No matter what you think of playing clerk, self-service is neither fad nor trend: It's revolutionizing how we are separated from our money. IBM certainly believes so, as evidenced by Big Blue's November purchase of self-checkout vendor Productivity Solutions.My first experience with self-service checkout last year did not endear the retailer to me in that it was unacceptably long lines at the staffed lanes rather than curiosity driving the experiment. What? Drop a couple hundred bucks on a cartload of stuff and I have to do my own scanning? Geesh.By the second run through the self-serve line that resentment had dissipated . . . and by the third I was perfectly happy to never again speak to a clerk.Why the transformation?Time. It's just faster to get through self-service checkout, where lines are nonexistent most of the time and mercifully short the others. Self-serve is so much faster that I no longer care if the store is squeezing an extra couple of bucks out of my semi-forced labor. Give me the extra minutes, and we'll call it a reasonable deal.Many of you have reached the same conclusion, willingly or not, according to research from IHL Consulting Group, which sells market insight to retailers and the IT vendors that serve them. Consumers using self-checkout machines rang up $128 billion in purchases last year, an increase of 80% over 2002, says IHL. Similar spikes are expected this year and next, with the total haul from such transactions reaching an astronomical $1.3 trillion by 2007.So no matter what you think of playing clerk, self-service is neither fad nor trend: It's revolutionizing how we are separated from our money.IBM\u00a0certainly believes so, as evidenced by Big Blue's November purchase of self-checkout vendor Productivity Solutions."That's going to be a big jump to this market because IBM is the [point-of-sale] vendor for about 75% of the top 20 front-lane checkout retailers," says IHL President Greg Buzek. "Now that they own the self-checkout platform, they're going to be able to integrate stuff in a much more cost-effective way, thus bringing down the cost and increasing the payback" for retailers."The operating system that runs most front-end checkouts - the IBM 4690 - is extremely robust and reliable. It was designed for nuclear subs," Buzek says. "But the bad side of that is that it's difficult to be interfaced to, so up until the last three or four months - before IBM got in the business - you've got a [point-of-sale] system that is running this operating system trying to interface to a Windows NT [self-checkout] system, and then there's translation-layer software. So you actually have two to three boxes in that lane, thus increasing the cost."Better integration from IBM should knock a few grand off the price of self-checkout systems, which now run between $21,000 and $23,000 per lane, he says.And those shrinking capital costs are only the icing on an already attractive proposition."Typically a supermarket has to sell $15 worth of stuff before they break even on a transaction with a staffed lane," Buzek says. "With self-checkout you're able to take that down two or three dollars, so your profitability comes much faster. Once the unit's paid for it's even less; they can start making money on a $10 transaction."Unless the human beings screw things up."It is not a technology that you simply put out there and hope customers get it," he says. "Where it has failed, it's failed because retailers have done just that. Kmart did a horrible job implementing it in the fact they put it out there and they did not provide incentives to their cashiers to take over the role of recruiting customers and training them how to do it. That's something [retail chain] Kroger does well."Retailers often face another obstacle when attempting to implement self-service checkout: labor unions. . . . No need to ask why.Writing to Buzz has always been self-service. The address is firstname.lastname@example.org.