Lock Screen Content Prediction for Offline Interaction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advertisers face challenges in accurately measuring user interactions with content displayed on lock screens, especially in offline modes, leading to potential over-spending due to inaccurate response capture and double billing for online and offline impressions.
Innovation Solution
A method and system that predicts content consumption by identifying and expanding user groups based on initial and subsequent responses, both online and offline, using machine learning models to determine when offline users will reconnect, allowing for targeted and accurate content delivery and billing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If content is pushed to user devices during online state for display on lock screen, then content can be displayed to users, but inaccurate measurement of user interaction during offline state leads to extra spending
Solution Approach 1:
The system performs preliminary actions by pushing content to user devices during online state before users go offline. The server records which content was pushed and to which users, then uses this information to predict and measure interactions when users come back online, avoiding the need to re-push content and preventing double billing.
Solution Approach 2:
The system implements feedback mechanisms by tracking user interactions with content during online state and using this feedback to predict offline interactions. The server receives interaction data when users are online, processes this feedback to identify patterns, and uses these patterns to accurately measure and bill for offline interactions when users reconnect.
2Productivity
If content is displayed on lock screen to increase user response probability, then content visibility increases, but full content requires internet connection which limits offline display
Solution Approach 1:
The server pushes content to user devices during online state as a preliminary action, storing the content locally on the device. This allows the content to be displayed on the lock screen even when the user is offline, as the content is already cached in the device's memory from the previous online session.
Solution Approach 2:
The user device acts as an intermediary between the server and the user. The device receives content from the server during online state and stores it locally, then displays the content on the lock screen when the user is offline. This intermediary storage mechanism bridges the gap between online content delivery and offline display requirements.
3Reliability
If content is pushed again when user becomes online after offline interaction, then content is re-delivered, but advertiser may have to spend for two impressions
Solution Approach 1:
The server uses feedback from tracking user interactions during online state to predict and identify offline interactions. When users come back online, the server checks whether the user interacted with content while offline based on the recorded interaction data, and only bills for the impression if it wasn't already counted during the offline period, preventing double billing.
Solution Approach 2:
The server performs preliminary tracking of user interactions during online state and stores this information before users go offline. When users return online, the server retrieves the preliminary interaction data to determine whether content was already delivered and viewed during the offline period, allowing the system to avoid redundant content pushing and billing.
Data Source
AI summary
Provided is a method and system for predicting content consumption. The method comprises pushing one or more content into a lock screen of a user device (105) of an initial group of online users from the server (101) and thereby receiving a plurality of initial responses by the server (101). Based on the plurality of initial responses, the method comprises determining a first group of users including a first set of online users and a first set of offline users and thereby receiving a plurality of first responses. Based on the plurality of first responses, the method comprises expanding the first set of offline users by a second set of offline users and thereby receiving a plurality of second responses. The method comprises predicting content consumption based on the plurality of initial responses, the first responses, and the second responses.