Customize Predictions and Boost App Growth - Huawei Developers

"And then even the mightiest company is in trouble if it has not worked on the future."
- Peter Drucker
The soaring growth of the mobile Internet brings both a trend of homogeneity in products and fierce competition. Analyzing users with session path analysis allows you to tap into the value of your users and guide their conversion in the lifecycle.
HUAWEI Prediction anticipates user behavior in advance based on user behavior data, helping you make smart and quick decisions.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
1. What is custom prediction?
In our previous posts, we introduced some prediction tasks provided by Prediction. So, what's special about the custom prediction task? Simply put, custom prediction tasks can be used in multiple scenarios.
User churn and payment is an important factor to determine whether a product can be monetized successfully, which is a key concern during product operations. You may also take into consideration other in-app events and their conversion nodes, such as the click-through rate of an in-app ad, or the probability of a player passing a game level. With the help of custom prediction tasks, you can evaluate the probability of user behavior.
Sourced from diversified data reported by your app, and similar to the preset prediction tasks, a custom prediction task uses the active user data from the previous two weeks to train a model, which then predicts the probability that active users from the previous week perform a certain behavior over the next week.
2. Method for starting a custom prediction task as required
Different operations solutions apply to different industries and scenarios. For example, an e-commerce app wants to attract more users and for users to add products to their shopping carts or tap an ad on the home screen, and a game app may want to know the probability of a player passing a level or consuming coins and items.
This post takes game apps as an example for evaluating the probability of a player passing a level using the custom prediction.
To achieve this, your app needs to report game level completion events. First, sign in to AppGallery Connect, click My projects, find your app, go to HUAWEI Analytics > Behavior analysis > Event analysis, and create a COMPLETELEVEL event, which is a preset event. Then mark it as a conversion event, go to Prediction, and create a custom prediction task. After you have selected the COMPLETELEVEL event, click OK.
Currently, you can create custom prediction tasks for automatically collected events, preset events, and custom events in HUAWEI Analytics in the same way.
3. Use cases of custom prediction task results
Since we have introduced the creation of a custom prediction task, let's see how its result can help with your business growth.
From the detailed result, it can be seen that the number of users predicted to have a low probability of completing the game level share a similar location and app version. To solve this problem, you can release different app versions for different countries and regions. Alternatively, you can develop an operations plan to improve the audience's experience and aim to turn them into paid users.
* The screenshot is from the custom prediction details page of HUAWEI Prediction. The data is for reference only.
Your game can recommend improved equipment and game coins to these players, as well as increase the probability of them making in-app purchases. By guiding the players to complete payments, you can turn minnows into dolphins and, eventually, whales.
Implementing this method couldn't be simpler. Prediction can work with AppGallery Connect Remote Configuration without you having to do extra coding.
On the Remote Configuration page in AppGallery Connect, you can choose to display item recommendation pop-ups just to players who are less likely to complete the game for promoting payment conversions. You can also attract players by setting a time limit on a discount for game coins. When your prediction task is updated, its audience is also updated, allowing you to reach target players without disturbing other players.
To learn more about HUAWEI Prediction, feel free to check out this document.

Related

How to Tailor Mobile App Content and Styles Through User Segmentation

To improve user retention in your app, it is crucial to optimize user experience and improve user loyalty since it is becoming more and more difficult and expensive to increase traffic and acquire new users. One solution is to customize the content, appearance, and style of your mobile app for different users.
HUAWEI Analytics Kit, together with Remote Configuration, can provide such functions in a convenient and efficient way.
What are HUAWEI Analytics Kit and Remote Configuration?
HUAWEI Analytics Kit is a free-to-use data analysis service for app operations personnel to track how users behave in apps and facilitate precise data-driven operations. Applicable to multiple platforms such as Android, iOS, and web, and various types of devices such as mobile phones and tablets, it can automatically generate more than 10 types of analysis reports based on users' behavior events.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Remote Configuration provides a console and a client SDK, and periodically obtains configuration items from the cloud. You can then use the user attribute and audience information provided by Analytics Kit to easily tailor app content and appearances for different audiences, without the hassle of releasing a new app version. Remote Configuration allows you to perform the following:
Flexibly control the target user groups to which configuration items take effect, in dimensions such as app version, operating system, country or region, language, time, audience, and user attribute.
Manage up to 300 configuration versions from the last 90 days, and perform one-click rollback to a specified version.
Push configuration changes in real time to speed up app updates on devices.
Examples of scenarios where HUAWEI Analytics Kit and Remote Configuration are applicable
Example 1: HUAWEI Analytics Kit found that the top 3 types of active users in a video app were animation, costume drama, and suspense drama fans, respectively. Therefore, different app styles were used depending on the user type, helping improve user activity.
Example 2: After HUAWEI Analytics Kit was integrated into an e-commerce app, it was found that purchase behavior differed greatly depending on the age of the user. Therefore, the app developer decided to change the appearance of the app and the products displayed depending on the age of the user, in order to provide the user with a better experience and increase in-app purchases.
It takes you only 5 minutes to integrate HUAWEI Analytics Kit, which allows you to precisely segments users and customize app content and styles accordingly.
Integration guide:
Android
iOS
Web
Sample code:
Android
iOS
Web
If you encounter any problems during the integration, you can submit a ticket online.
We look forward to your participation!

Predict Users with High Value and Send Them In-App Messages

A product will go through many stages in its lifetime. Of these, product maturity involves long-term exploration of how to mine value from existing users for monetization. At this stage, operations personnel commonly launch promotions to encourage purchases and cultivate regular payments. These activities advertise themselves with in-app and push notifications. In-app messages reach users in the app and appear in diverse formats, such as modal, banner, and image. Such messages redirect users to activity landing pages with purchase options, a shortcut to monetization.
However, in-app messages are not always welcome. Not everyone wants their app usage interrupted, and their affected experience can cause user churn. To wisely wield this double-edged sword, operations personnel should first target audiences by attribute and behavior so in-app messages they send are tailored to the audience. Since users in these audiences are more willing to make purchases, they are considered of high value in these operations activities.
Let's learn how a tool app attempted to grow revenue from member subscriptions. It sent its users a daily modal in-app message at a scheduled time, but this activity resulted in a less than 0.1% payment conversion rate. Operations personnel then enabled the Prediction service and adjusted their strategy to message only users with high payment potential. Such users were offered limited-time-only promotions that were tailored to their subscription periods. This adjustment increased the payment conversion rate to over 20%, while also increasing overall revenue and retention rate despite the message audience being smaller.
So how do we leverage the Prediction service to mine users with high payment potential, and App Messaging to send them relevant messages?
i. Identifying high-value audience.
First, target users who are more likely to make purchases based on their attributes and behavior.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
ii. Configuring a modal in-app message.
Next, create a modal message in App Messaging, and set the title, image, buttons, and so on for your upcoming activity.
iii. Selecting a trigger event.
Choose when to display the message. This could be upon app launch or whenever a user launches the password protection module.
iv. Targeting users
Use the Prediction condition and then select the target audience.
Boost your own app's payment conversion with Prediction + App Messaging today.
To learn more, please visit:
HUAWEI Developers official website
Development Guide
Reddit to join developer discussions
GitHub or Gitee to download the demo and sample code
Stack Overflow to solve integration problems
Follow our official account for the latest HMS Core-related news and updates.
Original Source

HMS Core Analytics Kit - A Quick Glimpse of Real-Time Overview

Core Value
Real-time overview provides real-time data feedback and analysis, which are significant to improve the efficiency of product operations. For key marketing scenarios related to user attraction, such as online operations activities, new version releases, and abnormal traffic warnings, its low-latency data feedback can benefit your agile business decision-making.
Application Scenarios
Scenario 1: Real-Time Evaluation of Activity Traffic
In most cases, after a new user acquisition activity is rolled out online, traffic is monitored hourly or daily. This makes it difficult for operations personnel to accurately locate the root cause of an exception and make timely adjustments, which may hinder the effectiveness of the activity.
Luckily, real-time overview can analyze traffic by the minute and present real-time fluctuations of new users in an app accurately, indicating when the best activity effect is achieved and how to optimize subsequent activities.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
* Test environment data is for reference only.
Scenario 2: Optimization of New Versions
Operations personnel require real-time data to measure the performance and acceptance of new versions, in the face of fast product iterations driven by ever-changing user requirements.
For example, after a game update is released, how the players respond to the new content directly impacts the game's revenue.
To understand how users respond to the update and mitigate its problems, real-time overview can be used as a reference. By referring to real-time overview, you can easily spot abnormal fluctuations, and then quickly optimize your app and take corresponding operations methods.
* Test environment data is for reference only.
Scenario 3: Real-Time View of User Characteristics
Real-time overview helps you understand whether the in-app journey of users matches the product design, whether you have attracted the target users who use specific device models and come from specific places, as well as their in-app behaviors.
The User analysis report clearly displays the real-time distribution of users by each attribute, like channels and countries/regions, in the form of cards.
* Test environment data is for reference only.
With the Event analysis report, you can learn about users' frequent in-app behaviors, so that you can identify the best time to send push notifications and in-app messages.
* Test environment data is for reference only.
How to Use Real-Time Overview
Sign in to AppGallery Connect, click My projects, find your project, and go to HUAWEI Analytics > Overview > Real-time overview.
Visit our official website to learn more.

Try Updated Functions of Analytics Kit 6.6.0

For a business to grow, it must be capable of fine-grained and multi-dimensional user analysis, facing the popularity of precise operations. HMS Core Analytics Kit, which has been dedicated to exploring industry pain points and meeting service requirements, can do that. Recently, it released the 6.6.0 version, further expanding its scope of data analysis.
Here's what's new:
Updated Audience analysis, for even deeper user profile insight.
Added the function of saving churned users as an audience to Retention analysis, contributing to the multi-dimensional analysis on abnormal user churn, and boosting timely user retention with the help of targeted strategies.
Added the Page access in each time segment report to Page analysis, making users' usage preferences even clearer.
Added the function of sending back day 1, day 3, and day 7 retention data to HUAWEI Ads along with conversion events, to help you evaluate ad placement.
1. Updated Audience analysis to Audience insight, and added the User profiling report, for deep knowledge of users
In the new version, the Audience analysis menu is changed to Audience insight, which is broken down into the User grouping and User profiling submenus. User grouping contains the audience list and the audience creation function, while User profiling displays audience details. What's more, User profiling has added the Audience profiling module, which presents basic information about the selected audience through indicators like consumption in last 7 days, so that you can make a practical operations plan.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
* This data is from a test environment and is for reference only.​
2. Saving churned users as an audience in a click to enhance winback efficiency
Winning back users is vital to any business and this can be even more achievable thanks to clear churn analysis, which helps boost winback efficiency with less effort. In Analytics Kit 6.6.0, we have updated the retention analysis model and added the saving churned users function. This allows you to analyze the behavior features of churned users, and what's more, by combining this function with the audience insight function, you can customize differentiated and targeted operations strategies to win back users effectively.
* This data is from a test environment and is for reference only.​
3. Displaying users' preferences for page access time segments, to pinpoint the best opportunity for operations
An abundance of different app types and page functions inevitably leads to varying user preferences for access time segments, making selecting the proper time segments to push content complicated. Fortunately, with Page analysis, you can view the access time segment distribution of different pages. By comparing the number of accesses and users in different time segments, you can fully understand users' product usage preferences and seize proper operations opportunities.
* This data is from a test environment and is for reference only.​4. Evaluating ad placement effects through detailed user loyalty indicators
Analytics Kit can send back conversion events, which provides data support for ad effect evaluation and placement strategy adjustment. In the new version, this function has been updated to send back day 1, day 3, and day 7 retention data along with conversion events, helping you better evaluate user loyalty. By using this retention data, you can further evaluate whether the user groups you advertise to are your target users and whether they are loyal, and adjust ad placement to improve the ROI.
Moreover, Analysis Kit 6.6.0 has also optimized functions like Event analysis and Project overview. To learn more about the updates, refer to the version change history. For more details, click here to visit our official website.

[CLOSED] Try Updated Functions of Analytics Kit 6.6.0

For a business to grow, it must be capable of fine-grained and multi-dimensional user analysis, facing the popularity of precise operations. HMS Core Analytics Kit, which has been dedicated to exploring industry pain points and meeting service requirements, can do that. Recently, it released the 6.6.0 version, further expanding its scope of data analysis.
Here's what's new:
Updated Audience analysis, for even deeper user profile insight.
Added the function of saving churned users as an audience to Retention analysis, contributing to the multi-dimensional analysis on abnormal user churn, and boosting timely user retention with the help of targeted strategies.
Added the Page access in each time segment report to Page analysis, making users' usage preferences even clearer.
Added the function of sending back day 1, day 3, and day 7 retention data to HUAWEI Ads along with conversion events, to help you evaluate ad placement.
1. Updated Audience analysis to Audience insight, and added the User profiling report, for deep knowledge of users
In the new version, the Audience analysis menu is changed to Audience insight, which is broken down into the User grouping and User profiling submenus. User grouping contains the audience list and the audience creation function, while User profiling displays audience details. What's more, User profiling has added the Audience profiling module, which presents basic information about the selected audience through indicators like consumption in last 7 days, so that you can make a practical operations plan.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
* This data is from a test environment and is for reference only.
2. Saving churned users as an audience in a click to enhance winback efficiency
Winning back users is vital to any business and this can be even more achievable thanks to clear churn analysis, which helps boost winback efficiency with less effort. In Analytics Kit 6.6.0, we have updated the retention analysis model and added the saving churned users function. This allows you to analyze the behavior features of churned users, and what's more, by combining this function with the audience insight function, you can customize differentiated and targeted operations strategies to win back users effectively.
* This data is from a test environment and is for reference only.
3. Displaying users' preferences for page access time segments, to pinpoint the best opportunity for operations
An abundance of different app types and page functions inevitably leads to varying user preferences for access time segments, making selecting the proper time segments to push content complicated. Fortunately, with Page analysis, you can view the access time segment distribution of different pages. By comparing the number of accesses and users in different time segments, you can fully understand users' product usage preferences and seize proper operations opportunities.
* This data is from a test environment and is for reference only.
4. Evaluating ad placement effects through detailed user loyalty indicators
Analytics Kit can send back conversion events, which provides data support for ad effect evaluation and placement strategy adjustment. In the new version, this function has been updated to send back day 1, day 3, and day 7 retention data along with conversion events, helping you better evaluate user loyalty. By using this retention data, you can further evaluate whether the user groups you advertise to are your target users and whether they are loyal, and adjust ad placement to improve the ROI.
Moreover, Analysis Kit 6.6.0 has also optimized functions like Event analysis and Project overview. To learn more about the updates, refer to the version change history.
For more details, click here to visit our official website.
MOD ACTION:
Thread closed as duplicate of https://forum.xda-developers.com/t/try-updated-functions-of-analytics-kit-6-6-0.4473115/

Categories

Resources