The research paper addresses two major objectives, one is to evaluate the results and the model used to conduct a multi-product direct mail campaign for Bank A. The second major objective is to build and improve on the original direct mail marketing model utilized to choose which postal codes to target with direct mail. The results of the campaign, including the distribution of respondents are studied within the Toronto CMA. A number of aspects of the respondents are examined including their proximity to Bank A branches to provide a clearer picture of some of their characteristics. The variables used to choose which postal codes to target are examined for their predictive power, against where actual response took place. Finally, a new model is created through testing customer and market-related variables. The new model was found to better predict actual response and thus can be used in future campaigns.