uSamp Unveils 'Adaptive Profiling' Relevancy Model

Online sampling specialist uSamp is launching a system called 'Adaptive Profiling', which involves asking questions to panellists based on their correlated attributes. The firm claims this will reduce the total number of profile questions asked by half, while doubling completion rates.

Using predictive analytics, Adaptive Profiling employs complex statistical analyses to build relevancy models between different data points. So, for example, if a client needs to target electric vehicle owners, uSamp can help target these individuals by examining previously correlated attributes of other groups, such as wearable technology users. uSamp says that the approach will also help with live incidence checking before each study, and once implemented, the firm expects overall time to field to decrease by 20% or more.

Andy Jolls (pictured), SVP of Global Marketing, explains: 'Most profiling models used by the industry don't paint an accurate picture of panel size and makeup. For instance, a panelist who is profiled as being pregnant may be counted this way for years because she hasn't been asked any follow-up profile questions. Adaptive Profiling is more accurate because it allows us to continuously collect data points on our panel and archive responses over time.'

Web site: www.usamp.com .