An Events Precision Marketing Research Applied On The Small Businesses Credit Of Harbin Bank

Macroeconomic growth is slowing, while service leading economy isgradually replacing industrial leading economy and consumption leadingeconomy is gradually replaceing investment leading economy in2014. Monetarymultiplier of China is the highest in this year. It means that banks still need toface a high level of legal deposit reserve rate and handle great pressure. Thedocument of People’s Bank of China on further promoting the reform of interestrate marketization notice was issued.The people’s bank of China fully liberalisedfinancial institutions loan interest rates, so that deposits and loans of commercialBanks’ competition would reach a new height and the bank’s profits will decreasesharply.The weakness in capital and technology of city commercial Banks wouldmake them face a huge challenge. The events precision marketing will improvethe efficiency of marketing for long-term development of Harbin bank.The Events precision marketing applied on small businesses credit businessin Harbin bank application is an interdisciplinary research, involving ma rketingmanagement, economics, computer science, using commom research approachesasinterview, questionnaire, expert scoring, historical literature analysis, as wellas some qualitative research methods as C5.0decision tree algorithm model andboostings promote empirical methods. This bank is suitable to do this marketingbecause of its Very good database structure and data quality, excellent smallbusinesses credit experience, various products and mature channels. The eventsdatabase has been built based on the interview and loan due to the end datebecame the experimental unit. The explained variation was the selections toreloan or not and the explain variation included3parts having both numericvariations and numeric variations. Reloaning was more important in themodeling and forecasting because of the practical significance.The databasecould only provided part of the data, which had such problems as Outliersmissing value. Boostings and sample segmentation helped develop themodel.The prediction accuracy of training sample set and test sample set wasclose to95%., it was close to90%when the model was applied to predict newdata.It suggested that the model can be used to predict reloan customers,whichgreatly reduced the marketing costs and improve marketing efficiency.