"Even the most powerful companies will get into trouble if they don't act on the future."
--Peter Drucker
With the rapid development of the mobile Internet, the homogeneity of products has also become increasingly fierce competition. Paying attention to the user's every page conversation path and guiding them in a virtuous cycle in the life cycle can dig more value, and this is the most concerned issue of every developer.
Based on historical user behavior data, Huawei's prediction service focuses on the possible actions of users in the future, allowing you to predict potential user behaviors in advance, and make scientific decisions one step faster.
1. What is a custom forecast
In the previous article, we introduced forecasting tasks such as the loss of services and payment. What is special about custom forecasting? To put it simply, the focused scenes are different and more diverse.
User churn and payment are important considerations related to whether the product can be realized, and it is also a core issue that everyone pays attention to in the process of product operation. However, in the process of user visits, we will also pay attention to some key event conversion nodes, such as the probability of a user clicking on an in-app advertisement, the probability of a user completing a game level, and other in-app conversion events. Custom predictions can be used to evaluate the user’s future Probability of these behaviors.
The diversified forecasting tasks come from the diversified data reported by your application. Similar to the preset prediction task, the custom prediction also trains the model based on the user data reported in the last two weeks to predict the probability of a certain behavior of the active users of the application in the last week in the next week.
2. How to create custom forecast tasks according to business needs
Due to the different application industries, the focus of product operations is also very different. For example, e-commerce apps may pay attention to the probability of users adding products to the shopping cart, the probability of users clicking on the homepage to promote advertisements, etc.; while game apps may focus on the probability of users passing a certain level and the probability of consuming game coins and props. Wait.
Taking the above-mentioned APP in the game industry as an example, we will introduce in detail how to use custom predictions to evaluate the probability of users passing a certain game level.
Players pass the game level, which is inseparable from the "level completion" event. First, we create a COMPLETELEVEL event in the event management of Huawei Analysis Service (this event is preset in the analysis service), and then mark it as a conversion event, so that we can select it as the target when we create a custom prediction task Forecast events.
After selecting the event, click "OK", and the custom forecast task is created successfully. At present, the automatic collection events, preset events and other events that you customize provided by the analysis service support the creation of custom prediction tasks. In the process of use, after marking the target event as a conversion event according to the above method, the event can be marked As the target event for custom predictions.
3. Practical application of custom forecast results
Above we introduced the actual creation scenarios of custom forecasting tasks, then the next step is how to use the forecast results to assist your business growth.
When the prediction task is completed, through the prediction details, we can see that the users who have completed the level with a low probability predicted by the system have very obvious common features in their location and application version. So, when the product is announced, can the region and application version be differentiated? Is it possible to optimize the product experience of this group of people? In view of the common characteristics of the crowd, we can formulate a set of operating plans to tap the opportunities for users to pay for conversion.
*The screenshots are from the customized forecast details page of Huawei Forecast Service, and the data is virtual
The probability of users completing levels is low. Trying to recommend solutions to these users during the game, such as game equipment optimization, virtual currency purchases, etc., can effectively increase the probability of users paying in-game. Through the guidance of the page path, promote the target users to complete the advanced conversion from "Xia Mi users"-"Dolphin users"-"Whale users".
The implementation of the above operation plan is very simple, no additional code development work is required, and it can be easily implemented only by combining with Huawei's remote configuration service.
We need to set the corresponding in-game props recommendation pop-up window on the remote configuration cloud interface to be displayed only to users with a low probability of passing the level. In order to further promote the conversion of users' payment, we can also add configuration virtual currency limited-time discount packages, etc. Method to attract users. With the daily update of the prediction task, the remotely configured audience objects will also be updated simultaneously, so that the target user can be effectively reached without affecting the game experience of other players.
The above is a brief introduction to the custom forecasting task of Huawei Forecasting Service. Come and create your custom forecasting task! To learn more and get started, click here to view the documentation.
For more details, please refer to:
Huawei Developer Alliance official website:https://developer.huawei.com/consumer/cn/hms?ha_source=hms1
Get the development guide document:https://developer.huawei.com/consumer/cn/doc/development?ha_source=hms1 To
participate in the developer discussion, please go to the Reddit community:https://www.reddit.com/r/HuaweiDevelopers/ To
download the demo and sample code, please go to Github:https://github.com/HMS-Core to
solve integration problems, please go to Stack Overflow:https://stackoverflow.com/questions/tagged/huawei-mobile-services?tab=Newest
Original link:https://developer.huawei.com/consumer/cn/forum/topic/0204462615445810035?fid=18
Author: Pepper