In simple plurality voting systems smaller parties facing resource constraints may struggle to field candidates, particularly when the number of electoral districts is large. In the absence of a strong coordinating party organization, the pattern of contestation may also be sub-optimal – the small party fields candidates where support is minimal, ignoring other electoral districts where voters would support the party if it had stood a candidate. This article considers how and whether the separate operation of an Additional Member voting system running parallel to a simple plurality system assists smaller parties by providing low-cost information about the spatial distribution of voter support. Using aggregate voting data collected at the London ward level for both simple plurality local council elections and the separate mixed-member proportional system used to elect the Greater London Assembly we test whether the Green Party demonstrated an ability to modify and optimize its pattern of contestation. There appears to be little evidence of either a contagion or learning effect in respect of the location of Green candidates, suggesting that a free but valuable data source is unexploited.