Accurately forecasting volatility is key in many financial applications. In this study, I suggest that individuals gather information online before implementing their trading decisions. In periods of higher investor concern, online information seeking intensifies. By analysing Google search data for a selected set of keywords, I find that changes in Google hits lead changes in market volatility. I show that a regressor based on search engine data can provide a meaningful complement to a two-factor EGARCH model. Results suggest that the augmented model significantly outperforms its restricted counterpart from a forecasting perspective.