Free-floating carsharing systems have become increasingly popular over the last years. The new mobility offer was launched without performing comprehensive target group analysis and establishing a well-structured fleet management. This dissertation reflects and optimizes the learning-by-doing-process by giving answers to the questions which external influences have an impact on the booking demand and how booking forecasts can be performed with time series analysis models. The basis of this work is booking data from a free-floating carsharing operator for the period of November 2011 to December 2014. At the beginning, the data is analyzed in detail on temporal and spatial level. The external influences are next to the weather land-use data, the citizens' election behavior and the local parking situation. The impacts of parking management zones and the weather turn out to be rather negligible but the other data allow to draw conclusions about the carsharing user through regression models. The typical characteristics of the customers resemble those which are found out by numerous methodologies in literature: financially well off and open for new, sustainable technologies. The time series analysis performed better by modeling with exponential smoothing using a Holt-Winters-Filter than with ARIMA. For model calibration it is sufficient to use booking data from a period of three months.    «

Free-floating carsharing systems have become increasingly popular over the last years. The new mobility offer was launched without performing comprehensive target group analysis and establishing a well-structured fleet management. This dissertation reflects and optimizes the learning-by-doing-process by giving answers to the questions which external influences have an impact on the booking demand and how booking forecasts can be performed with time series analysis models. The basis of this work...    »