High fidelity website research : using a browser extension to provide a natural environment

[摘要] People are spending ever increasing amounts of time online. As a result, companies are investing greater amounts of money into online advertising in an effort to influence their behavior. The impact and effectiveness of these ads is still an open question. One possible method of analyzing the effectiveness is through the analysis of clickstream data. However, this data may be difficult to obtain and does not measure behavioral change. Behavioral change is a change in consideration or preference. Surrogate sites can be used to study behavioral change but are difficult and time-consuming to create and do a poor job of mimicking certain classes of sites. This is particularly true of social media sites where the amount of content is impossible to fully reproduce and yet this content is the defining component of the web site. In this thesis, I present a Chrome extension that can be used for conducting high fidelity web site market research. The framework provides the opportunity to measure behavioral change and provide a natural environment almost identical to the actual sites. I detail the implementation of the extension and its use in a web-based media experiment with a sample size of 13,000. Preliminary results and learnings are discussed. Results suggest that use of the extension is feasible and is capable of producing significant changes in consumer consideration.

[发布机构] Massachusetts Institute of Technology