Analysis Of The Real Estate Advertising Effect Based On Log Linear Model

The log linear model is a branch in generalized linear models, like the Logisticmodel, Probit model, plays an important role in the analysis of the qualitative data. Thispaper describes the basic principle of log linear model, and on this basis, makes adetailed analysis on the relationship between the real estate advertising types and visitscompanied with using of the VaR and the Monte Carlo simulation method. In the end,this paper draws some conclusions and puts forward reasonable suggestions.Through the research and analysis, the paper comes to the following conclusion:(1)Visit quantity showed regular changes with the month changes----increased first andthen decreased, and reached the peak value in February, then reached a trough inSeptember, it is suggested that put more advertising appropriately in February, andreduce the advertisement in September.(2) From the evaluation results, from high tolow is: Morning News, business, times, Soufangwang, SMS, evening news, activities,daily, sohu.com, and other media.(3) If the visits are little, advertising has a greatimpact on visits, and can bring greater growth; when the visits achieve a certain degree,the influence of the advertisement will be greatly reduced.(4) Choose six ads in16adsrandomly, we found, the gap between visits in the confidence level95%and the actualis not too large, therefore, for the visits of the ad can bring are no need to worry abouttoo much.(5) At last, this article recommend news, evening news, daily, times, SMS,Soufangwang as the best advertise and do the random distribution. In the process ofthese ads, advertising cost is the only factor which the company needs to consider.