Self-service setbacks at the high-tech ATM? Check

Technology can improve the customer experience. This usually translates into more money for the company doing the improving.

So-called self-service retail is a great example of this. For most purposes, I'm much happier using Expedia and similar services online than I was going to a travel agent's office in the old days. The recommendation engine, customer reviews, streamlined ordering system, and sheer scale of Amazon.com present for many things a vast improvement over traditional brick-and-mortar shopping.

This isn't always the case, of course. I find it stupefying that Lowes and Home Depot are the two chains I use that have jumped most enthusiastically into self-checkout. It's difficult for me to fathom a thought process that would appear to have gone along the lines of: We carry this huge mix of big/awkward (e.g. lumber) and small/unlabeled (e.g. bolts) items, so we should obviously lead the retail industry in forcing our customers to try scanning and bagging them themselves.

Actually, I have a pretty good idea what the thought process was. Some financial guys figured it would save money. The spreadsheet said so. I have to wonder, though, whether, before rolling this system out on a large scale, any of those financial guys watched the system in use at a store and saw 1.) how annoyed customers get with it much of the time and 2.) how often one of their employees has to step in and help with a transaction, anyway. This is basic user experience testing, which is often taken to be synonymous with user interface testing but is much broader.

On Monday, I came across another great example of bad user experience, courtesy of Bank of America.

Automated teller machines, or ATMs, are one of the great success stories of self-service enabled by computer technology. Those of you under 40 or so probably don't remember just how inconvenient it used to be when you had to get to a physical bank branch during "bankers' hours" pretty much every time you needed to withdraw money from or deposit checks to your checking account. And this was during economic times much more based on cash and physical checks than today.

ATMs were truly liberating. True, there weren't all that many of them at first, and the networks were a bit fragmented. But it was still a vast improvement. Liberating, even. Thus, it's really quite an accomplishment on the part of Bank of America that it has been able to implement brand-spanking-new technology in a way that makes using its ATMs worse.

Here's the background: Bank of America has been rolling out a new ATM system over the past couple of years in which, rather than stuffing your checks in a deposit envelope, you feed them one by one into the ATM, which sequentially tries to read the amount (sometimes successfully).

Analysis of this move by the Gerson Lehrman Group suggests that "customers are able to receive immediate credit in real time for cash and/or check deposits, which may improve customer satisfaction and retention, and increase deposits at the ATM."

But let's not kid ourselves that this is about the customer. Gerson Lehrman's money quote is that "banks and credit unions are reaping the benefits of imaging ATMs with no envelopes because the technology eliminates fraud at the ATM, which surpassed $700 million in 2006, reduces transportation and back-office processing costs, and improves cash float." This is primarily about saving money, not pleasing customers.

There's a problem with this system, though. It works well enough for a single check, and I'm happy to concede that it's even an improvement over the envelope system. But it basically breaks multicheck deposits at the ATM.

Although the ATM suggests that you can deposit up to 10 checks, in practice, this process takes about a minute per check. I was depositing a grand total of five checks on Monday, and I (literally) was nearly punched out by the person behind me because I "should %$%#@ go inside if I'm going to do that much banking." Hey, depositing five checks wasn't that much banking before Bank of America "improved" its ATMs.

Did anyone at Bank of America actually observe people using this system as it rolled it out? I'm admittedly not familiar with all the details of the system in question, but it would seem an easy enough thing to accommodate both envelopes and direct scanning.

But perhaps I'm unfair. Perhaps Bank of America did extensive user experience testing. And it just didn't care because it figured, rightly or wrongly, that even with forcing customers to take up teller time in a physical bank branch, it'd still come out ahead financially. And you don't really think customer satisfaction played a large part in any calculation, do you?