We all use purchasing data to guide our marketing plans. The recency since last purchase, frequency of purchases, and the total amount spent have been the cornerstones for predicting customer performance in direct marketing for over 50 years. We have gotten better at refining RFM using other purchasing signals, such as products or category purchased from, but our planning has been based entirely on past purchasing behavior. All of our planning is based on the 3 to 5% of the customers who responded.
What do we know about the 95% who didn’t respond? A few of years ago, we began collecting browsing data from many of our clients’ websites and started asking questions about the value of current web behavior as a tool for optimizing our clients’ marketing plans.
In analyzing the data collected over a six-month period, we made some interesting finds. First, we found recency of visit was predictive of response to a mailing. That made sense; they had an interest in an item and were browsing the site to finalize their decision. Going deeper, we found more orders came from the shoppers who visited in the first 4 months than came from shoppers who visited in the last 30 days. Maybe they got distracted while shopping and the mailing reminded them of the items they wanted earlier?
We then looked at frequency of web visits and found this to be highly predictive of response to a mailing. In fact, a four-time visitor to the site was 10X more likely to make a purchase compared to non-browsers. Whereas, a one-time visitor in the last 30 days was only 2.1X as likely to purchase. A two-time visitor in the last six months was more likely to purchase than a one-time visitor in the last 30 days. A significant portion of these shoppers were mulling over their purchase. Maybe they wanted to make sure they coordinated the drapes with the pillows? Valuable insights.
We checked out other measures, like how many times someone used onsite search. It too was very predictive. The response index for performing 2+ site searches over the last six months was again higher than visiting in the last 30 days.
Today, we are leveraging all of these behavioral tells and more to help our clients model their website visitors and maximize the ROI for their marketing programs.