In this article, we numerically resolve the flow profiles of tastant concentration in the pipe of a gustometer used to deliver alternative pulses in concentration, which is a typical case of Taylor dispersion. Using this model, we can define the cases where the experimenter will deliver to the assessors a concentration profile which is significantly different from that intended. This can be simply assessed a priori using a scaling argument which involves calculating a dimensionless frequency. This is a function of the pulses frequency, the dimensions of the pipe and the flow rate used. We show that unless this parameter is taken into account, modifying the pulse frequency will modify the pulse amplitude. This design criterion is absent from the literature but we suggest this is important for designing such experiments.