Can Investors use Social Media Data?


New York – Primary research services have taken a hit in recent months, as a result of the concerns of buy-side institutions about the stigma that could be created by using services that might be able to uncover insider information.  On the other hand, there continues to be an obligation on the part of buy-side managers to produce results for their clients.

Part of the investment process relies on identifying themes and generating ideas to capitalize on those themes. Yet, once this conceptual process has been completed, there remains the task of confirming the thesis. Typically this has been done though assessing primary data, such as conversations with experts, supply chain (channel) checks, or survey data. Assuming this is a necessary part of the investment process, how can investors check their hypotheses without the use of traditional primary research?

The answer is non-traditional primary research.

The vast acceptance and use of social media sites, such as Twitter and Facebook, is well known. We also know implicitly that this information is both primary and potentially helpful to the investors, but are unsure how to harness this data to best advantage. While still a young field, there are a number of data mining services that take the raw Twitter, Facebook, YouTube, Chat room data and analyze the “buzz” around a particular product or event. This is relatively straight forward in the case of a new consumer product or the release of a new movie.

Where there is some concern is that the democratization of the media, means that the source of the information is somewhat uncertain. For example, if a movie has a large and growing pre-release buzz, is this a result of consumers chatting up the movie, or is it a result of marketing firms tweeting up the movie?  So while we can assess the level of buzz, we are not sure if it relates directly into demand for the product or not.

There are a number of research providers that have created inventive ways to interpret the social media data, including but not limited to firm like PeopleBrowsr, Lexalytics, Radian6, Media Vantage/CNW, Peanut Labs and Scout Labs. These firms all have relationships that allow them to receive data from social media sites and analyze the data for clients. These services rely on textual analysis and sometimes analysts to assess the tone of the comments and ascribe a positive/negative value concerning the level of intensity of the comments for and against a particular product or service. Others actually embed pop-up questions into online games, like Farmville.

This is a very promising area of primary research, but really only for goods and services that touch consumers directly and hence potentially provide buying intensions of online public. There are of course concerns with the reliability of the data. As with any data analysis, garbage in, garbage out applies. Because the analysis is relatively new, there will need to be extensive testing of the relationship between buzz and buying habits. Additionally it is important to understand what demographic the majority of the data points come from and adjust the results accordingly.

For investors looking to confirm their investment hypotheses, social media data mining represents a new and interesting area of primary research.  Just how reliable the information proves to be is still up for grabs.



About Author

Leave A Reply