(FORTUNE Magazine) – It's hard to say just where Starbucks went wrong on June 30. After the end of trading, it announced that it would be sinking millions into a goony-sounding Internet-based retailing venture that would sell everything from gourmet foods to furniture. That same day it began acting like an Internet firm, trimming its earnings expectations to 54 cents per share, down from a consensus estimate of 60. By the laws that have governed Wall Street lately, the company's stock should have doubled; had it managed to eliminate its earnings entirely, it might have tripled.

But no. By the end of the following trading day, SBUX shares were going for 26 15/16, down from a pre-announcement 37 9/16. Much of this drop can be attributed to the reduced earnings outlook: Shareholders don't like surprises, and while the company's comparable-store sales are up by a solid 5% for its fiscal year to date, the reduction hinted that the demand for non-fat-triple-foam-vanilla-cilantro lattes may finally be slowing. But if you had listened to the conference call Starbucks CEO Howard Schultz conducted with analysts that week, you would have heard something rare: rumblings of skepticism about a company's online strategy.

Schultz's pitch to the analysts was sketchy; he was holding back on more information, he said, until later this summer, when Starbucks plans to announce "a major strategic partnership," rumored to be an acquisition of housewares cult Williams-Sonoma. (For the same reason, Schultz is currently declining interview requests.) But the plan, as he laid it out, rested on a series of safe propositions, adding up to a somewhat less safe conclusion. Starbucks is a premier consumer brand. (So far, so good.) There is a $100 billion market for gourmet foods, kitchen products, and home furnishings. (I haven't totaled the numbers myself, but I'll take his word for it.) Those product categories appeal to women ages 25 to 49, who are also Starbucks' core customers. (Again, if he says so, I'll believe him.) There is no dominant market leader in these categories. (True enough.) So Starbucks should sell these items online. (I'm sorry. That last part. Could you run that by me again?)

One analyst put it best when he asked whether anyone actually wants a Starbucks couch. Schultz quickly replied, "There won't be a Starbucks-branded couch in any way." Whew! What there will be is a "canopy site"--a website with its own separate brand identity, which will link to a portfolio of other web pages, one of which will sell Starbucks coffee, another of which will sell home furnishings, and so on.

The apprehension about Schultz's plan isn't so much about the idea itself--if people are willing to buy books online, why not a spatula or an ottoman?--as it is about the fact that there's no compelling reason for Starbucks to be in this arena. Unlike, say, Starbucks' coffee ice cream, it's hard to draw a straight line from the coffee bars to online armchairs. So, what's Schultz thinking? One clue may have come from his acknowledgement that "it is clear to all of us that our core retail growth rate is slowing." And if you're a company looking for growth, and you're located in Seattle, and your CEO has reaped the rewards of investments in drugstore.com and eBay--well, then maybe the Internet seems like the next logical step. Schultz's description of his plan for the chain is rife with Internet logic; indeed, it's based upon the Holy Grail of online commerce, customers' eyeballs.

As Schultz sees it, the reason Starbucks' Internet adventure makes sense is that the chain has millions of loyal customers who can be herded toward the website. Or, in his words, "With [Starbucks'] physical assets, we're going, with great precision and discipline, to drive the physical community that exists in our stores to the web." Rather than spend much of its budget on advertising, as most online startups do, the new site will be advertised within Starbucks stores--meaning that Starbucks' doors would actually be a very low-tech Web portal. A portal portal, if you will.

In short, Schultz is gambling millions on the belief that his customers will go where he tells them. That may not be sufficient to guarantee his site success, but then it shouldn't be entirely discounted, either. After all, it's never wise to doubt the mass-thought-control capabilities of a guy who's persuaded a great, God-fearing nation that $3 is a completely reasonable price for a cup of coffee.

--Tim Carvell