The next sentence is the most uninspiring opening line you will ever read – which is why I have put this one in front of it.

This morning on my way to work I got some money out of a cash machine.

I have no idea how many times I have done that before – a couple of thousand on roughest of rough guesses.  This time was just like all those other times. With almost perfect consistency, I use cash machines to get cash.  The same amount.  And no, I don’t want a receipt, thank you.

So why, if my answers are always the same, do I always get asked the same questions? A small frustrated tweet went out to the world:

Life would be better if the first question cash machines asked after checking PIN was “The usual?”

The unexpected response, from Ruth Kennedy (who in turn had got it from Amanda Gore) was the extraordinary news that such a cash machine might exist, with a link to a fantastic design project by Ideo for a Spanish bank, BBVA, where

The question was not how to further automate the teller, but rather how to humanize the machine…

It’s what happens when you build an ATM from user up rather than components down.

Watch the video on the project site for an explanation of what’s different, though in one sense the answer is much less interesting than the way of exploring the question. For a public strategist, I think there are two important points to think about.

The first is that even services which are mundane to the point of invisibility can be radically improved, if only assumptions about how it has always been done before are jettisoned and if there is a relentless focus on helping customers get done the things they want to get done. It also requires – though this is less immediately obvious from the BBVA case study – a willingness to change things deep in the structure as well as in what is commonly understood to be the user interface. The problem with today’s cash machines is not just that they haven’t been designed to ask me whether I want the same service as last time; it is also that they and the network of which they are nodes are not designed to remember what that service was in the first place.

The second may appear to contradict the first. It is that change can be disconcerting to customers even if it is a change which quickly becomes second nature. I wrote earlier this year about the fact that however obvious self-service supermarkets and station ticket barriers may appear to us now, there was a time when they were new and disconcerting. So it is as important to understand the obstacles and inhibitors from a customer’s perspective as it is to understand the improvements they seek and will value.

We have no shortage of user interfaces where the underlying design has changed as little for as long as the basic cash machine, where quite literally we ask the question we have always asked and where the opportunities should be so much greater than for a cash machine because the underlying service the process supports is so much more complicated.

And if this is how they do business, I might just want to move my account to BBVA.


  1. I love the concept. I really do. Seeing obvious ways to make a better service experience from a mundane starting point gets me fired up too. I once wrote a piece about lifts (perhaps slightly inspired by Douglas Adams and the lifts that could see into the near future and thus arrive on time). I argued that there were dozens of small design parameters in even the most linear, binary experience – a lift moves between two floors – that could make it slightly better. From pre-setting of the lift’s start position based on time of day, to leaving the doors open, to pattern-of-use learning.

    In the cashpoint case I think you do rather easily dismiss the additional complication that the network was never designed to capture and retain previous-use information. Yes, the exchange of credentials with the cash-issuing bank – and checking for available balance etc – could include a poll for “previous cash transactions on this card” – or a proxy parameter set by the bank, having established a pattern itself – but that puts a fair few new capacity and information design burdens on the bank.

    How far would one really go in redesigning the whole transaction recording paradigm in order to change the user experience from “pressing one button out of a choice of six” to “pressing one button”?

    But the point in general is very well made.

    1. I wasn’t intending to dismiss it at all. On the contrary, as you say, what may appear to be a superficial tweak to the user interface can easily become – and in this case would need to become – a significant change to the deep system architecture.

      From what’s apparent from the video, the Ideo work may be more vulnerable to that challenge. Quite apart from building in learning and history, I thought their idea of simply rotating the machines by 90 degrees was an example of inspired and elegant design. But that was quickly followed by a second thought, how exactly does that work in a hole in the wall style machine (dominant model) as opposed to an inside the branch model (much less prevalent than it used to be)? And all of a sudden we have to contemplate redesgning and changing buildings and streets, not just swapping one machine for another.

Comments are closed.