First, we must define online profiling, also known as contextual marketing in advertising circles. It is a marketing and data-gathering process which “takes into account the users’ needs, habits, and goals to create a personalised web experience” (Elhawli, 2017). While it may not appear to be what conventional wisdom defines as business intelligence, the same principles are there: data is gathered, made sense of, and served to an information consumer in order to increase the company performance. The information consumer, in this case, may be considered either the actual user of the website or the layer of web platform between the data (recommendations) and website user.
The steps necessary for making the items actionable follow typical business intelligence processes. First, the “sheer volume of data now available to marketers” requires focusing on what data is relevant to the desired outcomes (Pophal, 2014). These outcomes and relevant data vary by market and platform. In many cases, the data itself yields important clues on what exactly is important—akin to an exploratory data analysis in traditional business intelligence implementations.
Another similarity is the platforms upon which the data must be presented. This may be considered a parallel to the various visualization and reporting platforms available to an information consumer. There are multiple data points involving user device and platform (i.e., Mozilla on Mac OS or Chrome on Android), which must then be utilized to determine how the information is best served up.
These data points also influence the intake process, parallel to the ETL stage in traditional business intelligence. Imagine the “multiple inputs around consumers and the devices they’re carrying-or wearing-that receive those inputs” as systems of record contributing to a master data aggregator (Pophal, 2014). These must work in near real-time, driving outputs and giving context to other inputs.
Moving from descriptive to predictive analytics is another similarity between online profiling and conventional business intelligence. Current online profiling “is largely driven by what consumers have done, the future will focus on what they will do” (Pophal, 2014). The transition from descriptive to predictive is a milestone in business intelligence maturity (LaValle, et al., 2011) and this is no different in online profiling. Serving up mountains of What Happened? insights can only go so far. Transitioning from that to So What? and Now What? crosses a hurdle into the next realm of usefulness. This cannot be at the expense of best practices, though—as the celebrated rise and embarrassing fall of Google Flu Trends reminds us (Lazer & Kennedy, 2015).
Elhawli, A. (2017, October 12). Why content & contextual marketing are key to your success. Retrieved February 2, 2019, from https://medium.com/@aelhawli1/why-content-contextual-marketing-are-key-to-your-success-4509239b71b9
LaValle, S., Lesser, E., Shockley, R., Hopkins, M. S., & Kruschwitz, N. (2011). Big data, analytics and the path from insights to value. MIT Sloan Management Review, 52(2), 21–31.
Lazer, D., & Kennedy, R. (2015, October 1). What we can learn from the epic failure of Google Flu Trends. Retrieved February 2, 2019, from https://www.wired.com/2015/10/can-learn-epic-failure-google-flu-trends/
Pophal, L. (2014). The technology of contextualized content: What’s next on the horizon? Retrieved from http://www.econtentmag.com/Articles/Editorial/Feature/The-Technology-of-Contextualized-Content-Whats-Next-on-the-Horizon-99029.htm