Defining the Session Path Analysis ModelThe session path analysis model provides an intuitive way of displaying how an app is used. With this model, developers can clearly understand the changes in user behavior, frequently visited paths, steps where users churned, and steps with an unexpected churn rate. Such information helps developers enhance the app, user experience, and user conversion.
When to Implement Session Path Analysis
1. Guiding Updates by Showing How Users Actually Use the AppTo maximize the benefits of app optimization and updates, it's important to identify the differences between an app's purpose and its actual usage.
{
"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"
}
* Example of a session path analysis report
Let's take Now: Meditation, a leading meditation app in China, as an example of Analytics Kit being used effectively. Now: Meditation provides five types of online courses (pressure relief, emotional management, personal development, sleeping, and focus enhancement) and its revenue is mainly generated from member payments. Initially, the product team thought that pressure relief and emotional management courses would be the most popular, but in reality, the session path analysis shows that sleeping-related courses were the ones drawing most user payments.
In light of this discovery, the product team decided to adjust the display level of the sleeping-related courses so that users would spend less time making a choice. Now: Meditation has reaped the rewards of this change, which has boosted the next-day retention rate of new users by 15%, the number of daily active users by 17%, and the payment conversion rate by 20%.
2. Locating App Flaws by Comparing the Preferred Session Paths of Different UsersThe filter function enables us to compare the path preferences of different audiences. For example, we can check whether new users share the same path during different time segments, and learn which paths are preferred by loyal users. You can segment and further analyze different audiences, and determine the rule of each session path. In this way, session path analysis lets you best optimize your app.
Let's say that we have a lifestyle and social app that allows users to share excerpts of their life using short videos, images, and text, as well as follow people and like others' posts. The operations team found that recently the overall user retention rate dropped significantly, so they were eager to learn about the behavior characteristics of loyal users. Such information could be used to help them design operations campaigns to lead new users and those about to churn to have such characteristics. In this way, the team could improve overall user loyalty for the app.
* Example of a session path analysis report
The app's operations team utilized the session path analysis function. They selected Active users in the filter and set App launch as the start event, with the goal of analyzing the behavioral path of such users. They found that over 70% of active users launched the app three times per day, and were more likely to browse content and follow other users. With this information, they were able to implement two measures to improve user retention by improving the rates of push notification tapping and of users following each other:
First, by enhancing the push notification sending mechanism, drafting better push notification content according to the A/B testing result, and sending audience-specific push notifications. And second, by displaying a message to prompt the user to follow another user when the user browses the latter's home page for more than 3 seconds. After just one month since these two measures were implemented, the retention rate skyrocketed.
3. Finding the Path with the Highest Payment Conversion RateAn app tends to have different banners, icons, and content, designed to guide VIP members toward making payments. But which path best gets users to make payments? What is the difference in churn rate for each step per path? Which path has the best conversion effect? And which paths are worthy of more in-app traffic?
Session path analysis has the answers. First, we select only the events related to app launch and payment completion, and then set payment completion as the end event. Session path analysis will then automatically display the traffic resulting from each path to the payment completion event, helping us compare their payment conversion rates.
* Example of a session path analysis report
4. Specifying the Start Event, and Exploring a Diverse Range of PathsThe product management or operations team often pre-design a path for a function or campaign, which they expect users to follow. However, not all users will follow this path, and a certain number of users will churn during each step. This leads to some important questions, such as, what do users do after they churned? What attracted them in the first place?
Let's say for an e-commerce app, many users churned in the steps between order creation and payment completion. The operations team wanted to analyze why users abandoned payment. Was it because users browsed other products? Did they leave to compare prices? Did they set another order because they had provided a wrong address or ordered too little/much? To answer these questions, the team can set order creation as the start event, and checked what users did after this event.
* Example of a session path analysis report
5. Checking the Path with Unexpected Churn to Determine the ReasonWhen using session path analysis, if you find unexpected churn during certain steps, you can save these steps as a funnel with just one click. Then, you can use the funnel analysis function to identify the reasons behind user churn.
For example, if the session path analysis report has revealed that most new users churned from adding a product to favorites to placing an order and making payment, the events of this process can be saved as a funnel. You can then use the funnel analysis function to determine the reasons behind user churn, and analyze the conversion effects of the funnel.
* Example of a session path analysis report
Funnel analysis provides the filter and comparison analysis functions, which allow you to check data according to such conditions as app version, active users, new users, and download channel. By analyzing such data, you'll be able to locate the root causes leading to user churn, and take measures to optimize your app accordingly.
* Example of a funnel analysis report
These are about the major scenarios where session path analysis can play its role. To learn more, click here to get the free trial for the demo, or visit our official website to access the development documents for Android, iOS, Web, and Quick App.
Here car Bluetooth is visible and connected in mobile but music is not audible.. someone suggest me what is factory reset code fr this car music system
Thanks for sharing!!
Related
As competition among mobile apps gets more intense, converting users is more challenging than ever. You need to understand how people are responding to your apps, so you can identify any issues as soon as they arise, and deal with them promptly.
{
"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"
}
Funnel analysis enables you to do that precisely. You can use this service to monitor conversion rates at each stage of the customer journey, identify the stages where your conversion rate is low, and optimize your approach accordingly.
How the Funnel Model Drives Your Business
1． Identify weak points in the customer journey
The first step to boosting conversion rate is to design a funnel which measures behavior throughout the user journey. This starts from when the user registers an ID, then encompasses every subsequent touch point: when they sign in to your app, level up, spend virtual currency, and finally, make a payment. By looking over the data for each of these stages, you can see where your conversion rate is low, and think of a strategy to address this. In this way, you can improve your app's user experience and retain users.
HUAWEI Analytics provides three pre-defined funnel models: the registration conversion funnel, e-commerce conversion funnel, and search conversion funnel. You can also design your own funnel.
2. Find root causes with drill-down analysis
Different users have different requirements, which means the only way to understand why a certain stage has a low conversion rate is by segmenting users. Let's say your conversion rate drops at the member payment stage. To understand why this is, you need to segment users by region, age, member level, gender, and the app version they are using, and then compare data for these groups.
With HUAWEI Analytics' funnel models, you can filter by platform, app, user attribute, and audience, and perform drill-down comparison analysis to find the root causes of low conversion rate.
3. Create audiences with one click
To increase your app's conversion rate, you need to do more than simply focus on a percentage. HUAWEI Analytics' funnel model allows you to save users who have converted or churned at each stage of the journey as an audience. You can then analyze the behavior and attributes of these audiences using a behavior analysis model.
For example, by saving the users who churned at a certain stage as an audience, you can analyze their distribution in terms of events, the page of your app they prefer, app startup time, session paths, and activeness. This helps you gain deeper insight into your churned users, so you can find ways to prevent this churn.
How to Use a Funnel Model
There are four steps to use a funnel model:
1. Design or choose a funnel model and determine the conversion path.
2. Monitor the conversion rate at each stage and identify the reasons behind a low conversion rate.
3. Develop improvement measures.
4. Implement these measures and continue to check whether the conversion rate in the funnel increases.
When it comes to analyzing root causes in the second step, there are a few things to consider:
l Segment users and analyze their behavior: In the funnel analysis report, you can save churned users and converted users as audiences, and then analyze their characteristics with event analysis, page analysis, session path analysis, launch analysis, and retention analysis.
l Perform drill-down analysis: Compare conversion rates of different audiences using filters and comparison analysis, and find areas where your app can be improved.
l Optimize user experience: By looking at conversion rates, you can see where users are dropping off and find ways to improve the user experience.
l Take a broad perspective: Consider your app's overall design, rather than paying too much attention to small details, like the style of a button, or the color of a page. You can use other services like Push Kit, A/B Testing, Remote Configuration, and App Messaging together with HUAWEI Analytics to analyze your app more comprehensively.
Once you have found possible reasons for user churn, you can think of ways to address these. This is the third step of using a funnel model. In this step, you need to sort the factors that may be affecting the conversion rate by priority, and optimize key factors.
During the fourth step, you will check whether the conversion rate increases once you have implemented the countermeasures in step three.
By going through these four steps, you can find the causes of user churn and improve your conversion rate with data-driven solutions.
Typical Case
A travel app used HUAWEI Analytics to design a funnel which measured user behavior throughout the user journey, including when the user opened the app, browsed the travel guides on the home page, added a travel guide to favorites, tapped to share a travel guide, and finally successfully shared the guide. By monitoring the conversion rate at each stage, the developers found that the churn rate between the time where the user tapped to share a travel guide, and the time when they actually shared the guide, was very high. To find the problem, they saved users churned between the two stages as audiences, and analyzed their behavior using the path analysis function. When they did this, they discovered that among these users, most of those who have added a travel guide to favorites had taken a screenshot of the travel guide page.
Based on this discovery, the operations team made an assumption: when users share content to social media, they take screenshots, combine multiple screenshots into one image, beautify the image, then post. If this was true, the team could help their users by adding the option to generate a long image of the current page and share it to social media.
To verify this, the team conducted an online survey which found that 72% of users wanted to be able to save travel guides or notes as long images, while 51% would prefer to save content in the form of a post card.
Having established this, the team added two functions to the content sharing page: generating a long image, and generating a post card. These functions were immediately popular among users, and helped double the app's content sharing rate.
Easy Integration
Today, more than 6000 apps use HUAWEI Analytics around the globe. When you integrate this service, you can use data-driven insights to increase your apps' conversion rates. What's more, integrating HUAWEI Analytics only takes about five minutes!
☞ HUAWEI Analytics Integration Guide:
Android iOS Web
☞ Sample Code:
Android iOS Web
If you have any questions during integration, you can contact us by submitting a ticket online.
We hope you enjoy using HUAWEI Analytics.
It's quite worth trying
When your app user base reaches a certain threshold, you may find it harder to acquire new users continually. In this case, rather than struggling to attract new users, it's better to try to activate your inactive users. Let's take a look at how HUAWEI Analytics Kit can help you activate and win back inactive/lost users.
1. Defining Inactive and Lost UsersTo win back inactive and lost users, the first thing you'll need to do is reasonably define them. To do this, you'll need to take your business characteristics into consideration, and quantify key user behaviors.
For example:
If you have a game app, you can define these users by the number of consecutive days that they do not sign in to your app.
If you have an e-commerce app, you can define these users by the number of consecutive days that they do not place an order in your app.
If you have a video app, you can define these users by the number consecutive days that they do not watch a video in your app.
Analytical models in Analytics Kit are the tools you need to define inactive and lost users accurately and quickly.
Identifying turning points that lead to user churn
Let's use the demo app for Analytics Kit as an example. Its revisit user report indicates that the first turning point appears in day 31–90. This means that there is a small possibility that users who have been inactive for 30 days become active again. The second turning point appears in day 91–180. According to this information, it would be wise to define users whose consecutive days with no app use are greater than or equal to 30 days but equal to or smaller than 90 days as the inactive users, and define users whose consecutive days with no app use are greater than 90 days as the lost users.
Analytics Kit also offers the user lifecycle model, which allows you to customize the statistical scale for each phase of your app's users.
{
"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 figure shows the user lifecycle analysis report for Analytics Kit with virtual data.
2. Creating Lost User ProfilesA comprehensive understanding of your users makes it much easier for you to build the best possible app. Implementing an effective winback strategy requires that you create detailed lost user profiles that are based on a complete understanding of past behavior characteristics of inactive or lost users.
To create these profiles, you can refer to:
The number and trends for inactive/lost users
The user lifecycle report offers an intuitive glimpse at the number and trend of users in the inactive and lost phases.
* This figure shows the user lifecycle analysis report in Analytics Kit with virtual data.
Phases when users become inactive/lost
The user lifecycle report displays the ratios of inactive users converted from beginner users, growing users, and mature users to all inactive users.
From the figure below, we can roughly divide inactive users into two categories: those converted from the users in the beginner and growing phases, and those converted from the users in the mature phase. The ratio of the first category is much higher than that of the second — this indicates that the more dependent a user is on your app, the less likely it is that he/she becomes inactive or lost.
* The figure shows the user lifecycle analysis report in Analytics Kit with virtual data.
Users in the first category became inactive before they have fully experienced your app. This might have been due to a cumbersome user experience, undesirable product design, or failure to deliver the "Aha!" moment that hooks users to your product. Users in the second category became inactive or lost after they fully experienced your app. This is likely because the product failed to bring them the experience better than their expectation.
Whether most inactive/lost users have similarities
On the user lifecycle report page, you can save inactive/lost users as an audience with just one click on the number. You can then go to the audience analysis report to check whether most inactive/lost users have similarities in aspects such as the model, location, event, system version, or download channel.
* This figure shows the audience analysis report in Analytics Kit with virtual data.
* This figure shows the audience analysis report in Analytics Kit with virtual data.
Behavioral characteristics of inactive/lost users
Behavior analysis provides a filter function, in which you can select inactive/lost users. The session path analysis report will then tell you the behavior path of inactive/lost users before they became inactive/lost, and the session step where this occurred.
* This figure shows the session path analysis report in Analytics Kit with virtual data.
You can save nodes related to user churn as the funnel, and then pinpoint causes of churn by checking the funnel analysis report.
* This figure shows the session path analysis report in Analytics Kit with virtual data.
Analysis of the value inactive/lost users contributed in the past
It's a good idea to analyze the value inactive/lost users have contributed to your app in the past. By doing so, you can separate them into different groups and formulate winback strategies that have a greater chance of success.
The event analysis function enables you to identify previously paying users among inactive/lost users, and learn more information about them like total top-up amounts, gross merchandise volume (GMV), and top-up frequency. Thanks to this information, you can divide inactive/lost users into different groups according to the priority and difficulty of winning them back.
3. Specifying Winback Strategies
Determining which group should be won back first
In the previous step, we created inactive/lost user profiles, and separated them into different groups according to the difficulty of winning them back, through such indicators as the previous behavior and value of lost users, whether they have uninstalled the app, and whether they can be reached now. In principle, the users to win back first are those who have become inactive, but have not yet uninstalled the app.
Determining the focus of your winback strategy
Different winback strategies have different focuses, including:
Benefits: For example, you can send coupons to inactive/lost users, or remind them of existing coupons or virtual currencies that will soon expire.
User interests: Let's use a video app as an example. Some of the inactive users are interested in animation. To win them back, you can send them notifications about upcoming new animation series, or about activities related to animations they are interested in.
Emotions: Let's say you have a life simulation game app which features virtual pets. To entice users to open your app to take care of their pets, you can send them notifications about their pets' health or emotional status.
Selecting a winback channel
Ads: In addition to attracting new users, ads can also play a role in activating or winning back users. When you use ads for this purpose, you're likely to get better than expected results.
Push notifications: If you choose this channel, make sure that the time and frequency for pushing notifications are reasonable. Also remember to check the percentages of users who uninstall your app and disable the push notification after they receive the notification.
SMS messages: You should use this channel to reach only target users, since costs associated with SMS messages are a little higher. To achieve better winback results, you can tailor the content of messages sent out to different user groups.
Emails: This is one of the most common channels for reaching users. You'll need to consider how to best impress your inactive/lost users in a short email, in order to win them back.
Those are the three steps for targeting and wining back lost users with Analytics Kit. Though Analytics Kit makes it easier than ever to win back users, it's still better to create mechanisms that warn you about which users may become inactive, and take proactive measures to retain users who have been won back. This will ensure you to keep improving engagement and loyalty of your users.
About Analytics Kit:
Analytics Kit is a one-stop user behavior analysis platform for products such as mobile apps, web apps, and quick apps. It offers scenario-specific data management, analysis, and usage, helping enterprises achieve effective user acquisition, product optimization, precise operations, and business growth.
For more details, you can go to:
Our official website
Demo of Analytics Kit
Android SDK integration documentation
iOS SDK integration documentation
Web SDK integration documentation
Quick app SDK integration documentation
Sports and health apps are thriving as more and more people place greater value on their health and exercise. This has in turn increased traffic and brought more complex demands. To seize this opportunity and retain long-term users, what is needed is to perform precise operations.
To do so, a practical, industry-specific event tracking system is vital. It is the very start of data analysis and paves the way for data-based operations. In light of this, Analytics Kit offers the event tracking template for the sports and health apps, which provides E2E event tracking management, simplifying app development. Analytics Kit thereby facilitates event tracking and maximizes data value, driving sports and health apps towards digital transformation.
Intelligent Event Tracking1. Selecting a TemplateSelect Health of Sports and Health. The page displayed shows four templates: Behavior analysis, Account Analysis, Consumption Analysis, and Services and Other.
{
"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"
}
* Template for sports and health apps
2. Configuring Event TrackingAnalytics Kit supports event tracking either by coding or through visual event tracking. Tracking by coding can be implemented by copying the sample code, downloading the report, or using the tool provided by Analytics Kit. Tracking by coding is relatively stable, and can collect and report complex data. Visual event tracking comes with lower costs and technical requirements, and allows the visual event to be modified and added after the app release. To use visual event tracking, you need to integrate Dynamic Tag Manager (DTM) first. You can then synchronize the app screen to a web-based UI and click relevant components to add events or event parameters.
* Configuring event tracking
3. Verifying the Tracking ConfigurationYou can use the verification function to quickly identify incorrect and incomplete configurations, as well as other exceptions in events and parameters once event tracking is configured for a specific template. With this function, you can configure event tracking more accurately and mitigate business risks.
* Verifying the tracking configuration
4. Managing Event TrackingThe management page presents the event verifications and registrations, the proportion of verified events to its maximum, as well as the proportion of registered parameters to its maximum. Such information serves as a one-stop management solution, giving you a clear understanding of event tracking progress and the structure of tracking configurations.
* Managing event tracking
CaseReleased in February, 2016, Now: Meditation has become a leading app dedicated to meditation and mental health in China. Its user retention rate and payment conversion have significantly improved since it utilized the event tracking template for sports and health apps provided by Analytics Kit. The template provides analysis reports containing various indicators and comparison analysis by different dimensions, giving an insight into how users used the app, identifying unusual payment conversion rates, and showing conversion rates of different channels. All of these can illustrate what attracts users to the app and what can be done to encourage users to make payments.
1. Understanding App Usage to Guide Product OptimizationAfter event-related data was reported, operations personnel of Now: Meditation checked them in the session path analysis report. The report clearly showed how users behaved in the app, what they did following each step, the steps where users churned, and the steps with an unexpected churn rate.
This powerful function enabled the personnel to find that most active users tended to check content related to improving sleep. Operations personnel concluded that this type of content was most popular among users. Consequently, in the updated version of the app, the product team adjusted the display level for this particular content. They also optimized push notifications by using A/B Testing and sent targeted notifications to audiences. One month after these measures were taken, the retention rate skyrocketed.
* Example of a session path analysis report
2. Analyzing the Reasons Behind UninstallationsAnalyzing why users uninstall an app has become a must. Few data analysis platforms, however, could perfectly meet this demand. Luckily, with the uninstallation analysis function in Analytics Kit, capturing app uninstallation events is no longer a daunting task thanks to the system-level broadcast capability. This function shows the uninstallation status and characteristics of users' pre-uninstallation behavior. With this information at your disposal, you'll be capable of finding the root reason behind uninstallation and better optimizing operations campaigns and the product.
The uninstallation analysis report of Now: Meditation clearly showed that before users uninstalled the app, they tended to engage in three events: tapping push notification, ad display, and performing searches. Operations personnel believed the reasons why users uninstalled the app were due to inappropriate frequency, timing, and incorrect audience for push notifications and ads. Other reasons include lack of informative course content and wrong course recommendations. Based on this assumption, the product team decided to send push notifications and ads less frequently, opting to send different notifications and ads to different audiences by using A/B Testing, and in accordance to user attributes. On top of this, the team also improved the course recommendation mechanism. These small changes have delivered a sense of personalization to users, which in turn has led to a significant drop in the uninstallation rate.
* Example of an uninstallation analysis report
3. Attributing Contribution Rates of Slots to Convert More UsersAn app tends to have different banners, icons, and content, designed to induce VIP members toward making purchases. This leads to some questions: what is the difference in how much each slot and marketing campaign contributes to payment conversion? How to optimize the combination of slots? And how to allocate resources to them more reasonably?
In order to answer these questions, the operations personnel of Now: Meditation used the event attribution analysis and marketing attribution analysis models in Analytics Kit. With these models, they were able to evaluate the user attraction and conversion effects of slots by week and month and check how push notifications contributed to user conversion. Let's take the analysis of the home screen slots as an example. The operations personnel used event attribution analysis to conduct an analysis of the slots. They chose Payment completed as Target conversion event and selected Push notification tapping, Pop-up window tapping, Splash ad tapping, Banner tapping, Searching, Checking exclusive content, and Checking popular courses as To-be-attributed event. They then chose Last event attribution as Attribution model. The following day, the report showed how much each slot contributed to the target conversion event. With this information, the personnel then adjusted how traffic and marketing campaigns were allocated to slots. As a result, they could effectively plan resource allocation.
* Example of attribution analysis
4. Establishing a Churn Warning System to Win back Inactive and Lost UsersWhat's the top concern of apps now? Undoubtedly, it's how to retain users. In the case of Now: Meditation, operations personnel used retention analysis and revisit users analysis to reveal the causes behind user churn. The personnel then used the user lifecycle analysis function to save inactive and lost users as an audience. Once this audience was created, they analyzed the scale of such users, their ratio in all users, their behavior characteristics, the phase from which they turned, and whether they came from a specific channel. With such information, the personnel prioritized which audiences they should attempt to win back. Then, they tried to engage users through ads, push notifications, SMS messages, and e-mails according to users' interest, benefits, and emotions. By the end of this, they established a complete churn warning and user winback system.
* Example of a user lifecycle analysis report
To learn more, click here to get the free trial for the demo, or visit our official website to access the development documents for Android, iOS, Web, and Quick App.
Does it support the gaming application as well?
how we can check uninstallation.
Basavaraj.navi said:
Does it support the gaming application as well?
Click to expand...
Click to collapse
Hi, HMS Core Analytics Kit provides Game Industry Analysis Reports for the gaming applications. You can check this article published early: https://forum.xda-developers.com/t/...ry-analysis-reports-in-analytics-kit.4320363/
You can also access our official site for more details: https://developer.huawei.com/consumer/en/hms/huawei-analyticskit?ha_source=hmsxds
lokeshsuryan said:
how we can check uninstallation.
Click to expand...
Click to collapse
Hi, there's Uninstallation Analysis provided by HMS Core Analytics Kit. You can check this article published early: https://forum.xda-developers.com/t/...lation-analysis-to-reduce-user-churn.4305551/
You can also access our official site for more details: https://developer.huawei.com/consumer/en/hms/huawei-analyticskit?ha_source=hmsxds
As enterprises have placed higher requirements on data monitoring and analysis, marketing campaigns have changed at a breakneck speed. This phenomenon has posed a challenge for operations personnel, who find it difficult to make informed decisions based on data that's being updated on an hourly or T+1 basis.
This is especially true when a new marketing campaign is launched, a new version is released, or when ads are delivered in different periods and through multiple channels. Under these scenarios, product management and operations personnel need the access to minute-by-minute fluctuations in the number of new users and number of users who have updated the app, as well as the real-time data on how users are engaging with the promotional campaign. Armed with this timely data and real-time decision-making capabilities, the personnel are able to ensure that the results from a promotion, update, or launch meet expectations.
Analytics Kit leverages Huawei's formidable data processing and computing capabilities, offering a reconstructed real-time overview function that's based on ClickHouse. It makes operations more seamless than ever, by showing data such as the number of new users and active users in the last 30 minutes and current day, channels that acquire users, app version distribution, and app usage metrics.
Data Provided by Real-Time Overview1. Data Fluctuations from the Last 30 Minutes and 48 HoursThis section provides the access to the numbers of new users, active users, and events over the last 30 minutes, as well as comparisons between the numbers of new users, active users, and events by minute or hour from the current day or day before. You can also filter the data to meet your needs, by specifying an acquisition channel, app version, country/region, or app to perform more thorough analysis.
{
"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"
}
* Example of a real-time overview report
2. Real-Time Information about Events and User AttributesThis part provides user- and event-related data, including their parameters and attributes. When using it in conjunction with data from the previous section, you can get a highly detailed picture of app usage, getting crucial insights into key indicators for your app.
* Example of a real-time overview report
* Example of a real-time overview report
3. User DistributionHere you'll get in-depth insights into how your app is being used, thanks to a broad range of real-time data related to the distribution of users by channel, country/region, device model, and app.
* Example of a real-time overview report
When Should I Use Real-Time Overview1. To Identify Unexpected TrafficOnce you have delivered ads through various channels, you'll be able to view how many users have been acquired through each of these channels via the real-time overview report, and then allocate a larger share of the ad budget to channels that have performed well.
If you find that the number of new users acquired through a channel is significantly higher than average, or that there is a sudden surge from a specific channel, the report can tell you the device models and location distributions of these new users, as well as how active they are after installing the app, helping you determine whether there are any fake users. If there are, you can take necessary actions, such as reducing the ad budget for the channel in question.
You can also check the change in the number of users acquired by each channel during different time segments from the last 48 hours. Doing so allows you to compare the performance of different promotional assets and channels. Those that fail to bring about expected results can simply be replaced.
Take an MMO game as an example, whose operations personnel use the real-time overview function to determine changes in new user growth following the release of different promotional assets. They found that the number of new users increased significantly when an asset is delivered at 10 o'clock. During the lunch break, however, the new user growth rate was far lower than expectations. The team then changed the asset, and was happy to find that the number of new users exploded during the evening, meeting their initial target.
2. For Real-Time Insights on User Engagement with a Promotional CampaignOnce you've launched a marketing campaign, you'll be able to monitor it in real time by tracking such metrics as changes in the number of participants, geographic distribution of participants, and the numbers of participants who have completed or shared the campaign. Such data makes it easy to determine how effective a campaign has been at attracting and converting users, as well as to detect and handle exceptions in a timely manner.
For example, an e-commerce app rewards users for participating in a sales event. It used the real-time overview report to determine that the number of participants from a certain place as well as the app sharing rate of participants from this place were both lower than expected. The team pushed a coupon and sent an SMS message to users who had not yet participated in the campaign, and saw the participation rate skyrocket.
3. To Avoid Poor Reviews During Version UpdatesWhen you release a new app version for crowdtesting or canary deployment, the real-time overview report will show you the percentage of users who have updated their app to the new version, the crash rate of the new version, as well as the distribution of users who have performed the update by location, device model, and acquisition channel. If an exception is identified, you can make changes to your update strategy in a timely manner to ensure that the new version will be better appreciated by users.
Furthermore, if the update includes new features, the report will show you the real-time performance of the new features, in addition to any relevant user feedback, helping you identify, analyze, and solve problems and optimize operations strategies before it's too late.
Timely response to user feedback and adjustments to operations strategies can help boost your edge in a ruthlessly competitive market.
That's it for our introduction and guide to the real-time overview report. To learn more, click here to get the free trial for the demo, or visit our official website to access the development documents for Android, iOS, Web, and Quick App.
Thanks for sharing.
authorization_url, state = flow.authorization_url(
# Enable offline access so that you can refresh an access token without
# re-prompting the user for permission. Recommended for web server apps.
access_type='offline',
# Enable incremental authorization. Recommended as a best practice.
include_granted_scopes='true')
In the information age, the external market environment is constantly changing and enterprises are accelerating their digital marketing transformation. Breaking data silos and fine-grained user operations allow developers to grow their services.
In this post, I will show you how to use the prediction capabilities of HMS Core Analytics Kit in different scenarios in conjunction with diverse user engagement modes, such as message pushing, in-app messaging, and remote configuration, to further service growth.
Scenario 1: scenario-based engagement of predicted user groups for higher operations efficiency
Preparation and prevention are always better than the cure and this is the case for user operations. With the help of AI algorithms, you are able to predict the probability of a user performing a key action, such as churning or making a payment, giving you room to adjust operational policies that specifically target such users.
For example, with the payment prediction model, you can select a group of users who were active in the last seven days and most likely to make a payment over the next week. When these users browse specific pages, such as the membership introduction page and prop display page, you can send in-app messages like a time-limited discount message to these users, which in conjunction with users' original payment willingness and proper timing can effectively promote user payment conversion.
{
"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"
}
* The figure shows the page for creating an in-app message for users with a high payment probability.
Scenario 2: differentiated operations for predicted user groups to drive service growth
When your app enters the maturity stage, retaining users using the traditional one-style-fits-all operational approach is challenging, let alone explore new payment points of users to boost growth. As mentioned above, user behavior prediction can help you learn about users' behavior willingness in advance. This then allows you to perform differentiated operations for predicted user groups to help explore more growth points.
For example, a puzzle and casual game generates revenue from in-app purchases and in-game ads. With a wide range of similar apps hitting the market, how to balance gaming experience and ad revenue growth has become a major pain point for the game's daily operations.
Thanks to the payment prediction model, the game can classify active users from the previous week into user groups with different payment probabilities. Then, game operations personnel can use the remote configuration function to differentiate the game failure page displayed for users with different payment probabilities, for example, displaying the resurrection prop page for users with a high payment probability and displaying the rewarded ad page for users with a low payment probability. This can guarantee optimal gaming experience for potential game whales, as well as increase the in-app ad clicks to boost ad revenue.
* The figure shows the page for adding remote configuration conditions for users with a high payment probability.
Scenario 3: diverse analysis of predicted user groups to explore root causes for user behavior differences
There is usually an inactive period before a user churns, and this is critical for retaining users. You can analyze the common features and preferences of these users, and formulate targeted strategies to retain such users.
For example, with the user churn prediction model, a game app can classify users into user groups with different churn probabilities over the next week. Analysis showed that users with a high churn probability mainly use the new version of the app.
* The figure shows version distribution of users with a high churn probability.
The analysis shows that the churn rate is higher for users using the new version, which could be because users are unfamiliar with the updated gameplay mechanics of the new version. So, what we can do is get the app to send messages introducing some of new gameplay tips and tricks to users with a high churn probability, which will hopefully boost their engagement with the app.
Of course, in-depth user behavior analysis can be performed based on user groups to explore the root cause for high user churn probability. For example, if users with a high churn probability generally use the new version, the app operations team can create a user group containing all users using the new version, and then obtain the intersection between the user group with a high churn probability and the user group containing users using the new version. The intersection is a combined user group comprising users who use the new version and have a high churn probability.
* The figure shows the page for creating a combined user group through HUAWEI Analytics.
The created user group can be used as a filter for analyzing behavior features of users in the user group in conjunction with other analysis reports. For example, the operations team can filter the user group in the page path analysis report to view the user behavior path features. Similarly, the operations team can view the app launch time distribution of the user group in the app launch analysis report, helping operations team gain in-depth insights into in-app behavior of users tending to churn.
And that's how the prediction capability of Analytics Kit can simplify fine-grained user operations. I believe that scenario-based, differentiated, and diverse user engagement modes will help you massively boost your app's operations efficiency.
Want to learn more details? Click here to see the official development guide of Analytics Kit.