Evan Schuman's StorefrontBacktalk

When Hewlett-Packard this week unveiled its prototype Retail Store Assistant?a kiosk that offers customized individualized promotions transmitted directly to a customer’s smartphone?it took a business pragmatic position in the personal shopper battle.

There’s little debate that customized shopping?taking the CRM customer-specific data one major step beyond the minimally effective coupons at the end of checkout?is almost certainly going to be a key in-store technology over the next few years, especially with grocers. But the form that this customized shopping will take is a key debate.

At its most basic level, there are two opposing approaches: manufacturing a handheld mobile unit that is given to shoppers; and leveraging the technology (usually a smartphone) that the customer already owns, carries and is presumably comfortable with.

The manufacturing and issuing approach?with Motorola one of the dominant players–can deliver the most powerful, flexible and consistent experience, but it requires the retailer to purchase large numbers of the units, at perhaps $700 each.

The “leverage what the customer already has” approach?now supported by HP–means no hardware cost and no hardware to track, but it severely limits functionality because programmers must write to the lowest-common denominator of multiple operating systems, hardware platforms, carriers and applications.

Mohamed Dekhil, manager of retail applications in the Digital Imaging and Printing Lab at HP Labs, said weighing the two is easy when business realities are considered. The approach that will survive is the one that most quickly gets a dominant marketshare of major retailers. With that goal, the approach that promises much lower costs to each retailer is the way to go, he said.

“I’m not saying which (approach) is better. I’m saying which one is more feasible,” Dekhil said, adding that HP’s prototype approach uses a store-based kiosk. “You may lose some enhanced features” because you’re using the consumer’s phone, he said, but that functionality could be delivered from the kiosk.

The HP approach calls for consumers to first have to register online or in-store. That registration requires the store’s loyalty card, which ties in the customer’s CRM history. It also requires the customer to take out his/her Bluetooth-enabled smartphone and pair it with the kiosk, so that data can be exchanged. The phone then also can serve as an identification/authentication device.

Ultimately, HP wants the customer’s smartphone to be able to communicate directly with a store’s self-checkout system, but they don’t have that capability yet, even in a prototype. This system could also leverage future item-level RFID systems, Dekhil said.

The HP kiosks?which Dekhil said cost about $8,000 to $10,000 in small-volume production and could “go down to a few thousand dollars” in large-volume retail deployments?can perform several functions beyond customer interactions, including inventory, POS, supply-chain and employee training programs.

For the consumer, the kiosk-to-smartphone system can deliver realtime promotions based on that customer’s history as well as inventory triggers at the store. But it can also display store maps, print recipes, remind the customer of items the customer entered into a Web site and play private voicemail messages (for example, a family member asking for a particular purchase).

The kiosk can print out color coupons and recipes or can beam them to the customer’s phone. But HP likes to use the printouts because it conveniently sidesteps the technical issues with transmitting data to the consumer’s device.

Without an HP application installed on the phone, how will the customer know where the data is sent? Will it be in a consistent place and have a consistent look? If a carrier or a PDA hardware manufacturer or an operating system company issues an upgrade that interferes with the kiosk interaction, is it likely anyone in the store would be able to troubleshoot the problem?

Dekhil said one way to try and standardize the experience is to have customers install a small proprietary applet, but that gets into the same OS/platform issues. “We said, ‘Is there something already in the phone that we can use?'” and BlueTooth was the initial answer, Dekhil said.