Computational Demand Leakage (CDL)
CDLPortion of Estimated Relevant AI-Mediated Demand Not Captured Due to Representation Deficits
CDL measures AI-mediated demand not captured due to representation deficits.
Definition
CDL measures the portion of estimated relevant AI-mediated demand that is not captured due to representation deficits. CDL = 1 − Captured Relevant AI-Mediated Demand / Estimated Relevant AI-Mediated Demand. The denominator is latent and must be independently estimated. CDL must not be defined simply as 1 − ARS.
CDL quantifies the gap between estimated AI-mediated demand and captured demand. Distinguishes transmission leakage from recommendation share.
Conceptual Formula
CDL = 1 − Captured Relevant AI-Mediated Demand / Estimated Relevant AI-Mediated Demand. Denominator is latent and requires estimation.Methodology
Type
index construction
Data Sources
Confidence Level
low
Description
CDL = 1 − Captured Relevant AI-Mediated Demand / Estimated Relevant AI-Mediated Demand. Denominator is latent and requires estimation.
Limitations
- Denominator is latent and must be estimated
- Not simply 1 − ARS
- Requires estimation methodology
Key Takeaways
Key Points
- CDL is leakage, not share
- Denominator requires estimation
- Distinct from ARS
Target Audience
Relevance Tags
Source Paper
Citation
For the Computational Demand Leakage (CDL), see HomeSelf Research (2026), The Balance-Sheet Economics of AI-Mediated Demand.