Rewarding Experience

When Ramer Holtan joined Drugstore.com Inc. in 2000, his current title—director of customer retention—didn’t even exist. But that was before the Internet investment bubble burst, sending a slew of dot-coms into oblivion and forcing those that survived to rethink strategy.

Today—in contrast to earlier days when the Bellevue, WA-based Internet drugstore and just about every other online retailer were out to acquire new customers at seemingly any cost—retaining customers is a primary driver. “Getting back to business fundamentals dictates that you have to recover those acquisition costs to make a profit. So we’ve implemented a huge shift in our strategic and marketing focus to do that,” says Holtan.

Program in three parts

At Drugstore, a carefully developed loyalty program focused on non-prescription merchandise carries much of that load. To keep customers coming back, Drugstore has put deals and discounts at the center of a three-part initiative that it institutionalized a year ago:

— Free shipping on purchases of $49 or more

— “Diamond Deals,” an assortment of six products likely to interest the shopper based on past purchases and

— “Drugstore Dollars,” a 5% rebate that can be applied to future purchases.

The offers have had an impact: Free shipping, for instance, has helped raise the average basket size by 12%.

But any retailer offering points, rebates, deals and discounts to attract shoppers walks a fine line between pricing to entice and giving away too much, and Drugstore has yet to earn a profit. “Part of the problem is their business is built on scale, and they are not getting the scale they need,” says Stewart Barry of the Seattle-based investment banking firm Delafield Hambrecht. “It’s easy to sell something when you are giving it away, so they are doing a lot of things like discounting to get the volume up, but then it doesn’t make them any more profitable.”

There’s no denying that Drugstore has posted a string of quarterly net losses though sales continue to rise. Second-quarter net sales this year, for example, reached an all-time high of $58.2 million while the company reported a $4.6 million loss. Yet the gap continues to narrow, with Q2’s losses a significant improvement over a $14 million net loss in the year-ago quarter. Holtan sees the loyalty program and its cost not as a treadmill but as a path toward profitability.

“Loyalty programs are popular in many retail industries. Do they cut into your margins? Yes, absolutely,” Holtan says. “So the goal has to be that you make up for the reduction in your margins by getting more reorders out of your customers.”

To hit that sweet spot, Drugstore performs an ongoing analysis of program performance. It conducted extensive testing of customer reaction before settling on the loyalty program it has now. The free shipping offer, for example, has ranged from no minimum purchase requirement to one of $75, and from being posted periodically or delivered via targeted e-mail to being a standing “everyday” offer.

What Drugstore found was that all free shipping offers got a positive response from shoppers, so it kept increasing their frequency. At one point, Drugstore was sending free shipping promotions via e-mail almost every other week to most of its customers. The popularity of the promotion started to backfire as the number of customers who would visit without the prompting of an e-mail declined. Without intending to, Drugstore had trained customers to wait for a promotional e-mail before buying.

Promotional dependency

That was the downside of the periodic promotions, says Holtan. “Customers were becoming dependent on the promotions, but what we were really after was to get people coming to the site on their own without having to be lured in,” he says. Instead, Drugstore decided on everyday free shipping with minimum purchase. The $49 threshold was set at slightly higher than the average order size; Drugstore correctly guessed that a significant number of customers—Holtan won’t say how many—would stretch their order size by the difference to take advantage of the offer.

IRCE 2008 Conference Multi-Media CD-ROM

Once it had a base of regular customers, Drugstore created its Diamond Deals loyalty program, built on the foundation of the site’s longstanding virtual sales circular, Great Deals. Holtan and his team noticed the feature was one of the most popular on the site, and also that, depending on the number of promotions offered and the type and breadth of categories represented, conversions off Great Deals rose and fell. The broader the category representation and the greater the number of deals, the better the conversions. “We thought, why don’t we make a circular that is guaranteed to have the products customers care about and put them on sale,” Holtan says. “Diamond Deals came from a blend of those two ideas: personalization with the clear desire of Internet shoppers to save money.”

The online advantage

Using algorithms to analyze a customer’s past purchases, Diamond Deals present a registered customer with a personalized grouping of six products. Customers can apply a dollars-off discount—the size of the discount varies weekly, as does the product mix in the selection—to one of the six items.

Internally written applications, including personalization and product affinity algorithms, drive the product selections served up for each customer. Drugstore’s experience confirms one of the elementary precepts of marketing: The more relevant and personalized the product, the greater the customer response. Further, the more history Drugstore has on a customer, the more spot-on the deals are to that individual’s tastes and the greater the likelihood of conversion. But the advantage in crafting offers that Drugstore and other online retailers enjoy is that they have a complete purchase history for each customer, unlike in stores where the record may be spotty at best, and so they can create offers with a high degree of confidence. “When the proportion of products in a customer’s Diamond Deals is more heavily weighted toward products they’ve bought before, or products that are related to them, the response is much better than when it’s more weighted toward products we just want them to see because our store carries them,” says Holtan.

Diamond Deals serve a purpose beyond enticing repeat customers with personalized discounts and deals. Drugstore also has experimented with using the heavily visited feature to introduce shoppers to new products and other areas of the store. Though conversions drop on items less familiar to customers, Holtan says that doesn’t necessarily mean the items don’t belong in Diamond Deals. “If you’re looking just for eyes on new products and new product categories, it’s a good place for them to be,” he says. “In that case, we put them there with the knowledge that we are trying to accomplish something different. I’d liken it to the difference between direct response marketing and branding. You don’t expect brand-building advertising to pay for itself immediately.”

Amid internal debate about how to design a rebate-based loyalty program and whether it would pay off, Drugstore launched an experiment for Holiday 2001 that evolved into the third component of its program, Drugstore Dollars. To tackle one of its biggest ongoing challenges—that of turning gift-buying holiday shoppers into repeat customers in the new year—the company tested giving registered shoppers a 5% credit on non-prescription purchases to spend on future purchases. Registered shoppers don’t need to opt into the program as Drugstore tracks each customer’s purchases and presents the rebate status on qualifying purchases to shoppers at checkout on return trips. The goal is to improve the frequency of purchase among repeat customers as well as boost the likelihood of retaining them over time.

Redemption windows

The key to accomplishing that—and making the rebate a cost-effective approach—depends in large measure on the windows set for rebate accrual and redemption, Drugstore found. In the initial test, customers accrued points for purchases during December, redeemable throughout the first quarter. Later, to promote multiple purchases during the accrual period and to give customers the opportunity to build a meaningful amount of store credit, Drugstore lengthened the accrual period from one month to three.

Then Holtan noticed that most of the lift in the three-month redemption period was in the first month, with the flow of redemption-driven traffic settling down to normal traffic levels by month two. Drugstore then set the time frame Drugstore Dollars uses currently: a three-month accrual period followed by a one-month redemption. “What we hope is that we get the same lift in the redemptions that we saw in the three-month redemption period, but we’re giving away a lot less money,” Holtan says.

Though the impact of free shipping is most clearly visible in increased average order size, Drugstore is still grappling with separating out the effects of all of its loyalty program elements and tracking the dollar value of each. “We launched them all at the same time, and they are all available to the same group of customers, so it’s hard to determine if one is making a difference, or if it’s all three,” Holtan says.

Yet whether individually or in combination, the programs have helped move the needle on some key global metrics at Drugstore, according to Holtan. As spending on retention has increased, customer acquisition costs, for instance, have plummeted from an average $42 in the first quarter of 2002, before the programs were established to $16 in the first quarter of this year. Average revenue per order, a figure that blends sales from both front-of-store merchandise and prescription drugs, rose from $67 to $71.

Constant tuning

Holtan says Drugstore brainstorms constantly on loyalty programs, though any testing between now and the holidays will likely center on adjustments to the current three rather than new ones. One possibility under discussion could increase the benefits of loyalty to the most loyal customers in the form of a preferred customer program. Drugstore already has tested a program in which registered customers who spent $150 on merchandise received 5% off merchandise purchases for the rest of the year.

Response fell into three segments defined by customer spending patterns. The offer had little effect on the best customers, who continued to spend at about the same level as before they reached the threshold, and no effect on customers who spent at the lowest level. However, the program did produce lift among customers whose spending fell into the middle level. “So is it better to reward the customer who already contributes the most to the bottom line, or to reward the average or less than average customer to try to get more out of them?” Holtan asks. “We are going to try to do both by being smart about how much we are willing to invest in different types of customers.”

mary@verticalwebmedia.com