Users Are Unpredictable Like Water

When unpredictability leads to misalignment between intention and outcome.

By Hans Sandkuhl, eolas – 8 minutes read

It happened again. We launched a beautiful feature that ticked every box, with a clean interface, a clear logic, internally tested, approved and applauded. And yet, three weeks after release, almost no one touched it.

There were no bug reports. No complaint was registered. And no engagement was observed. We tried to explain it away: maybe it was visibility, maybe the onboarding, or maybe timing. Eventually, I returned to something I keep relearning, that people rarely behave the way we expect. And almost never the way we design for.

Why Plans Fall Apart

It always begins with a plan. Blueprints, diagrams, flows and funnels, and carefully labeled zones. The architects step in, the I.T. leads, or the product owners, all with their wireframes and logic. We define entry points, sequences, and what success will look like.

And then people arrive.

They wander, hesitate, take two steps forward and backtrack. They scroll past the feature we spent months perfecting. Sometimes they ignore the main button entirely. Other times they find a workaround we never imagined. Or worse, they do nothing! No clicks. No feedback. Just… drift.

I have seen this happen in architecture, specifically in public parks, daily objects, in digital services, even in internal tools that were supposed to be intuitive. Over time, I stopped thinking of behavior as something we could control. It behaves more like water, flowing where it wants, unpredictable, finding the smallest cracks, ignoring the paths we paved.

How People Reveal What Really Matters

Take the classic example of park design. Have you ever noticed how the paved paths, carefully curved through grass, are often ignored? People cut corners. They trample the lawn. Not because they dislike beauty or order, but because they are simply trying to get somewhere and have their habits. And because the design assumed too much.

These informal trails are often called “desire lines” or “joy lines.” There is something joyful about people asserting their own logic over what was intended. They are not always being efficient. Sometimes they are lazy, distracted, or just following instinct, and even superstitions. Whatever the reason, it reminds us that people do not follow scripts. They respond to what feels immediate, not what was laid out.

My favourite story around desire lines comes from Ohio State University. Back in 1914, University Architect Joseph N. Bradford waited until winter. Once snow had fallen, he used a hot air balloon to get a bird’s-eye view of the criss-crossing trails students had carved into the snow. These trails became the blueprint for the university’s formal walkways.

Sometimes, the crowd sees what the plan could not.

In digital products, these behaviors are harder to detect and often more frustrating. You might launch a dashboard with five carefully designed features. Then you learn that users only touch one. Not because it is better. Just because it is the one they understood, or the one that appeared at the right moment.

We once launched a help system for a product filled with step-by-step guides and video walkthroughs. Not a single user opened them. We assumed the issue was visibility or relevance. Later, after speaking with users, the answer became clearer. They were not asking how to do something. They were still asking why they were doing it. The questions we answered were valid, just delivered at the wrong time.

How I Try to Understand What Design Cannot Predict

There is no single method to really anticipate what people will do. User Personas help, and Journey Mapping adds detail. Analytics leave a breadcrumb trail, but never give the full story. People’s decisions are shaped by need, memory, emotion, habit, and often by things they cannot explain themselves.

So I try to observe more like an anthropologist would. I listen before I track clicks. I sit beside them, ask softly, and let silence work. It sometimes takes hours or days for patterns to emerge. And even then, I hesitate. What I think I have seen, what I believe to be true, can still shift.

One method I return to is what I call “interrupted observation.” You set up a normal task, then make one small change. Move a button. Relabel a feature. Introduce something slightly unfamiliar. Then watch how people respond. Do they search? Do they freeze? Do they blame themselves?

Another technique is shadowing, ideally in their actual environment, not in a lab. Because distractions matter. Interruptions matter. The moments they sigh, or look away, or pick up their phone for something else, those are where the truth slips through.

Then there are the features that never get used. Entire sections of a tool left untouched, the ones we used to call failures. Now, I see them as signals. And not necessarily as bad ones, just clear ones. They tell us that the user’s need was somewhere else.

This is where so many assumptions about user experience fall apart. Good ergonomics, beautiful layout, smart logic, none of it guarantees usefulness. What feels “intuitive” on paper may collapse in practice. Think of those bathrooms where the toilet paper holder is mounted so far to the side that reaching it requires an awkward upper-body twist. It looks tidy, but fails the basic test of usability.

The same happens with touchscreen kiosks that bury simple options beneath marketing choices, or doors that need stickers saying “push” or “pull.” Or software tools so feature-rich they intimidate the very people they were meant to empower. When friction enters unnoticed, logic fails, and people walk their own path.

Digital tools are no different. Features are “logically” grouped, buttons placed with care, flows defined by best practice. Yet people still struggle, because real-world usefulness is shaped not by interface alone, but by need, memory, and context. 

What This Means for Those of Us Who Build

The danger lies in falling in love with structure. With our steps, our flows, our perfect user journey. We overestimate clarity, and we underestimate confusion. And then often ignore the emotional texture of using something that feels too complex or too irrelevant.

A beautifully paved road that leads to the wrong destination will always be ignored. Meanwhile, a bumpy shortcut that saves ten seconds, or just feels easier, will win every time. This reality makes product and service design uncomfortable. It means treating every plan as provisional. Every layout as a hypothesis, and every success as temporary. It asks for humility.

Every product you launch carries hidden assumptions. The sooner you see where those assumptions break, the sooner you can adjust and grow with clarity. I still catch myself slipping into the old mindset, wanting people to behave “correctly.” To admire the structure. To follow my logic. But people are not here to validate our thinking. 

They are here to fix a problem or to move on quickly. There is quiet power in accepting this and not to give up, but to listen better, to adapt, and to reframe. Because behavior is not clean or linear. It zigzags, it resists. And with enough attention, it reveals something honest. Something we never predicted, but now understand.

That, to me, is good design. 

Not controlling the flow. 

Just learning how to read the water.

Now That You See Differently

Understanding how people really behave is only the beginning. To translate that insight into strategy, courses like Introduction to Marketing explore how to position value clearly, while Introduction to Design Thinking helps you shape solutions that meet real needs in unpredictable contexts. These courses offer new lenses, ways to observe, interpret, and act with greater clarity, especially when certainty is out of reach.