Jeffrey got me thinking this morning about where people should start in choosing the right product to build. He suggested science and validation was key, and I replied that art and creative choice was key. Of course they both are, but since I've had this company for ten years where people ask us to research their products, I've grown to appreciate the artistic and creative side more, if I had to pick. It sounds like Jeffrey is encouraging people to lean more on science. What do you think?

I agree that it absolutely has to be a combination. I really want to believe in Nate's side of conversation, but startup history shows us that the more beautiful products don't always win (Gowalla vs. Foursquare, Rdio vs. Spotify). I'm not sure that's science, better marketing or luck.

I agree that art and creative choice is key – once you've decided what it is you're doing. Of course you can start the process if "wouldn't it be cool if..." but if you don't take the time to validate that assumption, then you're building a solution that's looking for a problem. IMO this is why most products fail. At the inception of an idea, one of the most dangerous things you can do is assume you know anything.

meant to say "...you can start the process *with*..." No edit button ;)

Oooh are we talking Tony Fadell (SVP of iPod and Nest creator) at Brooklyn Beta 2011?? The fact that he assigned most of iPod and iTunes product development success to them focusing solely on "wouldn't it be cool if..." really stuck out to me. And then his quote of "treat every member of your team like an artist" also seemed relevant. Heck, maybe I'll invite him to the branch. YOU NEVER KNOW.

Here's Richard Feynman describing the scientific method to students: youtube.com

"First, we guess. Then we compute the consequences of the guess to see: if this is right, what would it imply? Then we compare the results to nature. If it disagrees with experiments, it's wrong. In that simple statement is the key to science. It doesn't make a difference how beautiful your guess is, how smart you are, who made the guess or what his name is. If it disagrees with experiments, it's wrong."

Similarly, you could say every product is a hypothesis. You start with a guess: I think this will work, I think this will be better, I think this is true. And you build something to test that guess. Then you learn something, and you try again. That doesn't preclude art and creative choice and instinct and boldness at any point along the way. But it's a place to start, and it's a method to return to, to ground what you're doing. I like that.

The critical point where the scientific method breaks down in technology development is you can build something, put it with real users, test it, and have it essentially fail by all normal metrics, but still believe it's great for no reason than your own creative instincts. It's that point where the scientific process and art diverge - like with the iPod or tons of eventually successful products, there can be lots of failure at first, and no universal definition of a successful outcome.

That kind of failure is still failure. It doesn't always mean your guess is wrong, it can mean your experiment was wrong. And it's not a permanent state - it's a data point. The product evolves, and the environment evolves, and the really successful products are the ones where they converge at the right time. Inspiration and instinct are part of it, but they aren't enough on their own: it's fueled forward by a kind of disciplined ongoing observation, patience and persistence.

For me, art comes in along with analysis - usually in the form of metaphor. That's the connector between data and inspiration, and one of the most productive ways to "work" with qualitative data. (Though if you approach it as work, it's less effective than if you treat it as play.)

Been wanting to follow up for a while to clarify that I'm thinking of art as more than just the visual part of interaction design. Awesome technology experiences are damn near works of art. If it doesn't make everyone throw up in their mouths, I'm gonna quote Tolstoy:

"We are accustomed to understand art to be only what we hear and see…but this is but the smallest part of the art by which we communicate with each other in life." csulb.edu via @stephendodson

A dude named Jakob Bigelow, who wrote the first technology book in 1851, "Elements of Technology," believed tech was both art & science, but I think science gets more credit in technology design, and art deserves a little more.

Following up on Nate's point, it's helpful to define what exactly we're talking about. The scientific method is a tool -- by far the most effective one we have -- to understand how the world works. Art conveys human emotion and experiences; it is not aesthetics.

The scientific method is a superior mechanism to consistently reveal preferences and habits, like, do users prefer rounded buttons or squared ones? But if you need to create a truly wonderful technology experience for a user, it's limited. An understanding of what moves and awes people is what's required, and that's what art is.

In short, it depends on what you're trying to do.

"Every good scientist is half B. F. Skinner and half P. T. Barnum."
- Principal Seymour Skinner

Like most questions, the answer is probably grey rather than black or white.

...but if we're going to go all Celebrity Deathmatch, I'd rather put my money on art. Sorry, science.

Stephen sounds pretty right on to me. The scientific method is a framework – a structured way to question the world – and the output certainly doesn't have to be scientific. I'd go a step further and suggest that even the science-y bits of making the right product are themselves closer to art than science – say the creative process that goes into writing an elegant algorithm or repurposing a manufacturing process to fit your needs.

Go, art, go!