Web Traffic Peak Attribution via Trend Line Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for attributing web traffic to TV and radio advertisements are inaccurate and fail to effectively identify and divide peaks in online traffic among multiple ads aired at similar times, leading to skewed results and incomplete attribution of sales.
Innovation Solution
A computer-implemented method that receives traffic data and advertising spot logs, fits a trend line to the data, calculates variance, and uses advanced thresholding techniques to identify and attribute traffic peaks to specific ads, considering factors like cost and demographic data to accurately attribute traffic and sales.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If standard deviation and local average techniques are used to identify traffic peaks, then the method is simple to implement, but the results are easily skewed by traffic spikes and fail to capture the base of peaks
Solution Approach 1:
The patent changes the statistical parameters from simple standard deviation and local average to a more sophisticated model that fits a baseline trend line and measures deviations from it. This allows the system to distinguish between normal traffic variations and actual advertising-induced peaks, resolving the contradiction by improving measurement precision while maintaining reasonable implementation complexity through automated statistical analysis.
Solution Approach 2:
The patent replaces the mechanical threshold-based peak detection system with a statistical modeling system that uses trend line fitting and variance analysis. This substitution enables the system to capture the entire peak structure including the base, rather than just the tip, thereby improving measurement precision without significantly increasing implementation difficulty.
2Device complexity
If conventional peak detection methods are used, then the computation is straightforward, but the method fails to attribute peaks to multiple advertising spots aired at similar times
Solution Approach 1:
The patent segments the web traffic peak into multiple attribution components by analyzing the temporal patterns and correlating them with multiple advertising spots. The system divides the total peak traffic among different ads based on their individual impact patterns, thereby preventing loss of attribution information while maintaining manageable computational complexity through systematic decomposition of the attribution problem.
Solution Approach 2:
The patent adds a temporal dimension to the analysis by examining traffic patterns over time windows that encompass multiple advertising spots. This dimensional expansion allows the system to distinguish between traffic caused by different ads aired at similar times, improving attribution accuracy without requiring excessively complex computations through efficient use of time-series analysis.
3Quantity of substance
If viewer panels with small sample sizes are used, then the monitoring cost is reduced, but the representation of the total population is insufficient leading to inaccurate attribution
Solution Approach 1:
The patent uses web traffic data as an intermediary indicator to infer advertising effectiveness without directly monitoring large populations. By measuring the actual impact on web traffic and using statistical models to attribute this traffic to specific ads, the system achieves accurate measurement without requiring large sample sizes, thereby resolving the contradiction between sample quantity and measurement precision.
Data Source
AI summary
Systems and methods are disclosed for attributing web traffic to an advertising spot. The method may include receiving traffic data for a web page from a server associated with an advertiser and receiving, from a log provider, a log of a plurality of advertising spots related to the advertiser. A duration of time as a peak may be designated to identify the amount of traffic that is attributable to the one of the plurality of advertising spots.