Abstract:
|
With outdated assumptions governing the operation of Basic sample selection, actions were needed to counter-act the mounting inefficiencies and response rate impacts effecting Nielsen Audio's PPM (Portable People Meter) service. First, a correlation analysis was conducted that identified a one point consent rate decline for every sixteen days between the initial contact disposition and the subsequent recruitment attempt. Next, a safety stock algorithm was added to the process which calculates optimal inventory levels by taking into account selection rate and lead times. Also, a Rolling Horizon Optimization was implemented which utilizes an optimization function to minimize sample selects and to minimize the variation in sample select volume across weeks. Reducing the average days between contact is projected to yield up to a two point response rate gain over the next two years, while the sample efficiency efforts are projected to increase sample efficiency by 20% during the same time period.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.