Behavioural Analytics

Human Behaviour

Focal Research has been involved since 1991 in developing models that predict human behavior. Starting with data sets generated through large scale surveys and now using enormous customer data sets up to the terabit size, we have worked with organizations in Canada and around the world to develop models that accurately classify and characterise people. Dr. Tony Schellinck was the Management Informatics Chair Associate at Dalhousie University from 1997 through to 2011 during which time he and his graduate students worked with Canada Life, Meloche Monnex, Canadian Tire Acceptance, Bank of Montreal, Loyalty Management Group, Sobeys, Lawton’s, and the YMCA to help develop their CRM systems and create numerous models for deployment in the field. Learning from the experience of working with these large data sets Tony and several of his graduate students who were subsequently hired by Focal were able to apply advanced techniques to developing algorithms for Focal clients.

Focal uses where appropriate logistic regression, association analysis, neural networks, clustering algorithms, decision trees, memory based reasoning as analytics. We also create programs in SPSS, SQL, Access and other database packages in order to process customer data and to facilitate their deployment of the algorithms. To accommodate these initiatives we have created a safe and effective computer storage and retrieval system. We have also retained highly trained (Masters and PhD graduates) employees and associates who work as a team to create these algorithms. We have recently started creating interfaces/dashboards for clients to facilitate use of our algorithms where this does not conflict with our agreements with partner firms.

Focal has developed the ability to translate psychological theory related to the behaviours being studied to create as many as 500 variables measuring relevant behaviours or characteristics that can be used to predict behaviour or classify people. In some applications we have been able to achieve near 100% confidence in our classifications using our algorithms once they are deployed in the field. After a decade of developing and deploying these algorithms Focal has developed an effective and cost efficient system for creating, testing and deploying these tools in the field to the benefit of our customers.