The methodological framework and data quality in Internet surveys is not yet completely investigated. This paper focuses on the role of the solicitation plan in the participation into the survey. The survey period length influences data timeliness; the period length is connected to the solicitation plan. Starting from this point of view, in our paper, the reaction to the soliciting message is analysed modelling the individual survey period length. The event of interest is the self-interview after the solicitation and the aim is to determine how the occurence and timing of response depend on several covariates. It is proposed to use the competing risk model for the effect of the solicitation plan . This model is applied to the response process of a list based Internet survey on firms in four provinces of the Lombardy region in Italy.