A lesson from ATMs: Don’t add features your users don’t want!

Lately, every time I’ve visited an ATM, I’ve come away a little bit annoyed. This has made me think about what we do when we make websites, and about not saying “yes” to a new feature unless we really know our users want it.

Getting money out of an ATM must be one of the most meticulously designed user experiences in the world. ATMs have been around – at least in prototype form – since the early 1960s, so designers and engineers have had plenty of time to refine how they interact with customers.

Most of the time, they’ve got it down to a fine art. Getting cash from an ATM has become such a streamlined experience that we hardly need to think about each step as it happens.

But sometimes, someone decides to throw a spanner in the works.

About a year ago, the NAB, one of Australia’s “big four” banks, suddenly introduced a new step in their ATM workflow. Right after you’ve told the ATM whether or not you want a receipt, when the machine would normally start counting your money, you get this prompt:

“Do you want to save this withdrawal as your favourite transaction?”

You then have to select “Yes” or “No” before the transaction proceeds.

So let’s weigh it up.

This prompt appears every single time you use an ATM. At a rough guess, it adds 2 seconds to every transaction.

On the other hand, let’s just say there are people out there are actually grateful for this feature, and go ahead to save a favourite transaction. This will presumably save them time on future transactions, by reducing the number of buttons they have to click. Let’s be very generous and say it will save them 6 seconds a time.

So in order to save one group of customers 6 seconds per ATM visit, the bank is wasting 2 seconds per visit for the rest of its customers.

Even assuming a perfectly linear, utilitarian world where all time was equal, the bank would need at least 25% of its customers to use the favourite transaction feature, in order to “break even” on the time they wasted for the other 75%.

It’s actually not that simple, because all time is not equal: customers’ annoyance when you waste their time is stronger than their gratitude when you save them time. (Human nature, I’m afraid.)

Who actually wants this feature?

Let’s give the bank the benefit of the doubt, however, and say our 25% figure is valid as a rough guess. Does one person in four actually have a favourite ATM transaction? Even if they do, do they really want to “save” it? Is saving a few seconds worth adding mental clutter to an experience that’s otherwise second nature? (If they use the feature, they’re now going to have to make a choice every time they visit an ATM about whether to use their favourite transaction or a non-favourite one.)

Now, I don’t really know the answers to these questions. It’s possible that this favourite transaction feature comes out of extensive user research, and is actually something the bank’s customers are clamouring for. Sure, it annoys me, but maybe I’m the exception.

If that’s the case, I apologise, but I have to say, every instinct I have tells me that this feature isn’t based on user research at all. I’d be willing to bet that almost every customer, when faced with the “favourite transaction” prompt, scratches their head a bit and ends up choosing “No”.

What I suspect really happened is that one of the bank’s executives became obsessed with the idea of “personalisation”. You can almost hear the thoughts whirring away: “We want it to look like we treat our customers like individuals. I know, we’ll personalise their ATM experience!”

I’d be equally willing to bet that somewhere within NAB, there are user experience professionals who’ve spent their whole lives optimising the ATM experience, and are now tearing their hair out that a directive from above has effectively vandalised their good work.

Of course, I’m not saying a new feature can never be useful. There have been some really great features added to ATMs in the recent past. Being able to choose between different languages is a good example. Not only does this feature have a clear use case (you’ll appreciate it if you’ve ever travelled overseas), it was also added in a way that doesn’t interrupt your flow when you don’t need it.

The lesson for websites

So what should website owners take out of this? Don’t add features just because you think they’re cool, especially if they’re going to add extra steps or mental clutter to a common user task. If you’re selling underwear, your customers might not be keen to share their purchases with their Facebook friends. If you’re providing medical information, your readers probably don’t need to be able to customise the background image. If you’re a paper supplier, nobody wants to read your Twitter feed.

To get a website that works, find out what your users actually want to do, and help them do it.

« back to blog index