Today recommendations are used to improve the quality and the number of interactive TV services offered by providers all over the world. Their main goal is to recommend TV shows and TV events, sometimes combined with an automatic recording function. With the growing number of IPTV offers, the usage of recommendation systems will increase and help to personalize and individualize the TV viewing experience. This article studies recommendations for the home context investigating daily living habits and routines in 40 households in depth using playful and creative cultural probing. Study results are present design recommendations for the development of new forms of recommendation systems. Main results find that users prefer individualized recommendations in-time, either automated or user-oriented, but mostly prefer individualized recommendations for each member of a household rather than personalized for the whole household.