Love at First Sketch

TL;DR: Design fixation makes your first idea feel better than it is. Once that first sketch is on the page, most of your “exploration” becomes small variations on it instead of real alternatives.

You think you are exploring. You have filled half a sketchpad with different layouts, different flows, maybe one strange direction you would never show anyone. But you probably drew something in the opening twenty minutes, and every draft since then has been arguing with that one. You moved a button. You changed the hierarchy. You swapped the nav. That early draft is still sitting under all of it.

I know this move because I still catch myself doing it. You think you are still open, and then you catch yourself protecting the first route that felt right. That is why a sketch session can look busy while staying narrow. You get motion, polish, cleanup, rearranging, and refinement. What you do not get is distance, surprise, doubt, or a fresh angle.

That is how a workshop turns into orbit. You get scribbles, arrows, captions, reshuffling, tidy little upgrades, and a lot of movement on paper. What disappears is the awkward detour or the draft that breaks the frame open.

The first draft takes over

In 1991, David Jansson and Steven Smith ran a series of experiments with engineering students and working engineers. They gave both groups design problems: design a measuring cup for the blind, design a bike rack. Some people saw a sample solution first. The researchers told them it was flawed and pointed out its weaknesses in plain terms. Then everyone had to come up with as many new solutions as they could.

The engineers who had seen the flawed sample kept reproducing its features in their own designs. They were not copying it line for line. But they kept reusing a similar shape and the same hidden assumptions about how the object should work. Meanwhile, the engineers who had seen nothing produced far more varied concepts. They defined what they found as

Blind adherence to a set of ideas limiting the output of conceptual design.

— Jansson & Smith

The engineers were not blind to the problem. They had been told the example was bad. Still, they could not escape it. Knowing the trap was there did not help.

Design fixation does not only happen when someone shows you a flawed sample. It starts the moment you get a rough thought down. Then that reference belongs to you, which makes it worse. I know that feeling. You like it. You trust it. The later drawing looks new, but it keeps leaning back toward the draft you drew first.

The Einstellung effect describes a broader version of this. Once your brain learns one way to solve a problem, it keeps reaching for that path even when a better route is sitting right there. The old pattern does not just shape your thinking. Sometimes it crowds out every other option before it even has a chance.

Zune chased the iPod

In 2006, Microsoft launched the Zune to compete with the iPod. They had money, engineers, a clear goal, and a product team that had studied Apple’s player for years. What they did not have was much distance from the device they were chasing.

The Zune came out as a straight answer to the Apple iPod. Scroll-wheel navigation. Music-first interface. Desktop sync software. Physical buttons in almost identical places doing near-identical jobs. Even the hardware felt close, just with a different finish. Apple was in the room the whole time. Microsoft kept asking one question: how do we do what the iPod does, but better?

That question narrowed everything that came after. Once they started there, it was hard to ask a different one, because every comparison kept dragging them back to the same frame.

People already had iPods. They also had iTunes libraries, habits, and accessories. To switch to a Zune they needed a reason that went beyond “it is a bit better at comparable things.” They needed a reason to leave the whole setup they already knew. Microsoft never gave them one because it never stepped back far enough to ask what another response might be. The Zune’s market share stayed below Apple throughout its run, and even trailed SanDisk. Microsoft pulled the hardware from the market in October 2011.

The trap was set early. When you define your product as “a response to X,” you let X shape every choice that comes after. I have seen designers do this with competitors, old versions of their own product, and even a rough sketch from the first hour.

Your sketch already made choices

Every sketch you do at the start makes a choice you did not know you were making. Maybe someone taps a button to proceed. Or maybe you already assumed this belongs on mobile. You may also be assuming a certain kind of user or a certain mood. You did not sit down and choose those things. You drew the version that came to mind first, and now everything else sits on top of it.

Some early concepts are good. But you still cannot tell what kind of concept you have until you have seen something else. Fixation gets in the way. You are not testing that opening move against real alternatives. You are testing it against softer variations of itself.

A blank page scares people, but orbiting one early sketch is worse. It narrows the frame, trims off odd options, and makes safer variations feel like progress. A login button moves from left to right and designers start calling that a new direction.

Real alternatives often look awkward at first. They can be lopsided, noisy, bare, slow, oversized, cramped, upside down, or a little silly. But that ugly draft may expose a shortcut or a layout rule the polished version kept hidden. I have seen a bad paper sketch do more work than a neat screen full of polite variations.

Chrysikou and Weisberg’s 2005 study found that even when participants were told that prior examples were irrelevant to their task, those examples still shaped what they produced. Telling people to ignore an example does not neutralize it. The image is already in.

Check this before you iterate

Before you iterate, take your sketch and name the single assumption it takes for granted. Just one. Maybe a person starts the action. Or all this information lives on one screen. Or there is a navigation structure at all. Name it. Then do one rough draft as if that assumption were false. Not a variation. A contradiction. What does the interaction become if nobody starts anything? What if there is no screen? What if navigation disappears?

This is a way to break your own pattern. Force yourself to design the opposite of what you just made. It might be useless. That is fine. I only care that it shakes the early answer loose.

Most designers never run this check. They have one concept and a dozen versions of it, and they call that exploration. I still do that when I move too fast. The study found fixation in novices and in working professionals in matching proportions. Experience does not protect you from it. In some cases it makes it worse, because the more problems you have solved one way, the faster your brain grabs that route again.

The assumption break shows you what your sketch is actually made of. You see which constraints it accepts without question and which options it shuts down before you even notice them. After that, you can decide what stays. Without that step, the choice was probably made in the opening twenty minutes, and the rest just looked busy.

Your second answer usually shows up only after you stop protecting the earlier one.