Researchers often refer to investment habitats/categories to explain the patterns of comovement in asset returns that cannot be fully clarified by fundamentals. Many factors determine these habitats including investor preferences to size, industry, price-levels and risk-levels. This paper investigates a unique method to explore investment habitat based on the search behavior of investors on the Internet without actually using proprietary trading data. Using Yahoo! Finance data on investors' frequently coviewing stocks of Russell 3000 stocks between September 15, 2011 and February 24, 2012, we evaluate the return co-movement within investment search habitats/clusters. We find that stocks within a search cluster show strong co-movement with other stocks in the same cluster that is not fully explained by other traditional habitat characteristics. The behavior persists even when a stock moves from one cluster to another. The study provides direct empirical evidence that investors prefer to search among stocks that have similar co-movement.