Take the Money and Stand Still

Last week I tweeted this eye-opening guide to spotting an ATM skimmer published by the invaluable Snopes.com. Like a lot of people, I’d heard of ATM skimmers before — duplicate card readers and wireless cameras surreptitiously attached to cash machines with the intent of stealing your card number and PIN — but I had no idea what form they actually took. The visual evidence was striking; skimmers are uncanny mimics of the visual language of ATMs. The colors, shapes and peculiar plasticity that we’re all familiar with are faithfully reproduced in their ersatz forms. I had no idea they could blend into a cash machine’s hardware so expertly.


An Ugly Face

While I was surprised, scared and dismayed that I could have at some time in the recent past used a skimmer-outfitted ATM without realizing it, it also occurred to me that in many ways this is more than just a security problem. It’s a design problem too. In fact, design is at the heart of the problem: the thieves who install these skimmers are exploiting a profound flaw in the industrial design of ATMs.

Right: Swipe your card quickly. Snopes reveals what an ATM skimmer looks like.

Ask yourself: what exactly are all of those oddly proportioned boxes, varying planes, bizarre joins and strange angles that describe nearly every automated teller machine on the planet? Who among us who uses cash machines actually understands the purpose of all those expertly yet randomly fused-together shapes that are somehow intended to constitute a trustworthy money dispensing device?

The fact of the matter is, the superfluously futuristic form of these machines is so nonsensical, so utterly impractical and useless that even a quickly grafted foreign appendage like a skimmer is indistinguishable from the native hardware. The thieves who designed these admittedly ingenious tools had a much easier job of it because, like the ATMs onto which they’re attached, skimmers don’t have to look like they do anything that you would understand. They just need to look like they might do something you don’t understand.

Banking on Their Laurels

There’s no reason for this, no reason that these ATMs need to look like cast-offs from decades-old sci-fi television shows — except perhaps for the fact that the banking industry has left a gaping vacuum of thought where instead there should be a design vision.

Ironically, the ATM itself is perhaps the banking industry’s last great design achievement, at least according to economist Paul Volcker. The introduction of an an unmanned, relatively secure and continuously available cashier was a genuine advancement in commerce and an improvement to modern life.

The problem is that in the intervening decades since its introduction, the industry has spent all of its energies on things far more superficial, complex and useless: derivatives, credit default swaps and making ATMs look like Transformers among them. All of which has been intended to bamboozle customers into believing that the act of banking is so intricate and advanced that most of us shouldn’t bother trying to comprehend it.

That’s a digression though. More to the point is the question of what an ATM should look like if it shouldn’t look like the ATMs we have today? To begin with, ATM designers should abandon their strategy of intimidating customers through technologically imposing yet incomprehensible forms. Instead, they should focus on simple constructions, fewer planes, fewer parts, and a healthy dose of visual logic. New ATMs should be intuitive in the way that appliances and common tools are; the best designed of these forms communicate what they do at first glance and without ambiguity.

Should they be beautiful? It’s perhaps too much to ask banks to strive for aesthetic beauty in this endeavor because nearly everything they’ve ever produced in the past few decades has been blindingly ugly. The bar for success can be somewhat lower though: a new ATM design need only be simple and succinct enough in its form that it becomes difficult for a thief to attach something as flagrantly malicious as an ATM skimmer to it. It’s not much to ask, but it would be enough.

+
  1. Regarding safety, I guess the first step would be to make the interface a completely flat plane. This way, anyone could recognize extra seams and planes as “out of the ordinary.” As you’ve stated Khoi, extra seams and planes are completely within context.

  2. Especially unforgivable is how this confusing visual language transfers to other mildly bank-like, “secure”, and automated systems such as self-checkout lines at grocery stores.

  3. I’m not really qualified to comment on this, as I don’t have a bank account and have only used an ATM a few dozen times.

    Still, it seems obvious to me that the machine could be improved. And even though it probably wouldn’t make sense for a crook to run the above scam in most cases—when one can buy credit card numbers online for a few bucks, why incur the risks and costs sitting for hours near the ATM you’re monitoring?—that’s no excuse for it being possible in the few cases that running it would make sense.

    I think some of the IDEO solutions (privacy screen, touch ui, flat facade) are good. And if each bank only used one standard ATM model, as opposed to the hundreds of configurations they have now, it would probably improve security/credibility too.

    (One thing I’m not crazy about with the IDEO is how different their new shape is from that of existing ATMs—it’s throwing away a lot of visual equity and across-the-room recognizability that current ones have.)

    The only part of this post which I would have liked to hear more justification for was the idea that banks are trying to “bamboozle customers into believing that the act of banking is so intricate [...by...] intimidating [them] through technologically imposing yet incomprehensible forms.” To me, this seems like it would be harmful to banks in the long-term and that the banks may simply be contracting with vendors who are all building the ATMs from interchangeable parts thrown together as cheaply as possible.

    But again, given how little I know about the banking world, take the above with a huge grain of salt.

  4. The ATM has always been a topic of hot debate as more and more banks now-a-days prefer their customers to use teller and the internet for transactions than going to the bank itself. One of the biggest nuisances that I face is having to adjust to interfaces of different banks each with their own version of horrible UIs. I wish for these banks to get together and create a unified experience in their ATMs (of course, with privileges for their respective customers).

    Back in 2005, I had addressed some of the issues with ATMs and its definitely heartening to see the likes of BBVA and Wells Fargo take initiatives in creating a better ATM experience.

  5. This is absolutely genius, using the banks awful design decision against them. I’d imagine half the reason the started including all those awkward forms was to try and make it harder to attach things to the machines.

    Whole thing could be solved if the bank had used a flat cover around the card slot making it really obvious if something had been attached to the cover.

    Then again the banks don’t seem to have a clue about technology, last time I saw one crash I noticed the ATMs here appear to be running Windows 2000Ёаsurely you would have made a custom embedded system if you actually cared about security.

  6. The banks have minimal input. Even the largest constitute not enough market influence to make manufacturers like Diebold care.

    The manufacturers have the information needed to make better devices. I have talked to more than one person who worked on ATMs or self-checkout systems for these big makers. They came up with what you and I would: simple, seamless, easy-to-understand systems.

    Which are rejected out of hand. For cheapness (but we already have these components laying around). For engineering expediency (it’s easier to make it like this). And for no good reason (we’ve always done it like this).

    Anyone seen how contactless payment methods are being implemented? Existing pinpads with a block glued to the top replacing the magstripe reader. Lazy design, not helping anyone.

    With the amazing awfulness by which, say, eVoting machines are built — as all-new products — I have zero faith that card readers, ATMs or self-checkout devices will improve…ever.

  7. My solution is to only use Chase ATM machines where I am comfortable with the design palette enough to know when something is off. Ironically though banking with Chase can often be equated to being stolen from.

  8. The Chase branch near my house just got new machines with the check scanning feature. I noticed that they were now making the card reader area of the machine out of transparent green plastic, so you can see the card reading mechanism underneath. I believe this was done specifically to make skimmers easier to detect. The machines are still quite ugly, though…

  9. It seems to me very few banks care enough about extending the design of their service beyond even the UI of the ATM machine. Citibank and Wells Fargo in the past have done an admirable job in years past.

    Because this isn’t an iPhone-like landscape where banks own the hardware and software, but more like Android, banks have to balance requirements across multiple hardware and software vendors, legacy systems, etc. Vendors don’t seem to have a clue (or care) about ergonomics, interaction design, or fraud prevention. At the end of the day, cost is a major issue, the machines are probably leased, and banks have little interest in driving up costs with special requests from vendors.

    That said, a large enough bank should know it can pull off a JetBlue or Virgin-like experience by working with vendors to craft something better. Coupled with well-designed services and marketing, a bank might be able to gain some traction by thinking about how ATMs fit into the design of the entire service.

  10. In my opinion ticket machines on trams and trains have undergone quite a positive change (in some cities at least). I remember it to be a real pain in the youknowwhere to buy a simple ticket when I was younger, but some service providers have invested heavily in user testing over here in Europe.
    Some of those machines have interesting and innovative UIs and step by step process signage to make it easier to pick the right ticket.

    It might not be directly related, but I think ATM machines and other public machines like parking meters could benefit from such research.

  11. One of the first things that stood out about the seams and lines of current ATMs while watching the IDEO video was that the large curves and clunky features denote some sense of “heaviness” and by association security. Watching the IDEO video, it seemed like the slim touchscreen ATMs seemed easier to break(-into). The current ATMs have a look like an armored truck, so angular, bulky lines may lend the right effect.

    Also, I don’t think an all touch screen solution would ever work due to accessibility issues for the visually impaired.

  12. A significant number of UK banks changed their ATM designs in response to this a few years ago, and a number also took to displaying pictures of what the machine should look like – either in a window next to the ATM or on the screen of the ATM itself. These seem to be less common now but I’ll snap a pic if I see one.

  13. I agree with our host, but I have issues with the IDEO ATM redesign that I feel are crucially important.

    To me there are two great advances and one great drawback to the IDEO design for BBVA. The advances are the improved physical orientation of the machines, so the queue of next-users remains visible; and, the machine’s personalized memory of recent transactions.

    The biggest drawback is the use of a touch screen.

    While touch screens have improved tremendously in terms of accuracy and usability, they remain persistently out of reach for people with visual impairments.

    While it’s true that we can’t design a machine that successfully reaches every last person, the touch screen needlessly disenfranchises as many as 25 million Americans (“25.2 million Americans report significant vision loss,” source). People who cannot see can still read raised text (Braille), and can also learn the physical layout of the numeric keypad and the transaction buttons that surround the display.

    With a touchscreen, even if an audio-assist is available, what will the voice-over say? “Move your finger 47 millimeters from the left edge of the screen, and 24 millimeters up from the bottom edge of the screen, and press once. You will receive no tactile feedback.”

    Especially with a machine as vital as an ATM, it’s imperative that designers be as inclusive as possible.

  14. Steve H: That’s a very valid point. I’m out of my depth here in terms of accessibility, but is there any hope of haptic feedback improving the viability of touch screens for the visually impaired?