Perceived Value Attribution Model for Marketing Touchpoints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users are overwhelmed by multiple online and offline advertisements, making it difficult to attribute the impact of various marketing channels on their decision to place an order, as existing methods fail to accurately allocate credit to touchpoints beyond the last touchpoint, ignoring the cumulative influence of multiple marketing channels over time.
Innovation Solution
A perceived value attribution system using machine learning models tracks user interactions with marketing channels over time, constructs training examples from historical data, and allocates credit to each touchpoint based on its contribution to the order, considering both short-term and long-term effects of advertisements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional attribution methods are used that only credit the last touchpoint, then the implementation is simple, but the measurement precision of marketing channel impact is poor
Solution Approach 1:
The patent segments the attribution problem into multiple time-based components (short-term effects within 7 days, medium-term effects 7-30 days, long-term effects beyond 30 days) and channels (display, search, video, social). This segmentation allows the complex attribution model to be structured in manageable parts while maintaining high measurement precision through comprehensive touchpoint tracking.
Solution Approach 2:
The system performs preliminary actions by pre-processing and organizing touchpoint data into structured formats before attribution analysis. Historical touchpoint data is pre-collected and stored with metadata about channel, timing, and user interaction, enabling efficient processing when attribution calculations are performed without real-time complexity.
2Measurement precision
If comprehensive tracking of all touchpoints is implemented, then the measurement precision improves, but the loss of time and computational resources increases
Solution Approach 1:
The patent implements periodic action by processing attribution in scheduled batches rather than continuously. The system periodically computes attribution for completed orders using accumulated touchpoint data, which reduces real-time processing requirements while maintaining accurate measurement of all touchpoint contributions through comprehensive historical tracking.
Solution Approach 2:
Touchpoint data is pre-collected and stored in an organized manner as users interact with marketing channels throughout their journey. This preliminary data preparation includes capturing channel type, timestamp, and interaction details, so that when attribution is needed, the system can quickly process pre-organized data rather than collecting and structuring it at the moment of calculation.
3Measurement precision
If equal credit is allocated to all touchpoints, then the ease of operation is high, but the measurement precision of individual channel impact deteriorates
Solution Approach 1:
The patent applies local quality by assigning different credit weights to different touchpoints based on their specific characteristics rather than uniform distribution. Each touchpoint receives attribution proportional to its actual influence, determined by factors such as channel type (display, search, video), timing in the customer journey, and engagement quality. This allows precise measurement of individual channel impact while maintaining operational simplicity through automated machine learning calculations.
Solution Approach 2:
The system uses feedback mechanisms where the machine learning model continuously learns from observed conversion patterns and adjusts attribution weights accordingly. The model analyzes which touchpoints historically led to conversions and refines its credit allocation based on this feedback, improving measurement precision over time while requiring minimal manual intervention to maintain operational ease.
Data Source
AI summary
A system including one or more processors and one or more non-transitory computer-readable media storing computing instructions configured to run on the one or more processors and perform: tracking touchpoints by a user over a first time period; after receiving an order, determining, using a machine-learning model, a respective contribution of each of the touchpoints, wherein the machine-learning model is trained to predict a probability of the user placing the order during a second time period based on an input feature vector representing a set of touchpoints; and allocating a respective percentage of credit for the order to the each of the touchpoints based on the respective contributions of the each of the touchpoints. Other embodiments are disclosed.