[thinkcomp]

It's ironic to see this article coming from Farhad Manjoo.

I wrote a response to his original piece a while ago:

http://www.aarongreenspan.com/writing/essay.html?id=51

What Square has done with card case is pretty slick, but at present it doesn't scale. Paying with your name doesn't work at Wal-mart. And funding with your plastic credit or debit card doesn't work especially well for most coffee shops or other small businesses that want to save on interchange fees, not pay more of them.

I look forward to competing with Square again when I get my California license.

[gojomo]

Why wouldn't it scale to Wal-mart? Even with a dozen people with the same name in the store, a cashier can quickly pick a matching face: everyone is good at that (except for prosopagnosics).

And that's without assuming Wal-mart adds any extra hardware that could narrow the position of handhelds to individual registers.

[thinkcomp]

That's a pretty long scroll box to find the person in the first place. Or an added step as you key in the first few letters of the person's name. At large retailers or QSR chains that put a premium on fast checkout, neither option will fly.

I obviously appreciate the simplicity of paying with your face.

[gojomo]

I doubt any firstname+lastname (or even just firstname+last_initial) situation would require a scroll box... you could fit 20+ recognizable faces on a screen. Sort them by age and in the few cases where there are that many collisions, cashiers would know just where to look.

Keying the name is a bigger issue... so I suspect terminals in larger retailers will eventually just detect the N nearest pay-eligible handhelds.

[thinkcomp]

The additional hardware (plus integration cost) is something that reatilers don't want to spend on.

In contrast, barcode scanners are already there and they typically emulate keyboards, so no special integation is required.

I'm not convinced that paying with your name scales for most retailers. The retailers I've talked to aren't convinced, either.

[gojomo]

Barcode scanning is another good option to speed the initial account-identification.

I suspect in coming years, multiple systems will all work together to train users to bring up barcodes on their handhelds for specific purposes (as with boarding passes or "Jonathan's Card").

[pbreit]

Swiping a card is sounding a lot easier right about now.

[akavi]

A pretty obvious next step would be to pair this with facial recognition.

That'd cut down the search space dramatically.

[ch]

With NFC, they could just limit the list to devices in close proximity to the point of sale. That said iPhone doesn't have NFC so they need to solve that problem some other way.

The general point of scale issues still is valid, but culling the list of payers seems surmountable.

[sausagefeet]

I wonder how they'll keep people from putting products on someone else's tab.

[gojomo]

As noted in the article and alluded to in my comment, the cashier visually verifies that the person requesting the charge matches the photo that appears on their screen. That's a whole lot better than the pretend 'signature verification' purporting to deter other kinds of card misuse.

Identical twins and celebrity lookalikes may still present problems.

[sausagefeet]

> Identical twins and celebrity lookalikes may still present problems

Or just someone that doesn't care about their job that much putting their friends pack of gum on the same tab as the guy being 100 things.