In search engine optimization (SEO), advertisers bid on tons of keywords on Google, for example, so that their clickable ads can appear in Google's search results. In order for the advertisers to maintain their budget more efficiently and achieve the best performance, Google Adwords provides all the performance data for the advertisers’ keywords, but the advertisers need to know what is the best amount they should bid for each keyword, and what would be the keyword performance with that bid. A very important task for advertisers is to use the historical data to predict the click-through rate (CTR) and average cost per click (CPC) for a keyword with a set of features. The CTR and average CPC are two essential metrics to measure the paid search performance on keyword level, so that the advertisers will be able to optimize the bids and achieve the highest profits for their Google Adwords accounts. We try to predict the CTR and average CPC of a keyword using different machine learning methods. We use cross validation to evaluate the results and find the optimal predictors for the CTR and average CPC.