Evan Schuman's StorefrontBacktalk

And that explanation might have to be face-to-face because, under the trial’s rules, customers can ask for same-day delivery either from a Web site or in-person at a store (presumably that would be a store that doesn’t have the item in stock at the moment but is part of a chain that can get it into the system that day).

That 200-package limit isn’t a show-stopper, but it will mean chains in the test will have to find back-up delivery services or be willing to just say no to late-buying customers. This is, after all, a test, and 200 sets of data points per day is still going to be very useful to chains in deciding whether or not to

keep offering same-day delivery.

That data should also be enough to answer USPS questions about the viability of a service like this. The mail service is currently running a $15 billion annual deficit, because first-class mail has dropped off, but all the infrastructure for deliveries is still in place. The fact the U.S. Postal Service has that type of reach and capacity suggests it’s perfectly positioned for same-day delivery, for both chains and pure-play E-tailers.

The USPS is optimistic about Metro Post—USPS Spokesperson Friess calls it “a game changer.” That would sound like just a cliché, except federal law literally requires that nothing the USPS does can create market disruption, even though disruption (for the better) is exactly what same-day delivery is all about.

In fairness, postal regulators have to obey laws that are designed to keep the USPS from competing with for-profit delivery services. But so far, those for-profit efforts at same-day deliveries for retailers have been all over the map—everywhere but in the neighborhood called “successful.”

For example, ebay’s same-day experiment was always intended to be a loss-leader using commercial messenger services, and it still can’t manage to guarantee either delivery times or even product prices. (There’s an amusing blow-by-blow of ebay’s service in New York City, where it just started testing, by the Silicon Alley blog BetaBeat. Warning: If you’re offended by BetaBeat’s headline—which we won’t quote for fear of triggering spam filters—it may not be for you.)

Walmart, which has been doing same-day grocery deliveries in San Francisco since 2010 using its own delivery trucks, expanded that service this month to include some general merchandise. But unlike other Walmart same-day cities (Philadelphia, Minneapolis and the Washington, D.C., suburbs), where customers have until noon to order and deliveries start at 4:00 p.m., San Francisco customers must order by 8:00 a.m. but can choose delivery times with one-, two- or four-hour windows beginning at 2:00 p.m.

That type of tinkering with same-day schedules is exactly what chains should be doing at this point. It’s early in the development process for same-day delivery. Sure, Amazon has been offering it for years without making much of a ripple. But other big players are now experimenting and small, online apparel retailers like Net-a-Porter.com are trying it themselves in their home cities.

It’s just the beginning of what looks increasingly like a same-day arms race. Nobody knows what customers actually want or will accept from same-day delivery. At the same time, it’s holiday selling-frenzy time, when same-day delivery has probably its best shot at proving itself profitable.

And this is the time when regulators want to hobble both the USPS and retailers, with limits like 200 packages per day on Christmas Eve? Ho. Ho. Ho.