Why first-time riders
give up on the
NYC Ferry app
Four of five first-time riders couldn't get the NYC Ferry app to tell them whether the ferry reached their destination, so I traced the three places the app lost them and prototyped a fix for each one.
Study results · 5 participants
4 of 5
couldn't map a route, and all four said they'd switch to another transit option
5 of 5
found the home screen overwhelming or confusing on first open
3 of 5
couldn't confirm they'd found the right terminal, even after finishing the task
- Role
- Sole UX researcher, self-initiated
- Methods
- Generative pre-task interviews
Think-aloud usability testing
Emotional journey mapping - Participants
- 5 people, ages 25 to 35, little or no ferry experience
About 30 minutes each - Tools
- FigJam for coding and clustering
AI-assisted first-pass clustering
Figma for the prototype - Timeline
- About 2 weeks, September 2025
01 · Starting point
Can the ferry
get me there?
Starting point
The NYC Ferry reaches neighborhoods the subway serves poorly, like Red Hook and Greenpoint, so a new rider's first question for the app is whether the ferry can get them where they're going. The app has that answer, but when I tried planning trips with it, the answer sat several screens deep or behind features I had trouble finding.
Why it matters
If a rider can't find the route, they assume there isn't one. In testing, participants who couldn't dig the answer out decided the ferry didn't go where they needed and picked other transit, even though the route was in the app the whole time.
Research question
Can a first-time rider find what they need in the app (whether the ferry reaches their destination, how to buy a ticket, and when the next boat leaves) without already knowing where to look?
02 · Approach
Ask what riders expect,
then watch the app answer
Sample
Five people aged 25 to 35 with little or no knowledge of the ferry network, because the question was whether the app works for someone new to it.
Why this method
Each session opened with generative questions before anyone touched the app, which captured what they expected before the app could shape it. Think-aloud testing then showed where those expectations and the app's answers came apart.
Analysis
I coded recordings and notes by task step, tagged each moment as success, hesitation, or failure, clustered the tags into themes in FigJam, and counted how many participants hit each one. AI helped with first-pass clustering; I checked every theme against the raw sessions.
Journey map
After each session I charted the action, the reasoning, and the emotion at every step. All five arcs followed the same shape, so I merged them into one composite map.
Core tasks tested
01
Find a route from the Upper West Side to IKEA
Route discovery, map orientation, and choosing a stop. Participants named routing as the main thing they wanted from the app before the test began.
02
Buy a ticket for the trip
The ticketing flow, from picking an option to finishing the purchase, which is where a curious person becomes a rider.
03
Find departure times
Schedule navigation and whether terminal names make sense to someone who doesn't already know them.
03 · Who it's for
Meet Rachel,
the rider the app is losing
Persona
Rachel is a 29-year-old social copywriter in Manhattan who gets around the city by subway and takes the ferry now and then, mostly for the views or to reach places like Red Hook and Greenpoint the subway doesn't serve well. She's built from what the five participants told me before the test about why they'd choose the ferry and what they'd want the app to do.
Job to be done
Find out quickly whether the ferry can get her to a specific place, so she can reach parts of the city the subway can't without piecing a route together across several apps.
Uses today
Google Maps, Google Search, and the MTA subway app.
Goals
Know whether the ferry can get her somewhere before committing to it
Reach neighborhoods like Red Hook and Greenpoint the subway doesn't serve well
Buy tickets in the app instead of at the terminal machine
Frustrations
The app can't tell her whether the ferry gets her there, and the Apple Maps handoff shows no ferry route
Terminal names like St. George and Stuyvesant Cove don't match any neighborhood she knows
The live map home screen is hard to read on first open
Session evidence
Six screens where the app lost people
Screenshots from live sessions, each one a moment where a participant stopped trusting the app to get them anywhere. Hover each to see what happened.
Hover to expand
Cluttered home screen
Task 1 · Map a route
What they were trying to do
Find out whether the ferry could get them to IKEA in Red Hook, starting from a clear place to enter where they were and where they were going.
What happened
The home screen showed a live map, a search bar, recent locations, and two buttons at once, and all five participants said they didn't know where to begin. Those who tried the search bar couldn't tell whether it was for routing or tickets.
Hover to expand
Schedules list
Task 3 · Find departure times
What they were trying to do
Find the next departure from a specific pier, expecting a schedule organized by route or stop they could scan quickly.
What happened
"Schedules" opened a flat list of all 22 stops sorted by distance, with no routes, no filters, and nothing showing which stops connect. Most participants scanned it and left within seconds.
Hover to expand
No trips found, attempt 1
Task 1 · Map a route
What they were trying to do
Confirm the ferry could get them from where they stood to IKEA USA Brooklyn 2, the place they wanted to go, which isn't the name of any terminal.
What happened
"We couldn't find any trips in that area," with no reason, no nearby stop suggested, and no sign a ferry route to the area existed. The only way forward was to leave the app.
Hover to expand
No trips found, attempt 2
Task 1 · Map a route
What they were trying to do
Route from Lincoln Center to the same place, to see whether a different starting point changed the answer.
What happened
The same empty state. The app only accepts a ferry terminal as a destination and never says so, and both participants concluded the ferry didn't serve IKEA when it does.
Hover to expand
Walking directions only
Task 1 · Map a route
What they were trying to do
Get to a ferry stop from their current location using the app's built-in routing.
What happened
The app offered walking directions to the stop and nothing else, no bus or subway. Anyone who wanted a faster way to the ferry had to leave the app to find one.
Hover to expand
No ferry in Apple Maps
Task 1 · Map a route, after leaving the app
What they were trying to do
After "No routes," participants followed the app's suggestion to open Apple Maps and switched to transit mode.
What happened
Apple Maps showed only subway options, even though a ferry route exists for this exact trip. Participants assumed the ferry wasn't running that day or didn't go there.
04 · Findings
Friction clustered at the same three moments
"At this point, I think I would give up and then find the answers on Google because I am feeling upset and defeated."
P1, route mapping task
01
Route mapping breaks before it begins
Information Architecture · Navigation
The app only accepts a ferry terminal as a destination, while participants typed the place they were going, IKEA in Red Hook. When the search failed, the app sent them to Apple Maps, which didn't show the ferry route either.
Why it matters
Participants read a missing route as missing service, which made this the one failure where losing the task meant giving up on the ferry itself.
"The app is making me do a lot of work in order to get a route."
Prototype change 01
Destination-first routing
Let riders search by where they're going and plan it as one trip: origin to ferry stop, the ferry, then stop to destination. When the app hands off to an outside map, pass the ferry route along so it shows up there too.
02
The home screen buries the most useful information
Content Strategy · Home Screen UX
All five participants called the home screen overwhelming or confusing on first open. The live map only makes sense if you already know the network, and the simpler routes map participants preferred sits in a corner.
Why it matters
New riders need routing help the most and are the least equipped to read the default screen.
"It took me a good number of clicks to find these times. I think the homepage is just so cluttered by the map. And that map is not very clear."
Prototype change 02
Routes map as the home screen
Make the simplified routes map the default, with the live map one tap away. Put a clear "Plan a Trip" entry point up front, and quiet the search bar, which left participants unsure what to type.
03
Terminal names only work if you already know them
Content Strategy · Vocabulary Design
Three of five participants couldn't confirm they'd found the right terminal, even after finishing the task. Names like St. George and Stuyvesant Cove were unfamiliar across the group, including to people who had lived in New York for years.
Why it matters
The doubt shows up right before someone buys a ticket or boards, the worst possible moment for it.
"I am not sure if St. George is in Staten Island, so it would be nice if that was included in the stop name."
Prototype change 03
Neighborhood-first terminal labels
Add the neighborhood or borough to every terminal label, like "St. George (Staten Island)" and "Stuyvesant Cove (East Village)." List nearby landmarks or cross streets on each stop's detail view.
05 · Outcome
Three findings,
one clickable prototype
Prototype
A clickable Figma redesign with one change per finding: destination-first routing, the routes map as the home screen, and neighborhood labels on every terminal. Tapping a terminal highlights its routes and connection points.
Deliverables
A usability report with task success and severity by feature area, the composite journey map, the Rachel persona, and the prototype.
Where it stands
This was self-initiated, so nothing has shipped, and the prototype hasn't been tested with riders yet.
Figma Prototype
The redesigned routing and scheduling flows
06 · Reflection
Designing for new
and experienced riders
Why this problem
I hit it myself planning trips, and before the test began, participants named routing as the main thing they wanted from the app.
Why this sample
The question was about new riders, so I recruited people who didn't know the network and asked what they expected before the app could answer for them.
Tradeoffs
Five people over two weeks is enough to find where the app breaks and too few to say how often. I also skipped a standardized score like SUS, which would have made the redesign easier to argue for in a stakeholder meeting.
What surprised me
The pre-task interviews ended up validating the task list, which I hadn't planned. They're also where the terminal-name problem first came up, something I hadn't thought to look for when designing the study.
What I'd change
I'd frame tasks more broadly, something like "using the ferry app, you're trying to get to IKEA," which is closer to how people plan a trip. Then I'd run the prototype through the same tasks with new and current riders to see whether the changes hold up.
Contact Information
Have a research question?Let's talk.
Open to UX research roles and consulting, especially in healthcare and behavioral health tech.
thornton.coen@gmail.comCopy