If you’re an online retailer like Revolve Clothing selling hip, upscale fashion brands, your customers are all the same, right? They’re 18-35 year old men and women with a high discretionary income, much of which is spent on haute couture.
But those actually in the retail industry, like Kobie Fuller, CMO at Revolve Clothing, know that customers are far more nuanced than that. There are female customers who are boho, those who are edgier, and those who prefer a more feminine look. Although some may think that male customers are more uniform (pardon the expression), Revolve Clothing has developed various male-centric personas. “You have guys who are more neutral in their dress, more interested in buying high-quality denim,” Fuller says, “Denim is a staple, so they pair it with T-shirts. Then you have guys who are edgier and street. So, we’ll be a little rougher as it relates to the type of style they wear—more graphic tees and trendy sneakers.”
And deeper than aesthetic differences, Revolve must also contend with variations in customer buying patterns. How, for instance, do you tell when a customer might defect? And what can you do to make sure that doesn’t happen?
For roughly a year Revolve has been using tools from Custora, a provider of lifecycle marketing solutions, to better understand its customers and segments. “Based on the different activities and offers, that purchasing behavior can shift positively or negatively,” Fuller says. Prior to the implementation, Revolve did its customer analysis in-house and estimated how much each customer was worth. But the company needed a way to better test messaging and measure how effective those messages were during various customer lifecycle stages—these included everything from email to catalogs. For instance, during Revolve’s last holiday campaign, the company wanted to see how high-value customer segments responded. The insight from this campaign (Fuller declined to specify) will influence upcoming mail promotions.
Another issue Revolve wanted to resolve was identifying and reacting to imminent defections, first flagging any deviations in purchase activity, and then emailing trigger messages designed to re-engage.
“We ran an A/B test by sending [customers] an email with no offer, but that reminded them of Revolve and our value proposition,” Fuller says. “Maybe they liked to shop at Revolve to see the latest trends. We’d give them a broader picture, a reminder of their activity.”
Through this messaging strategy, Revolve saw 40% lift in value of the orders compared to the control group. For instance, if the average order of the control group was $9, that average increased to $13 among customers who’d received the message.
Ultimately, incorporating behavioral segmentation into a messaging strategy can be more valuable than demographic segmentation. Many brands, says Custora CEO Corey Pierson, erroneously emphasize the latter.
“Most brands make the mistake of applying the same rule to all their customers,” he says. “For example, it’s common for a retailer to say a customer is ‘at risk’ if the customer hasn’t ordered in the past 90 days. In reality, if a customer usually orders every week, that at-risk status should be applied if the customer misses a few weeks–waiting 90 days is far too long. On the other hand, for someone who shops only once a year, setting an at-risk status after 90 days is premature.”
Despite Revolve’s early success, Fuller sees room to hone his messaging strategy, mostly by better understanding customers and provide more relevant messaging, without falling into the dreaded “creep factor” that deters more than delights. This entails clustering more purchases to better understand buyer behavior, and then sending messages that are relevant but overly personalized.
“People look at [a personalized message and] know they’re being tracked,” Fuller says. “They looked at a boot, and now that boot is front-and-center on the homepage. Stop it. It’s not done in a subtle matter. It’s blatant.”
Revolve is still testing to find that happy medium between personalized and creepy, and to identify messages that keep customers clicking the Buy button. There are certain common messaging tactics that Fuller knows don’t work—like discounts, which fail to stand out in customer inboxes already overloaded with digital coupons and offers.
And while Fuller doesn’t have all the answers, he has ideas: “If someone is prone to shop for boots,” he explains, “maybe the messaging highlights it in a more subtle way. I’m not saying there’s no place for showing that same boot that they looked at, but a lot of times [the messaging] doesn’t work, especially if done in a product-centric matter. You have to be careful striking the balance and we’re honestly still working through it.”
