An E Commerce Cautionary Tale

As Williams-Sonoma's site faltered, inventory piled up

For much of last year, a band of young Net-heads toiled away at Williams-Sonoma Inc.'s San Francisco headquarters, cooking up a glossy e-commerce site that launched Nov. 1. Figuring the site would generate a surge of holiday orders and help promote sales at its core catalog and retail units, the company bulked up inventory. "We didn't want to disappoint anyone by being out of stock," explains Chief Financial Officer John W. Tate.

It turns out that Williams-Sonoma disappointed only itself and investors. On Mar. 6, the marketer of upscale kitchenware and home furnishings announced that fourth-quarter profits would fall about 13% below analysts' estimates, largely because it stashed too much in the holiday cupboard. While the merchant chalked up $8 million in online sales--in addition to $530 million in catalog and retail sales during the quarter--it surely expected better. The news sent the company's stock plunging 37%, to $19.50, that day.

HIGH HOPES. There's a lesson here for brick-and-mortar retailers rushing to the Web. Williams-Sonoma's troubles challenge the notion that adding an e-commerce site to an established retail and catalog business will quickly lead to a sales boom. "It's not just a matter of `if you build it, they will come,"' says James W. Vogtle, director of e-commerce research for Boston Consulting Group.

For even the savviest direct marketer, consumer behavior on the Web remains uncharted territory. The ability to mine and track customer data--one of Williams-Sonoma's greatest strengths in boosting catalog sales--doesn't necessarily translate into a quick payoff online. With e-commerce in its infancy, "accurately forecasting demand is almost impossible," says Vogtle. Indeed, even Web vet Inc. took a hefty fourth-quarter inventory charge.

For its part, Williams-Sonoma was determined not to repeat past inventory shortfalls that have hurt its real-world business. And with a new Web site and widespread expectations for a booming e-Christmas, the company was anxious not to disappoint. As a result, "we overdid it," says Tate. It didn't help that the company had little time to promote the site before the Christmas rush.

Of course, Williams-Sonoma wasn't shooting completely in the dark. A bridal-registry site launched last summer proved that Internet presence could drive traffic to Williams-Sonoma's 350 retail stores and its powerhouse catalog operation: 10,000 catalog requests a week came through The site brought in new customers, too. Half of online sales were to people not already in the retailer's 19 million-name database.

TOASTERS, ANYONE? In the end, though, holiday Web sales amounted to just 1% of fourth-quarter revenues. Catalog sales rose 30%, but sales at stores open at least a year rose a modest 4.4%, leaving the company with loads of leftover toasters and countertop mixers. Post-holiday inventory was up about 40% above the prior year's level. As a result, the company said net income for the fourth quarter would rise about 9.3%, to $48 million, vs. the $55 million expected by investors.

To be sure, only half the earnings shortfall came from holiday-related inventory, labor, and warehouse costs. The other half reflected a noncash charge to account for projected merchandise returns, about 10% of sales. Unlike most retailers, the company did not maintain a reserve in the past, but says it was advised to set one up by its auditors.

Still, there's no denying the Web stumbles. Williams-Sonoma's woes are another reminder that glitzy sites alone don't guarantee e-commerce success. And it proves that adapting retail disciplines to the Web isn't a cakewalk.

Before it's here, it's on the Bloomberg Terminal.