Essay one, Do Internet Board Messages Predict Stock Returns? An Analysis with Explicit and Intensive Board Postings, tries to discover whether postings on message boards predict future stock performance. While many other studies find significant predictive effects from message volume, they do not find significant predictive effects from message content, which is the most important and meaningful part of online messages. We improve sample selection, using a posting medium that allows for explicit buy/sell signals, focusing on firms with a high intensity of postings, and reducing posting noise by developing a credibility index. After controlling for factors which affect next day returns and incorporating credibility, we find that a bullishness index predicts next day stock returns at a 1% significance level. Essay two, Attention: A Better Way to Measure SEO Marketing Impact, differentiates itself from previous SEO marketing studies by using the first direct attention measure: the user search frequency index from Google Insight for Search (GIS). We find that as an attention proxy, GIS index has significant power in capturing impact around the time of the offering. A one point increase in the pre-issue GIS index change corresponds to a future reduction in the offer price discount by about 3%. In addition, as the only direct measure in SEO research, the GIS index differentiates itself from previous indirect measures by capturing some attention effects that are not captured by previous studies. All effects corresponding to the GIS index change are statistically significant and economically important. Essay three, Online Search Frequency, Information Asymmetry, and Market Liquidity, investigates online search frequency of Google users to explore how traders' response to information asymmetry would predict future market liquidity. The findings show that GIS daily index, as an effective measure of the level of information asymmetry, predicts the future liquidity level. This predictive power is significant at a 1% level. More importantly, the GIS indicators remain strong and significant after including traditional information asymmetry variables in all regressions. Last but not least, we find that the GIS index can capture investors' attention change and provide indications on future stock return.