Our primary goal is to model office and retail userís preferences over various locations in greater Helsinki region. Acquired results will then later be tested with other Finnish major cities. Primary data is acquired with phone and Internet surveys for each user group. Survey respondents are selected to be those responsible of corporate/chain location decisions. Awaited number of responses is c. 100 for each group, consisting of all major retail chains and office space users. Questions are formed as multinomial statements on 9-step Likert scale and results are processed with applied AHP approach in order to come up with weight vector for qualitative variables. Weight vectors are then supposed to be concise representations of the importance of various qualitative attributes in making location decisions. The approach differs from ëtraditionalí AHP approach in a way that it uses ordinary multinomial statements instead of pair wise comparisons. Approach was selected in attempt to diminish the number of questions each respondent is expected to answer. By using acquired weight vectors it could be possible to construct a spatial preference geo-index for each submarket. The appropriate qualitative metrics that could be used for geo-indexes will be discussed more thoroughly. An attempt to verify these results with revealed preference methods ñ more specifically hedonic pricing method ñ will be made using KTI Finland rent benchmarking data and GIS-software.