Strategy: Learning From Fiascos
We all make mistakes. They are usually pushed out of sight and you don't hear much about them. Marketing articles are usually about successful campaigns, but there is a lot that can be learned from those mistakes.
1. Catalogers 'Mailing Down'
Story: A cataloger was trying to cut mailing costs on a comprehensive catalog, without losing revenue. That meant looking for ways to trim the mailing list.
Fiasco: The marketer ranked each of the catalog customers by revenue and then "mailed down" the list until the return was unprofitable. But costs kept rising faster than revenue.
Failure: Catalogers are under pressure from increasing print and postage costs, but more heavily from customers switching to the Internet for purchases. In this case, the cataloger failed to recognize that knowing a customer's main sales channel could offer a more profitable approach to a better mail strategy.
Lesson learned: Rather than using a historical metric like past revenue, use "expected revenue" from catalog sales to rank customers. To save postage, don't mail customers so often who are primarily buying through the Internet.
2. Ignoring Lifetime Value in Acquisition Marketing
Story: An e-commerce company with a preponderance of "one-and-done" customers was addicted to acquisition. The marketer was making deals left and right, in addition to offering rich promotions to acquire customers to feed the revenue stream.
Fiasco: This organization was giving away gross margin without realizing it because its marketers didn't know customer lifetime value (CLV) by segment. Its acquisition programs were costing more per new customer than the newly acquired customer would be worth.
Failure: Not recognizing easier revenue gains could be realized by smarter marketing to existing customers. But acquisition is important, too. Not knowing CLV took away an important data point from management in the evaluation of different acquisition programs.
Lesson learned: Knowing CLV can save a company a lot of money and help prioritize marketing efforts.
3. Trying to Cross-Sell Without Individual Analysis
Story: A large retail clothing chain realized many of its customers, including the "better" ones, were buying very narrowly from only a few categories. Marketers launched a cross-sell campaign based on RFM to promote purchasing across a broader spectrum of product categories.
Fiasco: The campaign flopped, with response rates under even the typical cross-sell response rate of 0.5 percent.
Failure: For a cross-sell campaign to work, customers need to be interested in the products. Generic offers won't induce purchase. There is no product information in RFM, so using those scores to determine an offering makes no sense. Demographics won't work here either, because not everyone in the same ZIP code will buy the same products.
Lesson learned: The way to make a cross-sell campaign successful is to offer products based on purchase propensities calculated from transaction data, one customer at a time. What someone is likely to buy is determined more by what he or she has previously purchased than by age or address. Your customer analytics must be done at the individual customer level for this to happen, and done at scale.
4. Testing, Sans Controlling the Variables
Story: A nationwide automobile tire retailer went all-in on individualized marketing, sending digitally printed postcards with individualized offers to targeted customers in two different markets.
Fiasco: The targeted customers who received the postcards did no better than a control group that was not mailed.
Failure: The control group was composed of similarly targeted customers who were not mailed. When a second control group of randomly selected customers was examined, it was clear that the targeting was accurate and the targeted customers purchased at a significantly greater rate than the randomly selected ones. However, poor collateral (a lot of legal caveats qualifying the offer on the postcard) evidently inhibited buyers. The postcard had no effect on the purchase rate of targeted customers. Unable to develop a more effective postcard, the company dropped the program completely.
Lesson learned: Your customer analytics could be great, but unless you can execute against them in a compelling way, you could be wasting money. Set expectations properly and use control groups to measure all the variables.
5. Wrong KPIs and Metrics
Story: A B-to-C cataloger recognized more of its newer customers ordered through the Web, but the marketer continued to focus most of the budget and efforts on the catalog postal channel, because the average order value (AOV) was higher for that channel than for orders coming through the Web.
Fiasco: Looking at the data again, the cataloger discovered that while AOV was higher for mail, annual spend per customer was 50 percent higher for emailable customers ordering through the Web. The marketer wasted millions mailing catalogs, impacted margins and slowed profit growth.
Failure: By looking at the wrong metrics, the company failed to notice this trend in a timely way. A likely explanation for higher revenue in the email channel was that the lower cost of email enabled more frequent communications and more total orders, even if AOV was smaller. Within two years, the Web was accounting for the majority of revenue.
Lesson learned: Don't blindly rely on what's worked in the past.
It's painful to think back about these mistakes. Marketers wasted money and missed opportunities, but learned a lot. We hope others can benefit from these experiences.
Mark Klein is CEO at Portsmouth, N.H.-based customer analytics software provider Loyalty Builders. Reach him at email@example.com.