Leveraging the power of data at Overstock.com
With data warehousing, customer relationship management and other technologies, marketers today are able to generate one-to-one personalized messages that dramatically affect the bottom line.
It is a fact that customers today have less time to shop and therefore demand more from their e-mail subscriptions. If e-mail marketers are not generating relevant content, customers can remove themselves from a business' only free tracked channel with a single click.
There is something that can be said about batch 'n' blast techniques if you subscribe to the more-sends-equals-more-revenue theory of e-mail marketing. However, until marketers harness the power of the hidden gems buried deep within the data mine, they run the risk of stale advertising and customer fatigue.
Mind the Data
About a year ago, we realized that we had a problem. The e-mail marketing department was responsible for a tremendous portion of total Web
site sales. But at the time there was nothing innovative about the process. There had been no attempts to tie multiple campaigns together, use customer lifecycle knowledge, leverage behavioral patterns or even deploy segmentation models.
We were in a holding pattern, driven by hunches rather than empirical evidence, which put a glass ceiling over our progress.
Management responded by dedicating resources to innovating our marketing efforts. The team, comprised of business users, data miners and designers, began exploring all data sources that could be tied back to specific customer activity touch-points.
The wheels really started turning after a business analyst overheard a conversation: "How great would it be if we were able to tie pages clicked back to the customer?"
Without the team's knowledge or suggestion, the data architects were, in a roundabout way, already collecting this information for unknown purposes.
Collaboration in Stock
So, the first lesson was to foster a working relationship between our marketers and the database team.
We immediately began leveraging data like never before, creating event-triggered dialogs, content-based recommendation engines and simple collaborative filtering algorithms.
We were able to generate personalized product recommendations, fire off reminders and effectively segment our customers based on a fusion of purchase histories and click logs.
Once we began testing how effective we could be when armed with data, everything that we hypothesized would be a good idea resulted in performance lifts, ranging from 5 percent to 300 percent, with an overall measured effect approaching 10 percent.
The control groups became a necessary evil in that they incurred an opportunity cost while being an essential piece to measuring return.
The initial investment of a data warehouse can seem insurmountable from an ROI perspective. However, with dedicated resources and measurable experimentation, the results in both direct marketing efforts and long-term customer retention will provide the farthest reaching net benefit to your business: a more efficient customer experience.