Evan Schuman's StorefrontBacktalk

That’s nice safe thinking. It may also be dangerously wrong-headed in several ways.

One wrong-headed thought is the notion that Amazon’s quick-delivery program hasn’t been successful. We don’t know if it has, but there’s a very obvious reason Amazon might be taking its time with expansion: Unlike brick-and-mortar retailers, Amazon has to build a distribution center wherever it wants to deliver. That costs money and also exposes Amazon to sales-tax requirements.

There’s only so fast that Amazon can build DCs, and only so fast that it can cut deals for tax relief in exchange for building those DCs. (That could explain why Amazon isn’t doing same-day in California—it just did its tax deal there last year.) Meanwhile, Amazon has more same-day delivery experience than all other retailers combined—except for pizza places, of course.

Another troublesome misconception is the idea that testing same-day should require a big investment. It shouldn’t. In fact, it’s doubtful how much useful information eBay’s same-day tests in SF and New York will generate, because the cost of the trial—hiring its own couriers, paying full price for the products it delivers, offering big discounts for the first few deliveries—will swamp any chance at profitability. The auction giant even admits that its trial is a loss-leader.

However, a brick-and-mortar retailer doesn’t have to do those things. In fact, chains really shouldn’t. Courier services already exist. Same-day delivery can be an option that only appears on an E-Commerce site depending on delivery Zip code. The hard part is working out the logistics. The prize is learning how much customers are willing to pay for quick delivery and, more important, how many customers will shell out extra for the service.

But the biggest bit of wrong-headedness in this “change is good, you go first” mindset about same-day is the idea that this is like any other tweak to retail. It’s not. It’s more like trying something different on Black Friday. You can’t do that in July—it only makes sense on the day after Thanksgiving. (OK, as Black Friday marches toward Halloween, that’s no longer quite true, but you know what we mean.)

Same-day delivery really is made for December. Any tests a chain doesn’t make now will have to wait a year for the results that really

matter. Any chain that’s testing now could have a huge advantage that competitors will literally have to wait a year to catch up to.

And if it sounds like that’s not a problem because the USPS may have only one still-unrevealed chain to start its trial next week, think about this: Walmart is already doing same-day on its own, and several other unnamed chains are doing Google’s trial.

If you’re not one of them, you’re already a year behind.