Online marketers are continuously bombarded with new tactics that they “should” use if they want to be state-of-the-art. Most of these wonderous new tactics sound great and some actually are highly effective. However, many mid-size e-tailers have small online marketing teams (in many cases only a few people) that do not have enough hours in the day to adopt and become proficient with the tactic du jour.
The Perricone MD Case Study
At the Internet Retailer Conference & Exhibition’s email marketing workshop, iPost customer Neil Kjeldsen, vp of ecommerce for Perricone MD, spoke on segmentation and gave a very actionable case study built around problem, solution and results (with actual specific numbers). Perricone MD is a prestige skincare company that was founded by Dr. Nicholas Perricone, a dermatologist, award-winning author, and anti-aging expert.
Kjeldsen helped online marketers cut through some of the clutter and find a way to efficiently reignite relationships and sales from dormant customers. When Kjeldsen joined Perricone MD, he found that the vast majority of their customers were inactive - something that Perricone MD had in common with most companies. To make matters worse, these inactive customers were far more likely to opt out and flag Perricone MD as a spammer - hurting both the company’s ability to market to them in the future and Perricone MD’s email sender reputation with ISPs.
Kjeldsen’s solution was to adopt iPost as Perricone MD’s new email service provider (ESP) and to implement the iPost Customer Engagement Program, a combination of strategic services, processes and software. Kjeldsen said that iPost first worked with Perricone MD to analyze their email, web analytics and purchase data. Perricone MD then implemented iPost’s Autotarget, which automates and updates daily Recency, Frequency, and Monetary value (RFM) segmentation.
Using Autotarget to automatically group customers into micro-segments, Perricone MD divided customers into engaged and unengaged segments. First Perricone MD started mailing safely (engaged customers tend to have very low opt out and spam complaint rates) to the engaged customers and tested a variety of methodologies to see what best improved revenue and campaign profitability.
Perricone MD next worked with iPost to develop a “trail size” promotion for the unengaged customers and began mailing to them in “safe” segments. The safe segment approach is central to iPost’s Rapid Reputation Repair process and enables e-tailers to mail to questionable lists without damaging their email sender reputation. According to Kjeldsen, in one month Perricone MD was able to implement and optimize the engagement programs. Because Autotarget automatically segments customer data daily, Kjeldsen said he and his team found that it only took them 5 minutes per mailing to use Autotarget.
Kjeldsen's presentation made a clear case for actionable analytics
Kjeldsen's presentation made a clear case for actionable analytics
Kjeldsen stated that the results from the iPost Customer Engagement Program included:
* A huge increase in conversion rates (from 0 to up to 33 percent)
* A dramatic jump in revenue per email sent from $0 to $0.26
* Growth in online revenue from unengaged customers from 0 to 7 percent
The Bandwagon
Amusingly, another ESP in the session presented a case study on the benefits of RFM that used data from another iPost case study on the Eastwood Company. To work around the fact that it was actually an iPost case study, the vendor referenced a DMNews article from October 2008. I encourage you to check the article out!
The ESP’s pundit, also, told the audience that RFM is - as a process - very easy to do, something that Kjeldsen countered when he got up to speak. Kjeldsen said that he used RFM analysis in previous jobs and loves Autotarget because it makes RFM easy and that without Autotarget RFM is simply not easy for small and mid-size teams.
The Workshop in General
I found the presentations to be far better at the workshop than the general sessions. The presenters came across far more knowledgeable but were still not consistent in highlighting quantitative results and making it clear what all is involved (time and resource) to achieve those results.
