The term was originally coined by social scientist William McPhee in 1963 who observed the phenomenon, first in awareness and liking scores for Hollywood actors, and later in behaviours (e.g. reading of comic strips and listening to radio presenters).
[1] Shortly afterwards, Andrew Ehrenberg discovered the double jeopardy law generalised to brand purchasing.
[3] This empirical law-like phenomenon is due to a statistical selection effect that occurs if brands are broadly substitutable selling to much of the same types of people (often referred to as a lack of product differentiation and market partitioning).
The double jeopardy empirical generalization is explained and predicted by the NBD-Dirichlet theory of repeat purchase.
[6] The main implication of double jeopardy is that market share growth depends substantially on growing the size of a brand's customer base.