The pollination specialisation/generalisation continuum is a basic concept in pollination biology, usually defined as the number of species or functional groups of pollinators visiting a plant species. The level of specialisation can be affected by the relative abundance of pollinators on flowers, the among-pollinator variation in per-visit efficiency and the intra-population variation in floral traits. Here, we explore how these components influence the degree of generalisation of a Mediterranean herb, (Brassicaceae). shows a remarkable intra-individual floral polyphenism, with large cross-shaped lilac flowers during spring and small rounded white flowers during summer. We quantified the pollinator assemblages, grouped into functional groups, of six plant populations from SE Spain during spring and summer, experimentally tested their preference for spring and summer floral morphs, their per-visit efficiency, and their effectiveness during each flowering period, estimated as their contribution to plant's fitness. The level of generalisation differed between floral morphs: spring flowers are visited mainly by large long-tongued bees, while summer flowers are visited by a diverse set including small short-tongued bees, large long-tongued bees, large butterflies and beeflies. The functional groups of pollinators differed in their preferences for floral morphs, with large long-tongued bees, small short-tongued bees and beeflies preferring spring floral morphs, whereas flies and butterflies preferred summer floral morphs. Pollinator per-visit efficiency also varied among functional groups. Bees and butterflies produced many seeds per visit, while beetles, hoverflies and flies produced few or no seeds per visit. By combining floral visits with per-visit efficiency (effectiveness), the spring pollinator assemblage became significantly more specialised and the summer pollinator assemblage significantly more generalised. Our study suggests that, although traditionally neglected in pollination studies, examining pollinator effectiveness is crucial to accurately characterise generalisation-specialisation gradients, rigorously categorising pollination niches and correctly describing the architecture of plant-pollinator interactions.