For part of my HCI course I had to pick a “good” interface and a “bad” one to compare them to. I picked Heroku scaling dashboard versus AWS elastic beanstalk scaling page.
To do this I had to actuall soon up an elastic beanstalk instance, something I’ve never done before.
I took my screenshots and wrapped up my report. The HCI instrctor said it was a “great” example of one interface being better. I thought that’s where the story would end. But...
A few weeks later I got an email with a bill. For an AWS service that is still running... a service that I thought I had terminated, but apparently didn’t. I had to contact support and even following their directions it still unclear if I’m being charged.
I thought I was putting together essentially a superficial argument for the assignment, but the financial repercussions kinda hammered home just how a bad interface can influence and encourage incorrect usage.
This experience made me think of two general conversations i've had with people over and over again since I started at Heroku:
The first goes something like this: "You're just a fancy layer on top of AWS, aren't you worried that they'll wake up one day, make a prettier dashboard and you'll be out of business?"
The first time I heard this, I did have a mini-panic attack. What if AWS suddenly did start caring about user experience, interaction design, and experience design? Are we enough of a value add to justify our continued existence?
I used to think that our saving grace was how far ahead we were, that if we kept improving and kept iterating. Even if AWS started today, it would be a long time until they were close to UX parity. However, 6 years later and I don't know if AWS is even capable of doing what we do
Shifting from "the user can technically do task <blank>" to "the user enjoys doing task <blank>" isn't a thing you can sprinkle on. It has to grow from the roots and be in the DNA of every part of your product.
This leads to my second question I got asked a bunch which is something like: "You're only a layer of paint on top of AWS, is what you do really that important? Why do you cost so much?"
I hopefully already covered why we're not just "a layer of paint", but what about the second part? Why should people pay to experience joy while using my service when they can technically get the same requirements elsewhere for cheaper? It's a valid question....
And I don't think I really understood until this beanstalk interaction. While the bill was only $10, there was a physical and tangible cost to me using a poor interface. While that $10 is fairly cheap in the grand scheme of things it also cost me my time and my attention.
It also cost me a piece of mind. I've told other people in the past "your time isn't free" when they compare Heroku costs, but i've never before had to try to open an AWS support ticket and navigate the AWS console on my phone while in the parking lot of my kid's daycare.
We as humans are bad at making comparisons, especially when one side is easily quantifiable with dollar signs and the other is very intangible. Not saying that the balance sheet will always win out for the "good design" product, and not saying that Heroku is the greatest
I am saying that a well designed product cannot be an afterthought, and is definitively worth something. This information is not new, or trans-formative. However it's different when you hear someone say it versus when you actually experience it.
