The paradox between the ubiquity of polyploid lineages in plants and the early obstacles to the establishment of polyploids is a long‐studied yet unresolved question in evolutionary biology. It is assumed that to successfully persist after emergence, newly formed polyploids need to display certain fitness advantages and show a certain extent of reproductive isolation with their diploid progenitors. In this study, we tested whether immediate floral changes following polyploidization can improve pollinator visitation and enable pollinator preference leading to assortative mating, i.e. build a premating reproductive barrier between diploids and polyploids. For this purpose, we generated synthetic tetraploids of Arabidopsis arenosa and measured insect visitor behavior on diploids and synthetic tetraploids. We found that the increased floral size that accompanied polyploidization did not lead to a measurable increase in visitor preference, with insects visiting diploid and tetraploid plants at equal frequency. Despite this observation, tetraploids set more fruits than diploids, suggesting a positive impact of polyploidization on pollen transfer via other means. In addition, polyploidization did not lead to assortative mating but instead promoted interploidy pollen exchange, since visitors preferentially switched between cytotypes rather than preferring one. Consistent with this switching behavior, most of the progeny from tetraploid plants were triploid. Our data suggest that polyploidization has an immediate impact on mating in plants, but in a more complex way than has been assumed previously.