• by Laurie Sullivan , Staff Writer @lauriesullivan, December 8, 2010

Historically there have been two preferred ways to gather information from consumers online. Either observe them like AudienceScience and BlueKai and infer something, or ask them. Now there's another way. Crowd Science, which supports research technology and services for brands with an online presence like NN and Federated Media and Time Warner, and plans to package and release the technology during the next year as free research apps.

Co-founder and chief crowd scientist John Martin describes the series of "bit size" apps as being similar to Google Analytics for market research. The first two apps -- Who, and Sat -- were released last week.

Perhaps they are crazy names at first glance, but once marketers know the Who app identifies visitor demography and psychographics, and the Sat app looks at site visitor satisfaction it all makes sense. The apps and market research aim to empower the long-and-mid tail of Web sites.

Co-founded by Martin, Paul Neto and John Wainwright in 2007, the group of software engineers and researchers want to build on a series of ideas spawned from years at comScore, Apple, Netscape, IBM, and Cisco.

Crowd Science's market research in a box relies on Web analytics data and tries to make decisions on the amount of Web traffic, where it originates, and who and when to ask questions and what questions to ask.

The company blends analytics with survey research. The combination allows Crowd Science to grab information and power products to measure ad efficiencies and Web audiences for companies like Money, National Hockey League (NHL), Turner, and Better Homes & Gardens.

The data drives survey based research that can plug into SEO and SEM strategies. Depending on the companies Crowd Science supports, the research firm retains the rights to review some of the anatomized data in aggregate that helps to support research. There are about 6 billion page views coming through the system, Martin says. All the page views are eligible for certain types of research. On occasion the company will pull aggregate data across the network to analyze non-response bias and understand the differences in audiences.