored in any way, it is impossible to definitively determine whether the experiment affects the user.
{% note warning %}
For reliable player activity data, run experiments for at least one week.
{% endnote %}
The countdown will begin after moderation review. The experiment will automatically end after the time you selected in the **Duration** field. To end it early, click **Stop** → **Yes, stop it**.
## Step 4. Results {#results}
After the experiment ends, click:
- **Publish** to make the new build the main one. You don't need to go through moderation again.
- **Don’t publish** to end the experiment and stop showing users the experimental build.
After making your decision, click **Close**.
[*experiment]: Since user behavior will be influenced by multiple factors, it will be impossible to determine the role of the originally tested changes.
[*draft]: The new build must match the genre, age rating, tags, and promotional materials specified in the draft.
[*audience]: Available options:
- 10% (5%/5%);
- 20% (10%/10%);
- 50% (25%/25%);
- 100% (50%/50%).
[*duration]: Available options:
- 7 days;
- 14 days;
- 21 days;
- 28 days.
[*archive]: The ZIP archive must:
- be no larger than 100 MB;
- contain only one `index.html` file.
[*positive]: With 95% probability, the metric improvement is related to changes in the new build.
[*negative]: With 95% probability, the metric decline is related to changes in the new build.