E-Commerce ROI: Many Happy Returns

By Jeffrey Rothfeder  |  Posted 05-07-2002

E-Commerce ROI: Many Happy Returns

In January, when Amazon.com reported its first profitable quarter ever—keeping its promise to Wall Street that at the end of 2001, the online retailer would finally go into the black after seven years—the business press covered the story as if it were the second coming of the Web.

But 150 miles to the south of Seattle, where Amazon was celebrating its "turning point," in the words of CEO Jeff Bezos, executives of another Internet book retailer had a very different reaction.

"What's the big deal? We've always made money since we've been online—and that's eight years ago," says Miriam Sontz, CEO of strategic development at Portland, Ore.-based Powell's Books, whose eight branches make it one of the Northwest's largest book chains.

Without the luxury of a nickel in venture capital backing or public financing, privately held Powell's Books simply couldn't afford to lose money for an extended period of time. "We had to start off with a different mentality than Amazon and play business the old-fashioned way," says Sontz. Since its founding, Amazon.com has received upwards of $2.1 billion in funding to take itself to profitability. Meanwhile, Powell's initial investment to design its Web site was kept to well under $1 million, because virtually all of the software and creative development was handled in-house, and the company earmarked new money for online operations only when Internet cash flow could cover it.

Moreover, Powells.com didn't build its own warehouse for more than three years after the Internet site first opened its doors. Until then, the bookstores served as the site's distribution centers. But when the warehouse was completed in April 1998, the cost of Powells.com inventory management dropped precipitously, because fewer employees were needed to manage the stock. That widened Internet profitability, and generated additional cash to increase the number of books in inventory and upgrade customer checkout procedures, among other things.

Powell

's Tactical Choices">

Powell's Tactical Choices

The rigorous eye Powells.com keeps on earnings stretched to tactical choices the company made as well. Early on, it decided to sell new books at full price and to concede bargain hunters to Amazon. That's because a discounted book is simply not a profitable sale, Sontz insists. Instead, Powell's staked out a claim to the higher-margin used-book business, which accounts for 75 percent of its online revenue.

Because used-book sales are moneymaking transactions, Powell's can offer free shipping on any purchase over $50 and still not lose money. The result: While many of its customers come to the site to buy a few less-expensive used books, after they surpass the no-cost shipping threshold, they're likely to purchase new books as well. "We aligned the free-shipping strategy with the profit returns we could expect from each used-book sale," Sontz explains.

Consequently, Powell's Web site has never drained cash from the company even as it significantly expanded the bookstore's customer base well beyond the Northwest, bringing in millions of dollars in new revenue from other parts of the country that Powell's otherwise wouldn't get. Powells.com accounted for nearly 30 percent of the book chain's sales last year of more than $80 million, up from just 10 percent in 1999.

The success of sites like Powells.com turns the already somewhat hackneyed idea of "path to profitability" on its head. While pure-play e-commerce companies have struggled in vain to eke out earnings using the traditional measure of revenue minus expenses—hoping that, eventually, sales volume would simply overwhelm costs—a variety of traditional retailers are engaged in an innovative effort to rethink the concept of Web earnings and what the Net delivers to overall operations.

What makes this re-examination significant is that it's producing new and more precise methods for tracking profitability and return on investment from online operations as well as fresh strategies for leveraging operations across the retailer's multiple channels.

"Retailers have figured out that there are multiple advantages and beneficial ripple effects from having a Web site, and most of them have nothing to do with the initial fantasy e-commerce business model—that somehow if you just sell tons of stuff online, the earnings will automatically roll in," says Kate Delhagen, director of retail research at Forrester Research.

eROI thumbnail
How Companies Reap Returns: A variety of bricks-and-clicks companies are turning to more sophisticated metrics, which we call e-ROI, to determine the true gains they are receiving from their e-commerce initiatives. Click to enlarge.

"To measure the real value of the Web, companies have to examine the cumulative positive effect of the Internet on the total retail operation, and then look at this effect in real dollars and cents, not just assumed value."

At the core of this approach—let's call it e-ROI—are new ways to measure profitability that quantify the Internet's impact on some of the most important metrics of a retailer's performance, rather than isolating Web gains purely to what occurs on the Internet.

Retailers are now measuring how their Web presence affects inventory management, generates revenue in multiple channels, helps develop a much wider and more valuable customer database, and cuts costs across the corporation. By viewing these more holistic gains against the expenses required to run the Web site, a truer picture than ever before of Internet profitability can be drawn, and an actual value placed on the online operation.

e

-ROI: Keeping it Simple">

e-ROI: Keeping it Simple

Powell's tactic is the simplest: Don't spend more than you have, and extend your existing systems—computers, logistics, fulfillment and anything else that can be leveraged—to your Internet operations until the Web site needs and can afford its own. Many smaller retailers, including Godiva Chocolatier Inc. and food company Hickory Farms Inc., have adopted this model.

Other retailers are viewing these new e-ROI metrics through much more finely focused lenses, discovering ways to get value out of complementary operations they wouldn't have imagined before.

Wal-Mart Stores Inc.'s Web site turned a profit primarily on the strength of selling higher-priced goods such as DVD players, digital cameras and computers. Knowing this, Wal-Mart can potentially improve companywide margins by carrying fewer of these items in its stores, which tend not to sell particularly well to the typical Wal-Mart non-Web shopper but take up lots of precious shelf space.

Meanwhile, Sears, Roebuck and Co.'s Web site is paying unexpected dividends by being the first stop for more than 10 percent of its major appliance customers, who like to investigate and compare items online before buying them in a Sears store, and those shoppers represent $500 million in appliance sales per year. So Sears has expanded the product information available on its Web site.

Perhaps most surprising, apparel retailer J. Crew Group Inc. recently reported that during the months of January and February its Internet site generated more revenue than did its 20-year-old catalog business. J. Crew's coup is a financially significant landmark that is helping the retailer in two ways: According to the company, customers spend more on each Web order than they do on catalog purchases, so as more shoppers use the Web site, J. Crew's overall prospects improve. And thanks to the increase in Web sales, J. Crew has been able to save money by cutting down on the number of catalogs it prints and mails.

"Retailers' main task is to convert lookers into buyers in the least expensive way they can," says Kara Heinrichs, chief customer experience officer at Fry Multimedia Inc., an Ann Arbor, Mich., company that has developed Web sites for 1-800-Flowers.com Inc., Eddie Bauer Inc. and Domino's Inc.

"For bricks-and-clicks, profitability has to be viewed through systemwide metrics," adds Heinrichs. "Product research online for more expensive items can convert to valuable purchases offline, and low-margin everyday products can produce higher margins if sales are shifted from the stores to the Web sites. How well these types of activities are leveraged is the real determinant of success."

e

-ROI Metrics: Uncharted Territory">

e-ROI Metrics: Uncharted Territory

The use of e-ROI metrics is still a new phenomenon. Even retailers that are trying to adopt them admit that they're feeling their way through uncharted territory. Just a few years ago, companies trying to make a go of e-commerce were advised to build completely separate Web operations, unencumbered by the sclerotic strategies of their real-world managements, if they wanted to compete against fiercely nimble start-up Internet rivals in a venue where capturing customers required the dexterity of a juggler keeping a dozen balls in the air.

So Wal-Mart, Toys "R" Us Inc., Kmart Corp. and numerous other big-name retailers created their first sites as distinct entities, building high walls between their bricks and their clicks.

But the strategy was based on a seriously flawed business premise. It cut the Web sites off from critical lifelines—inventory controls, warehouse systems, customer service centers and relationships with suppliers, to name just a few—that the traditional retailers had spent years developing. E-ROI metrics, by contrast, can coordinate all aspects of a retail operation for the greatest gain, rather than separating them for the least.

To use e-ROI models, most companies start by calculating how much is being spent to develop and run a Web site. On the face of it, that's simple enough: They just add up the capital costs the way they always did. But that analysis raises a critical question: How much should companies spend on Internet operations?

Once a retailer has its Internet capital costs determined and set at an appropriate level, returns generated by the Web site that offset expenditures can be measured. That, of course, begins with revenues from online purchases themselves.

Often, a Web site can deliver 20 percent or more in incremental revenue over and above traditional retail store and catalog sales, especially with marketplace or high-end sites. But for companies selling products that consumers typically prefer to research before purchasing, the impact of the Web site on transforming browsing customers into actual shoppers extends well beyond the dollar value of the online sales to include purchases made in physical stores.

For every dollar spent online, for example, the typical auto parts retailer can expect $136 in store sales initiated by a visit to the Web site, according to Jupiter Media Metrix. The ratio for home improvement retailers is even higher: $163 to $1. Such cross-channel sales—a large part of e-ROI analyses—are critical for big outlet retailers that depend on high and continually increasing volume for profitability.

At electronics retailer Best Buy Co. Inc., 40 percent of its customers log onto its Web site before going to one of its stores to purchase major products like expensive home or car audio systems.

"That's a pretty astounding number," says Scott Bauhofer, Best Buy's general manager for online stores. "We have $18 billion in annual revenue, and if the Web site is at least largely responsible for $7 billion of that—being conservative, you can cut that down to $4 billion or so—the ROI is obvious. We're not spending $4 billion a year to manage our Internet operations."

Improved payroll productivity is another hidden benefit on the revenue side that shows up in e-ROI analyses. For many sales-driven operations, payroll can consume more than 10 percent of revenue. And while some companies treat these expenses as easy areas for cost-cutting, customer service usually suffers when that happens.

But a well-developed Web site that provides detailed information about products can significantly limit the amount of time spent on handholding when customers make in-person and telephone purchases—and eliminate it altogether when goods are bought on the Web. In many cases, this can save 15 percent or more in payroll expenses.

Toshiba America Business Solution Inc.'s online office machine sales site for dealers, called FYI, illustrates this well. When it was launched in 1997, the cost of information-intensive telephone orders were eating into sales just as Toshiba was being forced to drop equipment prices amid hardened competition.

But over the past five years, FYI has converted fully 96 percent of Toshiba's product dealers into Web customers. Now, more than 80 percent of the company's business comes through the Net. Calls to customer service centers have remained flat during the period, while revenue has grown 20 percent a year.

The upshot: Without FYI, it's estimated that Toshiba would have had to hire at least another dozen employees to handle sales. As a result, Toshiba has cut payroll expenses as much as 50 percent through its Web presence, according to a PricewaterhouseCoopers study.

Improved inventory management would seem to be a natural for e-ROI cross-channel benefits, but most companies are not taking full advantage of it. Too many retailers treat the Internet like a discount bin, selling products at deep markdowns that aren't popular anymore in stores.

That works to a point: Items tend to move out of inventory more quickly on the Web than if they're offered in real-world outlets at clearance prices. But the cross-channel and companywide gains are inconsequential, says David Fry, CEO of Fry Multimedia. "Discounting doesn't drive customers to stores, and bargain hunters are not loyal customers," explains Fry.

Keeping Sales Moving

Keeping Sales Moving

A far more valuable inventory management tactic, one that can reflect a tangible return on investment from the Web site when explored with e-ROI analysis, is illustrated by a Toys "R" Us campaign to promote Microsoft Corp.'s high-buzz Xbox game machine. Last August, Toys "R" Us let customers pre-order the device on its Web site, bundled with software and accessories, for $499. That was $200 more than the Xbox alone would cost in Toys "R" Us stores after the machine's official release in November.

The Web site promotion was a remarkable success: The company's entire first Xbox allotment—tens of thousands of machines—was snatched up in 30 minutes, even though no machines would be shipped for four months.

This was an especially lucrative use of the Web for inventory management. Before Toysrus.com, the retailer would have had to guess at how many game machines it needed to put on its store shelves, losing money if consumers didn't purchase the hardware quickly enough, and losing sales if the item was so popular that it sold out before the shelves could be replenished. By contrast, the Web pre-order strategy gave Toys "R" Us the luxury of not having to purchase that portion of its inventory blindly, without knowing how many it would sell.

And it also delivered more profit than store sales because Web customers were also required to buy the higher-margin peripherals. Toys "R" Us declined to discuss how much additional revenue grew out of the Web Xbox campaign, but analysts estimate Toys "R" Us generated well over $10 million in less than an hour.

The Web benefit that's the most difficult to measure or quantify—but that may turn out to be the most valuable for the long-term growth of a retailer's business—is what can be learned about customer behavior and preferences through the Internet's far-reaching data collection capabilities—and how this information can be used to increase sales across all of the company's channels.

Almost every e-tailer with a marketplace or high-end Net presence uses marketing analysis software to keep track of how customers navigate their Web sites—essentially, shadowing them click-by-click to keep a record of what they look at, what they purchase and how they respond to different features on the site.

Having access to this data can lead to invaluable sales tactics such as those implemented by Wal-Mart.com and Landsend.com, among others, in which products abandoned from online shopping carts are stored on virtual shelves and then remarketed to these customers when the item goes on sale, thus potentially reclaiming a lost sale.

"It's called next-purchase intent, and we could only know this by keeping a detailed log of every move the shopper makes on the Web site," says Scott Kauffman, president and CEO of Burlingame, Calif.-based Coremetrics Inc., whose clients for its online behavior software include Wal-Mart, Ann Taylor Stores Corp., Nortel Networks Corp. and the Sundance Catalog Co. "Such information is available in no other channel for a retailer. In some cases, we've found that customer analysis programs are responsible for double-digit revenue increases that would have gone untapped."

A few e-commerce companies are blending this information with data that confirms customer activities in other channels, asking visitors for their e-mail addresses when they shop in stores or purchase through a catalog, or offering discount customer loyalty programs based on cumulative levels of spending in either bricks or clicks outlets.

Then they combine offline and online purchasing activity for each of their customers into one database. Sharper Image Corp. has been particularly aggressive about amassing a database of this kind, compiling records of cross-channel buying behavior for upwards of 80 percent of its customers. This information has paid off handsomely: It's partly the intelligence behind Sharperimage.com's so-called dynamic browsing feature, which continually displays new products for people to consider based on prior store and Internet purchases, and it is the basis for e-mail and in-store marketing campaigns targeted at database-driven customer preferences.

The return from this extensive use of information, most of it gleaned from Web activities, is demonstrated in The Sharper Image's results for the third quarter of 2001, ended in October. In a difficult sales environment, when the retailer's overall revenue dropped 24 percent, the percentage of higher-margin, private-label merchandise rose to 73 percent of revenue in the nine months ended in October 2001, thanks in part to Web site customer data collection.

There are any number of other factors to consider, often peculiar to the operations of individual retailers, when calculating e-ROI from a Web site.

These include quantifiable cost savings from closing stores and using the Internet as a much less expensive alternative sales outlet, which Apex Stores, a 78-year-old New England discount retailer, did recently. And then there's the flip side: earnings gains from opening up real-world stores in areas where Web site data shows potential customers.

Retailing experts caution, however, that no amount of financial analysis and forecasts gleaned from statistical consumer activity can replace the need to deliver a shopping experience on the Web—and in stores—that attracts customers. Return on investment for retailers, no matter what the venue, is based solely on the way people feel when walking into a store, reading the catalog or browsing the Web site—and then whether making the purchase is a comfortable encounter.

"On the Web, it's called usability testing—measuring how easy it is to navigate the site, put items in the cart and go through the checkout process," says Angela Sinickas, president of Sinickas Communications Inc. in Costa Mesa, Calif., which customizes performance measurements for companies' online and real-world operations. "Retailers go through the same analysis in their stores all the time. And knowing or not knowing what customers want in a retail experience is why Wal-Mart succeeds and Kmart failed. There's no return on investment from a Web site for a retailer that doesn't understand that lesson."

Jeffrey Rothfeder writes frequently about business, security, environmental and technology issues.

How Much Should It

Cost?">

Getting and Spending: How Much Should Your Web Site Cost?

Simple marketing site
Includes static or easy-to-post content such as a store locator, sales announcements, product information and customer service. These sites, which can cost anywhere from $100,000 to $1 million up front, and about the same each year to maintain, are generally suited for retailers with customers who tend to reorder the same items on a regular basis.
Example: duanereade.com

Other retailers who don't want to spend a lot on fulfillment—especially those without existing catalog operations and logistics systems—can inexpensively produce a site like this to promote products and discounts, and direct customers to nearby stores.
Example: ames.com

Marketplace site
Features a wide array of product images and descriptions, sophisticated ordering capability and targeted or filtered promotions based on prior purchases. The cost is about $10 million to $20 million a year—although some companies are cutting corners to trim this expense by as much as two thirds. Such Web operations are sufficient for retailers selling low-margin commodity products that consumers are very familiar with, such as books, CDs and sporting goods.
Examples: towerrecords.com, thesportsauthority.com

Multi-channel site
Offers such features as interactive content that details and compares all aspects of a wide variety of products; programs that track customer behavior and link that data to purchasing activity in stores and by catalog; and customer incentives and rewards for buying items through any of the retailer's sales channels. These pricey Web sites—in some cases $40 million or more annually—are used mostly by retailers selling items that require significant up-front research, such as TV sets and digital cameras.
Examples: sharperimage.com, walmart.com, bestbuy.com