Web log file shows the behavior of user when they access the website. It is in text format and automatically generated by web server whenever user accesses the website. Entry of particular web page with date, time, cs-method, cs-uri-stem and other information is generated automatically in text file for every access of web page. Web Usage Mining is different mining technique apply on weblog file to discover the different pattern which is useful for efficient design of website, performance enhancement of server etc. Mining of web log file consist of three steps Data Preprocessing, Pattern Discovery and Pattern Analysis. Data preprocessing task convert the web log file in database by applying data extraction, data storage and data cleaning technique. In existing technique data extraction from the log file is perform line by line and store in multidimensional array and then extracted data field will be store in database and then data cleaning is performed. In this research paper, a new technique is being proposed for Data Preprocessing of web log using Microsoft Excel. The experimental results show that the proposed new technique using Microsoft excel file is an effective compare to existing technique.