User Conversion Graph Targeting via Similarity Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the context of cloud computing and SaaS, it is challenging to effectively target trial users for conversion to paid users, especially when the number of trial users is large or the marketing methods are costly, as existing methods lack a comprehensive approach to determine user similarity and conversion likelihood.
Innovation Solution
A method involving the calculation of conversion likelihood scores and similarity scores for users, constructing a graph with weighted edges representing similarity, and determining marketing potential scores to identify and target users with the highest potential for conversion, utilizing both dynamic and static data, including marketing response, service consumption, and sales data, with a blended score approach that adjusts for user similarity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional marketing methods are used to target trial users, then marketing coverage is broad, but marketing cost increases and conversion accuracy decreases
Solution Approach 1:
The patent segments the trial user base into distinct groups based on their similarity to paid users. By calculating similarity scores and constructing user graphs, the system divides users into target groups (high similarity to paid users) and non-target groups. This segmentation allows marketing resources to be concentrated on the most promising segments, improving conversion accuracy while reducing overall marketing costs.
Solution Approach 2:
The patent transforms the marketing approach by changing key parameters: instead of using broad demographic criteria, it introduces similarity scores based on usage patterns, feature adoption, and engagement metrics. The marketing potential score combines multiple parameters (similarity score, conversion likelihood, user engagement) to create a new targeting criterion that improves both accuracy and cost-efficiency.
2Measurement precision
If comprehensive user data analysis is performed to improve conversion prediction, then conversion accuracy improves, but computational complexity and processing time increase
Solution Approach 1:
The patent performs preliminary actions by pre-calculating similarity scores and constructing user graphs before the actual marketing campaign. User profiles are enriched with pre-computed features (similarity to paid users, conversion likelihood scores) in advance. This preliminary processing organizes the data structure so that during the marketing campaign, the system can quickly query and target users without performing complex real-time calculations, thus reducing computational complexity during execution.
Solution Approach 2:
The patent creates simplified representations (copies) of complex user behavior patterns through similarity scores. Instead of analyzing entire user activity histories during marketing decisions, the system uses pre-computed similarity scores that capture essential patterns. This copying approach preserves the predictive power of comprehensive analysis while reducing the computational burden during actual marketing operations.
Data Source
AI summary
A method for targeting marketing for user conversion includes receiving a list of users. Data pertaining to the users is received. A conversion likelihood score representing an estimation of how likely the user would be to converted from a trial user to a paid user is determined for each user. A similarity score representing how similar the users of the pair are to one another is determined for each possible pair of users. A graph in which each node thereof represents each user and edges between the nodes have edge weights representing the determined similarity scores is constructed. Each node is associated with a value representing its conversion likelihood score. A marketing potential score is calculated for each user using both the node-associated-values and the edge weights of the graph. A set of target users having highest marketing potential scores is constructed.