This presentation will demonstrate a different way to blend sample, using psychographics and other factors to create stabilization variables which can be used to create consistent samples, regardless of the number of sources used. Data from offline sources will be compared with that obtained from blended online sample to demonstrate whether or not a rich online source can improve on the performance of a panel against external benchmarks. Finally the presentation will consider the industry’s conflict between the need for better data and the need to avoid changes in the data; two often competing goals. Some practical steps will be presented for integrating improved methodology while minimizing disruption for the research buyer.