Byline: Vicki M. Young

NEW YORK — The Internet’s allure as a retail panacea is rapidly yielding to reality.
A report on Internet infrastructure issued by PaineWebber asserts that the upkeep of commercial Web sites is even more expensive than the costs of launching them. PaineWebber released the report, entitled “The Web at What Cost?”, based on data compiled with Forrester Research, the Internet research specialists who last week rocked the Internet world when it predicted the demise of more than half the dot-com retailers now online by next year (April 14, page 2).
While less dramatic, the report may be as sobering as Forrester’s earlier dot-com forecast. It notes that launching a pure-play Internet commerce site can cost as little as $1.8 million, with annual upkeep costs of $1.9 million, or as much as $40.5 million for a state-of-the-art site, which would cost $48.8 million a year to maintain.
Most traditional retailers entering the click-and-mortar arena spend somewhere in between. But with costs as high as they are, the average dollar amount per order needs to be quite high before doing business on the Web can approach the break-even point. Post-launch costs tend to reach the stratosphere because there is no such thing as a onetime Internet cost — the funds needed to stay fresh and retain customers can be equal to or greater than the costs of setting up or acquiring customers.
Forrester formulated three pure-play site models in analyzing potential cost structures: a basic transaction site, a midtier “share grabber” site and a leading edge anticipatory seller site. PaineWebber converted those dollar estimates into costs per share for a wide range of retailers, finding that the earnings hit can be as high as 59 cents per share for some.
The investment firm concluded that size matters when defraying upfront costs. Wal-Mart is so large that an annual $100 million loss wouldn’t affect earnings by more than 1 cent per share. Of course, the opposite is true as well — even a $100 million profit from would boost earnings just 1 cent per share.
PaineWebber assumed traditional retailers could save on some of the expenditures, up to 30 percent, if they took full advantage of existing purchasing power and brand name recognition. The investment firm also built three hypothetical models to provide an estimate of profitability. But the hypothetical models indicated that even firms with an average order size of $100 or more couldn’t be guaranteed profits.
And if overall Net costs can be staggering now, chances are they’re going higher.
“As the Web continues to proliferate,” the report said, “people/talent will become more scarce and more expensive. The same goes for marketing costs as each operator tries to be heard among the growing masses. Finally, customer expectations and resulting service levels will also rise… All of that will keep a lid on profitability until the e-commerce space consolidates dramatically.”
Forrester’s anticipatory seller model, at a launch cost of $40 million, is top of the line. With up to 10,000 transactions daily, this model allows for nonstop e-commerce. Site content — the graphics, text and entertainment component — costs $10.9 million, or 27 percent of the launch total, but allows multiple-language text.
Selling strategies such as collaborative filtering and other data mining follow at $10.9 million, or 25 percent. Core infrastructure costs for hardware and telecommunications connections and staffing needs hit $6.9 million, or 17 percent. Marketing expenditures at $3.6 million, or 9 percent, are sufficient for a major portal deal.
Post-launch, the expense eaters are presale customer service at $20.8 million, or 43 percent, and marketing at $18.9 million, or 39 percent of the estimated $48 million in annual maintenance. The post-launch marketing costs allow multiple portal deals.
Assuming 25 percent gross margin, a $50 average order results in $182 million in annual sales, with $45.6 million gross profit. While that appears impressive, PaineWebber noted that the $48.8 million in annual Web costs would leave the site operator in the red before even adding fulfillment costs of $20.1 million. But a $100 average order can generate earnings before interest and taxes of $16.9 million and a margin of 4.6 percent.
According to PaineWebber, several retailers use this approach. J.C. Penney is one, but it also has one of the highest average tickets, approaching $150. The Web enables Penney’s to leverage its significant 52 million customer base across its three channels — retail, catalog and the Internet.
Penney’s is embarking on aggressive sales targets, about $260 million in 2000, $500 million 2001 and $1 billion by 2002. Yet those targets pale in comparison with Penney’s $4 billion catalog business and $15 billion in sales at retail. The company is in the range, PaineWebber said, where it can take advantage of its “poor standing with the Street by taking risks” from which others shy away.
Other retailers using the cutting-edge approach are Federated Department Stores and Nordstrom. Federated’s model is spending between $75 million to $100 million in investment this year, plus a similar amount in capital expenditure. In comparison, Federated in 1999 spent $822 million in capital expenditures corporate-wide. The overall Internet business is projected to hit $200 million in sales, and the Fingerhut operation will be about half of that amount. Yet PaineWebber expects Federated’s online business to lose around $100 million in 2000 due to investment spending.
Nordstrom’s combined direct-to-consumer business grew 8 percent in 1999 to $210 million. Most of that gain, or $20 million, was generated by online revenues. Web site enhancement costs are expected to increase modestly above 1999’s $23 million figure.
The midtier model, dubbed “share grabber” by Forrester, can accommodate 1,000 transactions daily. This mechanism is built for growth, with mass appeal. It’s the model that describes many traditional retail players — the big three discounters, Sears, Victoria’s Secret, Pacific Sunwear, American Eagle Outfitters and Talbots.
Launching a share-grabber site costs $10.6 million. Site content represents 30 percent, or $3.1 million, of total pre-launch costs. Core infrastructure costs hit $2.8 million, or 26 percent. Marketing expenditures total $2.1 million, or 20 percent, and are enough to allow for a major portal deal. Post launch, however, marketing costs explode to 62 percent, or $8.9 million of the estimated $14.3 million in upkeep costs. Presale customer service, such as call center costs, follows at 14 percent, or $2 million, while basic infrastructure requires just 10 percent, or $1.5 million, of total maintenance costs.
Using the same 25 percent gross margin figure, PaineWebber estimates this midtier model would need at least an average order size of $200 just to break-even — not good news for the click-and-mortar crowd that use this model. A $50 average order projection at 1,000 transactions a day would generate $18.3 million in annual sales and $4.6 million in gross profit. But subtracting the $14.3 million in annual upkeep costs and $2 million in fulfillment expenditures leaves the site operator $11.7 million in the hole.
PaineWebber said in the report that even though Wal-Mart’s goal is to be dominant and a “trendsetter” in everything that it does, the firm’s “direction is still uncertain. Given the low average order size in its base business and a high number of sku’s per transaction, we are not convinced that a basic distribution model would serve Wal-Mart well.”
Talbots, with its catalog fulfillment infrastructure already in place, spent only $1.3 million to launch its site, which PaineWebber said was likely to be “additive to earnings from day one.” The site has been successful in attracting new customers. About 25 percent of the customers purchasing on the Web do note have Talbots credit card and are not in the company’s catalog database.
Victoria’s Secret, PaineWebber said, has a sophisticated distribution system that supports the 400 million catalogs mailed annually by the company. Not all of its online customers patronize its brick-and-mortar locations, suggesting that the Web site is not cannibalizing its retail sales.
PaineWebber said management disclosed that “half of its Web customers are new to Victoria’s Secret and that twice as many men shop online at the Web as do in the catalog or store.”
The in-store purchase ratio is 90 percent women to 10 percent men, while the online ratio is 40 percent women to 60 percent men. The average online transaction is $95, higher than the $50 spent at the stores, but currently below the $110 on catalog purchases. One negative feature of the Web is its lack of integration with the stores, since purchases online cannot be returned or exchanged at retail.
In contrast, PaineWebber said both Gap and Pacific Sunwear were likely to take longer to achieve profitability because neither had a catalog infrastructure in place. could reach the break-even point this year and Gap could turn profitable in another year or so.
The numbers paint a fairly dire picture for the most basic of site structures. According to Forrester, the transaction site is equipped to handle 250 orders per day, with the goal of selling goods in reaction to customer demand. The finite items on the site are a convenience for its current customer base.
A basic transaction site, at $1.8 million in prelaunch costs, has the cheapest startup of the three models.
Expenditures for hardware, telecommunications connections, personnel and software eat up 47 percent, or $841,000, of the cost structure. Marketing is a mere $305,000, not enough to score big portal deals, the report said. However, marketing costs jump to 53 percent, or $1 million, of the estimated $1.9 million needed annually for site upkeep. Hardware and other core commerce costs drop to $528,000, or 27 percent, in post-launch costs.
According to the PaineWebber profitability model, it’s only when a transaction site has an average order size of $150, with a gross margin of 25 percent, that the firm posts EBIT. Most e-tail pure plays using this transaction model, however, average only $50 per order, the report said. At that rate, companies never recoup their original startup investment.
Jewelry firms such Tiffany & Co. and Zale’s use the transaction site format. Tiffany, PaineWebber said, can use the Web to enhance its bridal registry business, while a small business-to-business opportunity exists with the firm’s corporate gift program. At Zale’s, the company’s nearly 1,350 store sites provide easy return, repair and support.