The SEC's Numbers Game

How well are you doing online? That, in a nutshell, is what the U.S. Securities and Exchange Commission wants retailers to tell investors. Since last spring, the SEC has been sending letters to retailers who have told investors that their e-commerce is growing phenomenally, that online sales growth is twice industry average—you've heard their lines.

But when the SEC pushed Target, Walmart, PetSmart and other retailers to get more specific, they resisted, arguing that the online sales figures were "not material"—or in plain English, "too small to matter."

This is a bad idea, this focus on e-commerce by the numbers. It's bad for retailers, bad for investors, bad for the SEC.

Consider three retailers whose numbers were reported by the Wall Street Journal, which wrote about the SEC's efforts this week. In the case of Walmart, about 2 or 3 percent of its sales are somehow attributable to online. That compares with roughly 11 percent for Nordstrom—except Nordstrom's number includes its substantial catalog operation, which gets thrown in because the category is really "not brick-and-mortar" sales. And Staples does more than 30 percent of its business by way of the Internet.

What do those numbers mean? Lots of things, but they're all specific to the retailers involved. They definitely don't mean that Staples is three times as e-commerce-savvy as Nordstrom and 10 times more advanced than Walmart. Maybe Staples is, but these numbers won't tell investors that.

Why not? Because for years, office managers have been faxing office-supply orders to Staples and its competitors. Now they just use an online form. That's technically e-commerce, and it's a big chunk of Staples' business, but it's not something that investors can use to compare Staples with Nordstrom, which is selling apparel and personal service to an affluent crowd, or Walmart, which is selling fill-your-cart-at-low-prices-right-now.

What's worse, of course, is that e-commerce as a percentage of total sales—this number the SEC seems to favor—is the murkiest of metrics. The easiest thing to count is where the customer pays—online or in-store. But how do you count ship-to-store? Or an "endless aisle" order at an in-store kiosk? Or a mobile-commerce order from an in-store customer who doesn't want to wait to checkout? Or an eBay Now order that's taken online but then purchased from a local retailer for quick delivery?

Or—my favorite—a customer who arrives to buy in-store, but ends up being walked through an online ordering process because that's what corporate is instructing associates to do, after which the customer picks up that item before leaving the store?

Should retailers count all those as e-commerce, Mr. SEC? Let's face it, the SEC wants a single, simple number that investors can use to compare retailers. But if those numbers aren't simple, they're useless. And omnichannel retail—which is really what we're talking about, not "e-commerce"—is anything but simple.

That's why it's bad for the SEC along with investors and retailers. It's junk data. It's useless for investors and easy for retailers to manipulate. It's a disservice to everyone involved.

So here's a modest proposal for the SEC: Forget about a simple number. Tell retailers that they can either release voluminous data about their omnichannel sales in all its endless detail, or they can say "we don't report that." No middle ground. No "three times industry average" or "growing spectacularly." Either open the kimono completely or shut up.

Every retailer would shut up. Analysts would make their educated guesses, but they'd be the ones accused of misleading investors, not retailers. There would be one less meaningless data point for us and our colleagues to report.

And instead of the online numbers, we could all focus on the bottom-line numbers—the ones that matter.