Notification Lift Metric for Incremental Engagement

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current notification systems often rely solely on predicted click-through rates to decide whether to send unsolicited notifications, failing to consider the user's likelihood of engaging with digital content without notifications, leading to user annoyance and resource wastage.

Innovation Solution

Implementing a system that calculates an incremental benefit by comparing the notification likelihood with a baseline likelihood of user engagement, using machine learning models to determine whether to send unsolicited notifications based on the difference between these probabilities, and considering disengagement probabilities to conserve resources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If notifications are sent based solely on predicted click-through rate, then the likelihood of user engagement with the notification is improved, but user annoyance and distraction increase

Engineering Contradiction:

Improvenotification engagement rateVSAvoiduser annoyance

Core Design Contradiction:

ProductivityVSObject-affected harmful factors

Solution Approach 1:

The system changes the decision parameter from单一的predicted click-through rate to a comparative metric (lift) that measures the incremental benefit of notifications. By calculating the difference between notification likelihood and baseline likelihood, the system identifies notifications that provide genuine incremental value rather than relying on absolute engagement probabilities alone.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system incorporates feedback loops where user interactions with notifications and digital content are continuously monitored. This feedback is used to update machine learning models that predict both notification likelihood and baseline likelihood, enabling the system to learn from past behavior and improve its notification sending decisions over time.

Inventive Principle:

Principle #23Feedback

2Productivity

If more unsolicited notifications are sent to increase engagement opportunities, then the potential benefit from user engagement is improved, but resource wastage increases

Engineering Contradiction:

Improveengagement opportunityVSAvoidresource wastage

Core Design Contradiction:

ProductivityVSLoss of energy

Solution Approach 1:

The system applies partial action by sending notifications only when the incremental benefit (lift) exceeds a threshold, rather than sending all potentially engaging notifications. This selective approach avoids the excessive action of sending notifications that would not provide meaningful incremental value, thereby reducing resource wastage while maintaining engagement opportunities.

Inventive Principle:

Principle #16Partial or excessive action

3Device complexity

If the notification system relies only on predicted click-through rate, then the system complexity is reduced, but the measurement precision of notification benefit is insufficient

Engineering Contradiction:

Improvesystem complexityVSAvoidnotification benefit measurement

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The system segments the notification evaluation process into two distinct probability calculations: notification likelihood (probability of engagement with notification) and baseline likelihood (probability of engagement without notification). This segmentation allows for precise measurement of incremental benefit by comparing the two separate metrics, rather than relying on a single aggregate measure.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11574339B1Modeling lift of metrics for triggering push notifications

Publication Date: 2023.02.07 GOOGLE LLC

AI summary

Processor(s) of a client device can: analyze one or more features of an electronic resource that is under consideration for solicitation to a user; determine a notification likelihood that the user will access the electronic resource in response to an unsolicited notification of the electronic resource being output to the user; determine a baseline likelihood that the user will access the electronic resource without being solicited; compare the notification likelihood with the baseline likelihood; and cause, based on the comparing, the unsolicited notification to be output to the user. In some implementations, determining the notification likelihood and/or the baseline likelihood is based on applying data associated with the electronic resource as input across a machine learning model to generate output indicative of the notification likelihood and/or the baseline likelihood. In other implementations, determining the notification likelihood and/or the baseline likelihood is based on past behavior or preference(s) of the user.