The
massive volume of online analytics data about customers has led to
novel opportunities for user segmentation. However, getting real value
from data remains challenging for many organizations. One of the recent
innovations in online analytics is data-driven persona generation that
can be used to create high-quality human representations from online
analytics data. This manuscript (a) summarizes the potential of
data-driven persona generation for online analytics, (b) characterizes
nine open research questions for data-driven persona generation, and (c)
outlines a research agenda for making persona analytics more useful for
decision makers.