Re-reading Frank Chimero’s excellent article, The Web’s Grain, I was struck by this simple point he makes:
"...the less you have to do, the less say you have"
A simple view of this could see us equating quite a few different things:
|more convenience||=||less stuff to do as a user||=||less control as a user||=||more choices to design on the users behalf|
It isn’t that simple though.
For example, there could be occasions where..
- it is more convenient to do more (e.g. a bit of extra work during on boarding could really help me get the most out a new product/service)
- doing less could lead to more control as a user (e.g. by removing the needless manual steps whilst retaining the more controlly features)
- fewer controls could lead to less choices to design on the users behalf (e.g. you could move to a simplified service proposition)
Perhaps we should equate these things instead
|Being overly driven by convenience / automation||=||Removing the wrong manual steps||=||Taking the wrong choices / controls away||=||Over designing (of a type)|
Automation saves time?
Of course it does. But at the same time it really hasn’t, has it?…
Automation is appealing. Remove the drudgery. Make things easier. Free up time to do more important things. For example in banking, automation might allow customer service people to give more attention to the customer. Nice!
However while basic banking (checking balances, making transfers and whatnot) can be largely automated, this doesn’t mean that we can remove the manual / human assisted option. There are just too many people who still need / prefer it. And when a question pops up that our automated process can’t handle / hasn’t catered for.. we need a manual process there too (or one that involves simpler ways to interact with systems of record).
The service we provide is always going to be bigger than the digital element. Likewise, the digital estate for staff will always more complex than for customer self-service. They have to help with more complex journeys – coaching customers through their financial decision making, their emergencies or difficulties. Automating these experiences doesn’t feel right. Nor does creating a self-service experience. These are moments for staff to shine.
Automation makes us more efficient?
Again the answer is yes, of course. And no.
Humans are emotional and fallible. We might one day wonder why they were ever trusted to perform surgery! Or why they were ever allowed to drive cars! We’re not there yet – and there are still many important things humans do best. There are situations where throwing an algorithm into the mix just doesn’t help. For example, asking a distressed and worried customer to view a tailored set of FAQs, really doesn’t help with efficiency. Nor would asking a financially astute customer to step through an attitude to risk questionnaire.
We need to learn when to get out of the way of a human to human interaction
Designing for a better human and computer symbiosis
Here’s a set of things to watch out for then (I’m sure there’s more):
- Remember what humans do better
- Allow for humans to use your software in creative ways
- Be humble about what we know (create interactions that can flex and learn)
- Allow for non-linear self-directed flows with only occasional modal interactions
- Interpret data but be as transparent as possible about how you did it (so it can be checked if necessary)
- Detect moments when humans might need to take over quickly (especially customer service humans)
- Never take on more than you can be completely trusted with
- Don’t take on sexy challenges (e.g. an AI chat interface) until you’ve got all the simple stuff working really really well