Probabilistic Media Attribution System for Cross-Channel Conversion Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in measuring and optimizing television advertising, as customers often convert through channels other than where the ad was viewed, leading to difficulties in attributing conversions to specific media sources.
Innovation Solution
A system and method that infer attribution between competing media events by analyzing characteristics such as demographics, time, and other attributes, using a probabilistic model to attribute conversions to media events, even when some media events are unknown.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individualized tracking systems are used to attribute conversions to media events, then measurement precision is improved, but customer privacy is compromised
Solution Approach 1:
The patent extracts and removes individualized tracking mechanisms from the system. Instead of tracking specific users across channels, the system aggregates data at the population level and uses probabilistic models to attribute conversions to media events based on demographic and temporal characteristics without identifying individual customers.
Solution Approach 2:
The patent introduces probabilistic models and aggregation layers as intermediaries between media events and conversion tracking. These intermediaries process data in a way that preserves privacy by removing individual identifiers while maintaining the ability to measure conversion attribution through statistical relationships.
2Device complexity
If multi-channel media events are tracked separately, then device complexity is reduced, but measurement precision deteriorates due to inability to attribute cross-channel conversions
Solution Approach 1:
The patent merges separate multi-channel tracking systems into a unified probabilistic model that handles television, digital, and other media channels simultaneously. The model integrates data from multiple channels and uses demographic-temporal matching to attribute conversions across channels without requiring complex individual user tracking infrastructure.
3Adaptability or versatility
If probabilistic models are used to attribute conversions, then adaptability to unknown media events is improved, but calculation time increases
Solution Approach 1:
The patent performs preliminary actions by pre-processing and aggregating media event data into demographic and temporal profiles before conversion events occur. This pre-aggregation allows the probabilistic model to quickly match conversions to media events using pre-computed statistics rather than processing raw data in real-time, reducing calculation time while maintaining adaptability.
Data Source
AI summary
This paper presents a practical method for measuring the impact of multiple marketing events on sales, including marketing events that are not traditionally trackable. The technique infers which of several competing media events are likely to have caused a given conversion. The method is tested using hold-out sets, and also a live media experiment for determining whether the method can accurately predict television-generated web conversions.