Hannah Keller

Hello, I am Hannah Keller,a product designer who writes.

Recently graduated from the human computer interaction program at the Limmat School of Design.

I design journey planners, checkout flows and claim forms, then write down what I got wrong.

Selected works, written out in full

Four projects from the last two years, each told as a short essay.

Why riders skip the fastest route

Ranking trips by transfers as well as time cut abandoned searches by 21%.

Seeblick Mobility's journey planner always put the quickest trip first, yet riders kept scrolling past it. Over six weeks I sat in on 14 commutes, read 900 support tickets and ran a diary study with regular tram and bus riders. The pattern was plain: a trip with one change and four extra minutes felt safer than a tight connection that might be missed. We redesigned the results list to show transfers, walking distance and how much slack each connection had. The ranking now weighs all three.

Context
Seeblick Mobility
Role
Research and interaction design
Year
2025

Designing for the bottom shelf

Screen reader checkout completion rose from 54% to 88%.

Kornfeld Grocers asked why older customers abandoned their delivery app halfway through checkout. I tested the flow with eleven people who use screen magnifiers or screen readers and found the problem was not one big barrier but a stack of small ones: unlabeled quantity steppers, a delivery slot grid that read as forty unrelated buttons and a total that moved every time an item changed. We fixed the order of focus, grouped slots by day and kept the total in one place.

Context
Kornfeld Grocers
Role
Accessibility research and design
Year
2024

Where the night shift loses the thread

A one page handoff sheet is now in a pilot on two hospital wards.

My master's thesis followed nurses through 22 shift changes on a surgical ward to see where information about patients went missing. Most of it was not lost in the software at all but in the four minutes between a verbal handoff and the first round of the night. I mapped every note, sticky label and whiteboard entry used during those minutes, then designed a printed sheet and a matching screen that follow the order nurses already speak in.

Context
Limmat School of Design
Role
Field research and thesis
Year
2024

Writing the empty states first

Calls about claim status fell by a quarter within three months.

Talbach Insurance's claims app looked finished in every mockup because every mockup was full. Real customers mostly saw blank screens while a claim waited for review. I started the redesign from those waiting moments instead, writing what each empty screen should say about what happens next, who is looking at the claim and when to expect news. The visual design came afterward and stayed simple.

Context
Talbach Insurance
Role
Content design and interaction design
Year
2025

One piece in full

Why riders skip the fastest route

A journey planner can be correct and still feel wrong. This is what six weeks of riding along with commuters taught our team about ranking trips.

The brief from Seeblick Mobility was narrow: people searched for a trip, looked at the results and then left without choosing one. Analytics showed the drop happened on the results list, not in search. The team's first guess was speed, so the first fixes made the list load faster. Abandonment did not move.

So I went riding. Over six weeks I joined 14 people on their usual commutes and asked them to think out loud while they planned. Nearly all of them skipped the top result when it had a short connection. One rider put it simply: four minutes saved is not worth the chance of standing on a cold platform for twenty. Support tickets told the same story in angrier words. What the app called the best trip was, to riders, a gamble.

We stopped treating time as the only score. The new results list shows the number of transfers, the walk between platforms and how much slack each connection has. Slack is drawn as a small bar under every trip. The ranking now weighs those three alongside travel time, and riders can still sort by speed alone if they want to. We shipped it to a tenth of users first and watched for three weeks before rolling it out.

What we measured
Searches that ended without a chosen trip fell by 21%, and the share of riders picking the top result rose from 38% to 61%.
What surprised us
Riders read the slack bar faster than any number. Several said it was the first time the app felt like it understood a missed tram.
What I would redo
I would bring an engineer on the ride alongs from day one. The ranking change took two sprints to agree on because only I had seen the platforms.

Back to the list of works

About the person writing this

I grew up in Winterthur and studied German literature before I ever opened a design tool. Reading closely and writing clearly turned out to be most of the job, so I kept doing both when I switched to design.

I finished my MS in human computer interaction in 2024 with a thesis on hospital shift handoffs, and since then I have worked as a product designer at Seeblick Mobility on journey planning and ticketing. I also take on a small number of accessibility reviews each year.

I write a short essay after every project, mostly about what did not work. Some of them end up as talks at local meetups, and a few have been used in design courses at my old school.

Currently
Product designer at Seeblick Mobility
Studied
MS in human computer interaction, Limmat School of Design
Works in
Field research, interface design, writing
Based in
Zurich, Switzerland

Notes and talks

Write to me

I am glad to hear about research questions, accessibility reviews and teams that want a designer who will also write things down. I usually reply within a few days.

hannah@example.com
The EssayBuild this oneAll templates