The probability based web panel Elipss ensures that all respondents answer identically designed questions by equipping the entire panel with the same touch-screen tablet. However two specific challenges have to be overcome: to deal with the heterogeneity in digital proficiency of a general public and to meet the expectations of researchers in this digital mobile environment. In this paper we will present and critically discuss the choices made during pilot study of Elipss to conceive and produce an appropriate design for questionnaires.