Two experiments are reported that investigate nonlinear patterns of feedback effects on family brands across close, moderate, and far brand extensions. When extensions are unsuccessful, patterns of negative feedback effects are of inverted U-shape showing greater impacts in close and far extensions than moderate extensions. When extensions are successful, positive feedback effects show greater impacts in moderate or far extensions than in close extensions. The newly observed nonlinear patterns cannot be explained by affect transfer theory or diagnosticity theory. I find that perceived diagnosticity of extension information varies in nonlinear patterns similar to those of feedback effects. Our regression analysis indicates that the nonlinearly shaped diagnosticity explains the observed patterns of positive and negative feedback effects.