We use a stratified customer survey data to study the patterns of residential telephone subscriber listings. A conditional logit model to capture the unobserved individual-choice-specific effects is suggested to provide revenue impact analysis under different price scenarios.