• I've worked on the web since it's been possible, typically for start-ups, agencies, and new product groups where everyone wears a lot of hats. Nearly every role has been a hybrid of UX / visual design, content strategy, and marketing. I love clean, artful, scalable information design that makes it easy for people to do what they want to do. (And what you want them to do.) I'm a strong team lead, mentor, and an exceptional verbal and visual communicator.

Measuring Magic: How We Can Do Better than “I Know It When I See It”

Designing the future of tech using the HAPI emotional model

Transcript from my lightning talk at Conflux: Future Possible, the Amazon design conference on September 21, 2017.

Maybe like you, I’ve been asked to design “magical” experiences. I’ve worked on early-stage voice shopping, predictive carts, and now physical stores like Amazon Go.

Magic IS our business.

But sometimes our greatest achievements are met with a mixture of excitement and fear, such as around privacy or intrusion:

So how do we know what is… “cool” or “creepy”, “seamless”, “intuitive”… or even “magical”? Emotional qualities like these can be hard to define and, when you ask someone to do so, you often get a response like — as in U.S. obscenity law:

But come on… This is Amazon.

We can do better than “I know it when I see it.”


An example is the “Uncanny Valley” concept by Japanese robotics professor Masahiro Mori in the 1970s, which I think still holds up. The general idea being that between a robot looking and sounding like a robot, and a human looking and sounding like a human, there is a range of combinations that may seem friendly or familiar… or “off” or creepy:

In a similar vein, I’ve developed the HAPI Model — a framework or language we can use when designing new products, on a spectrum of Normal to Magic to Fear:

Just to run through these quickly:

Humanity: This is basically the “uncanny valley” concept I showed you earlier, for humanoid things, like with voices or faces.

Autonomy: This goes from a [monthly automatic delivery] — “set it and forget it”— to random packages appearing on my doorstep because maybe a machine thinks I wanted them.

Predictability: This is the difference between what we usually call “surprise & delight,” like Alexa’s new skill last week: Alexa, sound like a cat… to Alexa making animal noises day and night whenever she feels like it.


Intimacy: This goes from “Hi Jill” on the [Amazon.com homepage]… to walking in the door and hearing “Hi Jill. I don’t really like that dress on you, and you look a little tired to today.”

And the perception of these shifts to Normal…

over time… such as generation acceptance, or

over value… what discomfort people are willing to trade for convenience or benefit.

Here is favorite tweet on the topic:

Another fun example is how when trains first came out, newspapers had scary headlines about the dangers of your organs flying out, because it was unnatural for humans to travel at such high speeds.

So could we apply this to what we design?

Here are a few quick examples. Now, I just made these up based on general media response — but how we could know this in practice is through user research, usability testing, and all the other ways we know our customers.

Echo when it first came out 3 years ago: People knew Siri and AIs, but… It’s always listening?

Alexa today: Little kids love her and she appears on Saturday Night Live.

But earlier this year with the release of Echo Show and Look, you see some of that fear factor bubbling up again, around things like the camera, or Drop In feature, or placement in the bedroom:

So back to my earlier question:

How can we do better than “I know it when I see it?”

I think it’s through having a model or language we can use to talk and think about designing the future for our customers.

Know the magic to grow the magic.

Thank you!