Abstract: To improve pricing models for location based services, established geocoding/reverse geocoding services are used to enrich the locations defining the price item (start location, multiple end locations) and the pick-up/drop-off location with geocoding information like zip codes, coordinates and geographic boundaries. A multi-step matching algorithm then uses the geocoding values to determine 1) if there is a match between locations and 2) how good of a match there is. The match is measured by its “order.” For example, if the places are identical, they have a matching order 0, identical coordinates a matching order 1, matches based on zip codes have a matching order 2, matches based on geographic boundaries have a matching order 3, etc. Other sub-criteria, such as the number of zip codes or the size of an area, may also serve as tie-breaker. The order of the match may then be used to establish the correct predictive pricing model.