Fips Mobility UX Case Study | Francis Haupolter

In short

Problem

Existing usage data showed it: 8 out of 10 seniors stopped using the on-demand bus after their first trip.

What I did

10 interviews, COM-B diagnosis, choice architecture and a new web booking flow designed for seniors on a computer.

Outcome

Root cause found (time pressure + no trigger at home), two high-impact interventions with measurable KPIs and a hi-fi web booking flow designed for seniors.

My role in this project

What I owned

Research, interviews and behavioral analysis
Wireframes and hi-fi prototype
Testing the proposed solution

With stakeholders

Design reviews with managers and stakeholders during prototyping
Adapted the design to their feedback, business requirements and brand guidelines

Given at the start

Existing usage data: 8 out of 10 seniors stopped after their first trip. This was the problem I was asked to solve.

01 · Overview

Seniors tried it once and never came back.

Fips Mobility offers affordable, on-demand transport for seniors (65+) living in the German countryside. Despite strong initial interest, seniors were dropping off after just one trip, leading to service shutdowns.

The goal was simple: find out why seniors were leaving and design a seamless web version that makes booking completely effortless. Together with the project team, we focused on one business goal, retention, and on the 70% of seniors who prefer using a computer over a phone.

80%

of seniors stopped using the shuttle after their very first trip

61%

of people 65+ find AI-powered apps too complicated (Bitkom)

53%

of people over 65 don’t own a smartphone at all (Bitkom, 2021)

02 · Target behavior

Turning a business problem into a measurable action.

I defined the target behavior so success could be observed and measured, not guessed.

Actor (as is): an inactive senior after their 1st trip
Context & device: a computer at home
Target action: clicks the final reservation button
Behavioral KPI: repeat bookers ÷ first-time users × 100

03 · Research

10 interviews, one clear pattern.

After secondary research on how seniors interact with digital products, I ran remote interviews with 10 seniors (65+) from the Rhein-Neckar region who had used Fips once and stopped. A familiar person helped each of them join the call. Questions were open, neutral and focused on real past experiences. We stopped once answers started repeating.

9/10

had at least one bad experience

7/10

needed family help to register

8/10

stopped after the first or second trip

Fips Mobility interview data clustering and findings — Interview data clustering and key findings.

Proto persona Annemarie Bauer, 75, retired teacher.

04 · Diagnosis

Mapping the friction with the COM-B model.

Mapping the as-is app flow revealed three problems: seniors had only one minute to choose an offer, the service area wasn’t marked on the map, and an invalid address was only flagged at the very end.

Capability, cognitive overload: a 1-minute booking timer forces fast decisions and causes acute stress.
Opportunity, no desktop trigger: nothing at home reminds seniors to book again.
Motivation: no friction found, they wanted to ride.

Fips app user flow as is with pain points — As-is app flow: 1-minute timer, no service area on the map, late address errors.

Behavioral diagnosis with the COM-B model.

05 · Choice architecture

Two quick wins with the highest impact and lowest effort.

Stress reduction

Replace the 1-minute timer with a 15-minute price & seat guarantee: “Your seat is reserved. Take your time!”

Hypothesis: removing time pressure reduces cognitive overload, so checkout drop-offs decrease.

Timely reminder

An automated email 3 days before the regular trip day with a one-click rebook link.

Hypothesis: an external trigger in the home decision context increases repeat bookings.

Behavioral intervention impact effort matrix — Impact / effort matrix to pick quick wins.

Choice architecture interventions for Fips Mobility — The two interventions: 15-minute guarantee and one-click rebooking email.

06 · Experiment design

How we’ll know it works.

Stress reduction: usability test. Primary KPI: checkout completion rate. Secondary: time on page.
Timely reminder: scenario test. Primary KPI: 7-day desktop rebooking rate. Secondary: link click-through rate.

Status: I tested the proposed solution with users. The two KPIs are hypotheses that can only be measured once the new flow is live.

07 · Information architecture & wireframes

A calm, step-by-step booking flow for the desktop.

Fips web sitemap and to-be user flow — Sitemap and to-be user flow, based on a benchmark of booking platforms.

Fips lo-fi wireframe find a ride — Lo-fi: find a ride

Fips lo-fi wireframe pick date and time — Lo-fi: pick date & time

Fips lo-fi wireframe registration — Lo-fi: registration

08 · Final design

A calm booking flow for seniors at home.

The hi-fi web version turns the research into the interface: large type and touch targets, one decision per step, and the two interventions built right into the flow.

Fips web booking homepage header with booking form — Home: one clear question, “Wo möchten Sie hinfahren?”, the whole booking form right away, and below it booking explained in just 3 steps.

Fips pickup address field with saved home and recently used addresses — 1 · Pickup: “Mein Zuhause” and recently used addresses, so repeat trips take one click.

Fips destination field with address suggestions — 2 · Destination: suggestions while typing, plus a fallback to search on the map.

Fips passenger panel with large plus and minus buttons — 3 · Passengers: large plus and minus buttons with clear age groups.

Fips date and time picker with large buttons — 4 · Date & time: “Heute / Morgen / Später”, large day fields, past days clearly marked and times as big buttons.

Fips connection choice with 15-minute seat and price reservation — 5 · Connection: the 15-minute seat & price guarantee replaces the 1-minute timer, and the best option is preselected.

09 · UI decisions

Designed for seniors at a computer. Every detail has a reason.

I built a small design system for the booking flow. These are the decisions behind it.

20 px

Smallest text

Body 20 px, buttons 22 px, in Lexend, a typeface made for readability.

64 px

Buttons & fields

Large click areas: for radio buttons the whole row is clickable.

7.5 : 1

Minimum text contrast

All text colors meet WCAG AAA and are documented in the design system.

Navy text on orange, not white

White on our orange would only reach 2.4 : 1. Navy text reaches 7.5 : 1. Every button has normal, hover, focus, disabled and loading states.

Contrast is documented, not guessed

All colors are variables, and every text-on-background pair is listed with its contrast ratio.

Addresses with house numbers

Suggestions are full 64 px rows including the house number, so there is no extra step. “My home” and recent addresses come first.

Limits explained in words

56 px plus and minus buttons, 28 px numbers and a sentence instead of a silently greyed-out button.

Tiles instead of checkboxes

Walker, wheelchair or luggage as large tiles, easier to hit and to understand than small checkboxes.

Big time slots, no dropdown

Times as 128 × 64 px cells: one click instead of scrolling through a list.

A calm reservation without a countdown

“Your seat and price are reserved for 15 minutes. Take your time.” No timer and no red, because time pressure was a key reason seniors gave up.