Ongoing Debates: Social Media, Agency Pay, Questioning Survey Questions

A recap of the 5 most retweeted links on the #MRX community this week:

HP Tests Predictive Capabilities of Social Media Data – HP has completed a pilot project, dubbed “Project Fusion”, in which it merged social data with internal company data to better predict customer behavior. The process involved using a proprietary text-analytics technology to bring structure to unstructured data (social media, Amazon reviews, customer surveys), after which the data was mined together with more typical customer data. 

What Facebook Thinks About Market Research – Tom Ewing recaps Esomar 3D’s session with Sean Bruich, Facebook’s head of measurement. The article offers insights into Facebook’s perspective on research, including the goals of Facebook research, the three kinds of data that Sean uses to accomplish these goals, and the company’s views on text analytics.

What the Foucault? Uses and Misuses of Social Media Research – Tom Ewing also provided his summary of presentations by Ray Poynter, Gregor Jawecki, and Annie Pettit – all on the topic of the good, the bad, and the ugly of social media research.

Should Research Agencies Be Paid for the Value of Their Insights? – Client-side researcher Edward Appleton outlines his personal perspectives on whether or not market research agencies should be compensated for performance. He wonders “if the debate over ROI for research is a productive one” and suggests that value perception is perhaps a more relevant context for understanding the cost drivers in the industry. 

Conflicting Data in Online Surveys: The Rorschach Test of Market Research – In Quirks' Research Industry Voices blog, Lev Mazin of Ask Your Target Market discusses strategies for obtaining more accurate data when surveying two notoriously tricky topics: brand preference and price sensitivity.