Nobody focus-grouped the iPhone

Part two of a series. Part one, What Towns Could Learn from Apple, argued that someone has to hold the whole. This one asks where that person's answers should come from.

The meeting

If you've ever been to a public meeting about a new development, you know how it goes.

There's a rendering on an easel. There's a man in the third row who has done the math on traffic himself. There's a woman who moved here for the quiet and would like it to stay that way. Somebody asks about parking. Somebody else asks about parking. The developer's consultant says "great question" eleven times.

And at the end, the plan has more parking, wider streets, fewer of the things that were interesting, and everyone leaves a little less sure why they came.

I don't say this to make fun of anybody in that room. They're right to show up. It's their town. But I've noticed that the process almost never produces a better place. It produces a safer, blander version of the place, plus more parking.

Steve Jobs was asked once why Apple didn't use focus groups. "It's really hard to design products by focus groups," he said. "A lot of times, people don't know what they want until you show it to them."

It's an arrogant line. And it leaves out the most interesting part, which is that Apple was listening obsessively. They just weren't asking people to design the answer. They were asking them to reveal the problems.

In the last post, I argued that a place needs someone with the authority to say what it should be, which raises the question I'd be asking if I were you: so the designer just knows best? We skip the meeting and trust the person with the sketchbook?

No. But the real answer is more interesting than yes or no, and Apple is a surprisingly good way into it.

Where does "this is good" come from?

Here's the question under the question. If a designer shouldn't simply build what the public asks for, then when she says "this is good," what is that claim resting on?

There are really only three answers.

What I like. That's the answer people are afraid of, and they're right to be. Taste is real, but it's not an argument, and a town is too big a thing to hang on one person's preferences. (More on that in the next post.)

What we say we want. The most people wanted it, or the loudest did. That's the public meeting, and we've already seen what it builds.

What actually works. Something outside both of us. Something you could check. That's the answer I want to make a case for, and it's the one Apple actually used, even if the focus-group line makes it sound like the first.

How Apple actually listened

Apple didn't skip listening. It skipped asking.

Those are different things. Asking is a survey, a show of hands, a comment card. Listening is watching what people actually do with a thing, over and over, until you understand it better than they could have told you.

The people who designed the iPhone didn't poll anyone about what a phone should be. They watched people fumble with the phones they had. They watched themselves. They noticed which pockets things went in, which buttons nobody pressed, and how long a person will tolerate a stylus before losing it. Then they built something, held it, and paid attention to their own hands.

Build, hold, watch, change. Repeat until it's right, then repeat some more.

They even did this for the store. Around 2000, when Apple decided to open its own shops, Steve and Ron Johnson didn't sketch a store. They built one. A full-size prototype, in a warehouse near the office, with the real fixtures and the real lighting, so they could walk around in it.

They walked around in it for months. Then Johnson realized the whole thing was wrong. It was laid out by product, the way a computer company thinks, and not by what people were trying to do with their lives. Steve's response, more or less, was: well, we have to start over.

So they started over. Built the whole thing again. Delayed the launch. And then opened stores that, by most published estimates, made more money per square foot than any retailer in America. (Which was not their point, but it is data.)

Now think about how we prototype a neighborhood people will live in every day for a hundred and fifty years.

Mostly, we don't. We draw it, render it with a flattering sky, show it at the meeting, and build it once, at full scale, for good. Apple built a store twice before it would let a customer in. Most towns get built once, and the residents find out.

I say "mostly" because I've seen the other way. On the town I worked on, we built the main street in CAD and walked it in VR before anything was poured, and it caught things: it needed to feel narrower than it was, the center needed something to make it an unofficial meeting point, a corner that needed a rooftop deck, dimensions that looked fine on plan and felt off on foot. None of that would have been fixable a year later.

Obviously, nobody can build a town twice. Nobody can afford even the warehouse. But if we can't prototype the way Apple did, we'd better be very good at the other ways of finding out.

Now think about how we prototype a neighborhood, which people will live in for a hundred and fifty years.

Obviously, nobody can build a town twice. Nobody can afford even the warehouse. That's true, and it's exactly why the rest of this post matters. If we can't prototype the way Apple did, we'd better be very good at the other ways of finding out.

Preference and pattern

There's a second thing Apple's designers were leaning on, besides watching. They knew things. Decades of knowledge about how hands work, how eyes track, what a hinge should feel like. None of that was opinion. It was the accumulation of thousands of things that had been tried.

Places have that too. We just forget it's there.

The architect Christopher Alexander spent his life writing it down. His argument, stripped to the bone, was this: human beings have been living in places for a very long time, and certain arrangements reliably support life better than others. A room with light from two sides. A square small enough that you can recognize a face across it. A front porch deep enough to sit on. A path that passes something worth passing.

These aren't merely someone's preferences. They're solutions people have arrived at again and again, in cultures that never met, because they work on the creature we are. The best patterns work with physics, climate, culture, and context. They are not law; they're evidence.

And “works” still requires us to say what we're working toward. A highway and a main street can both work perfectly and do opposite things. That's why the first question from the last essay—what should this be?—still matters.

He called them patterns, and he collected a few hundred of them, and a great deal of what makes an old town feel good is that somebody, over centuries, followed them whether or not they knew they were following anything.

This is the distinction I most want a developer to take from this post:

Preference is "what do you like?" Pattern is "what helps people live here?"

The first one is a matter of taste, and taste can't be argued with, which is why it's such a bad foundation for a town. The second one is a matter of observation, and observation can be checked. You can walk to the square and count the faces. You can stand in the room and feel the light.

When a good designer says "this is good," that's what she's saying. Not "I like this." She's saying: I've seen this work, many times, in many places, and I've watched it fail when it was done the other way. That's expertise. It's not the same as having opinions, and it's not the same as a vote.

This last month, I've been doing resident meetings for a condo renovation coming to a vote. Sixty or so people in a room, some who love the proposal and some who'd rather nothing changed. At one point, someone said she didn't like the black-and-white walls I proposed, and I saw heads nodding. So I asked for a show of hands: who's worried about painting the entry black? Maybe seven hands. I asked why. One woman said it felt trendy. Another resident, incredulous, said black and white is about the most timeless palette there is. A third said, honestly: I just don't like it.

All three were telling the truth. None of it was an argument, including the one on my side. And tensions were rising.

So I said what I actually think: it isn't about anyone's preferences, mine included. I'm looking at wayfinding. At giving the space a hierarchy so it reads clearly, because clarity is calming. At what the market will reward when you might go to sell. At how to get the most bang for the buck and still have it look good in fifteen years.

Those are things we can check together.

It shifted the energy in the room. It stopped being what I like versus what you like versus what she likes, and started feeling like sixty people working on the same problem. Which we were.

Listening differently

So I'm not saying don't listen to people. I'm saying stop outsourcing the design to a show of hands, and listen the way Apple listened.

Residents are experts in their own lives. Designers are experts in turning needs into form and function.

Good engagement needs both kinds of expertise, and it goes wrong when either one tries to do the other's job.

Here's roughly what it looks like when it's working.

Watch what people do, not only what they say. Where do they actually cut across the grass? Where do they stop and talk? Which bench gets sat on and which one is just furniture? Which corner do kids gravitate to and which do they avoid? A neighborhood is already showing you how it works.

Ask about needs, not solutions. "More parking" is a solution. The need under it is usually something like "I don't want to circle for ten minutes with groceries" or "I'm worried about strangers on my street." Those are real, and people are excellent witnesses to their own experience. A designer can solve those needs a dozen ways, most of which aren't a parking lot. But you only get to the need if you stop taking the solution as the request.

Go stand in places that already work. This is the closest thing we have to Apple's warehouse, and in one way it's better, because the experiment has been running for three hundred years. Every old town people love is a full-scale prototype that somebody built, altered, and lived in long enough to refine. Stand in it. Count the faces across the square. Measure the street, and then notice that the fire trucks got there somehow. The patterns are the results. We don't have to run the test again. We have to read it.

Model what hasn't been built. We can't build a town twice, but the 3D modeling and VR available now would have been science fiction to the people who wrote our codes. We did this on Main Street. It works. Almost nobody does it.

Mock up the parts you can. You can't prototype a town, but you can narrow a street with planters for a month and see what happens to speeds and to the sidewalk cafe. You can tape out the plaza and have your team stand on the other side. You can put the market in a tent before you put it in a building. The pieces are cheap. The whole is expensive. Test the pieces.

Refine it. Nobody gets it right on the first try. Jony Ive famously had a hundred prototypes of everything. Give it the time it deserves to get it right.

None of that is anti-public. It's the opposite. It takes people more seriously than a comment card does, because it's paying attention to how they actually live, not just to what they said at seven p.m. on a Tuesday after a long day, with a microphone in their face.

The meeting is still worth holding. It's just not where the design comes from. It's where you find out what you missed.

The back of the fence

Steve Jobs told a story about building a fence with his father. His dad insisted the back—the side nobody would see—be finished as carefully as the front. Why? Because you'll know.

That instinct ran through Apple: the inside of the machine, the packaging you'd throw away, the underside of the trackpad. Nobody in a survey says, "I'd really like you to obsess over the underside of the trackpad." And yet everyone can feel it when someone did, because a thing made that way holds together in a way you can't quite name.

Places have a back of the fence, and it's most of the place.

It's the soil the trees went into, and whether they'll be alive in twenty years. It's the drainage nobody photographs. It's whether a town with a normal budget can actually maintain the street. It's the side of the building facing the alley. It's the hundred decisions that never make the rendering and that decide, almost entirely, whether the place is still beautiful when your kids are grown.

Nobody at the meeting will ask for most of it. They shouldn't have to. It's our job to know.

Not because people are wrong about what they want. Because much of what makes a place last is something you'd never think to ask for. You just live inside the result, and you know.

Who informs the vision

So, back to the question. If the designer shouldn't just do what the room asks, where do her answers come from?

Not from what she likes. Not from what we say we want. From what actually works, as best anyone can find out. From watching how people live, from the patterns that have held up for a few thousand years, from standing in places that already ran the experiment and modeling the one that hasn't, and from caring about the parts nobody will ever ask about.

That's what Apple was doing behind the arrogant line. It wasn't "we know better than you." It was "we've looked harder than a survey can."

Put it all together, and it's not "trust me," and it's not "what does everybody want?" It's something more like: look harder. Know the craft. Make a proposition. Test it against reality. Let people show you what you missed. Refine it.

And here's why this matters for the next post, which is the uncomfortable one. If a designer's answers come from her preferences, there's no way to check them, and giving her authority over a town is exactly as dangerous as it sounds. But if her answers come from something outside her, from patterns you can go stand in and models you can walk through, then they can be checked. Argued with. Corrected by the people who'll live there.

That's what makes holding the whole something more defensible than “trust me.”

But what if she mistakes her preferences for patterns? What if she reads the evidence selectively? What if the person holding the whole has a blind spot, and nobody around her has enough standing to say so?

That's the next essay.

Next: What Apple got wrong, and what it costs a town when one person's worldview gets set in cement.