This paper explores how tourists value the presence of certain types of attractions in a city that they visit as part of a longer vacation trip. In developing new attractions tourism managers need to decide which additional attractions will add most value to the tourist offer in a city, which may already have multiple attractions such as museums, parks, shopping, amusement and historical sites. They also need to decide where to locate these attractions. Our study uses a simulated Google maps environment to present respondents with a hypothetical city offering a mix of geographically distributed attractions. Respondents indicate their preference for the city as a place to visit. Using latent class regression methods we identify for different motivational segments which attributes of attractions (e.g. location or co-location next to other attractions, star rating, recommended visit duration) contribute most to the overall appeal of the city as a destination.