Open Seat

Insights

I used thematic analysis and inductive coding to code and identify themes from my user interviews and my notes from my findings of existing solutions. I also created an affinity map to identify more important details and group them accordingly. From this data, I was able to create a manager and driver persona.

Opportunities & Requirements

After my analysis, I identified some key opportunities for my design.

1. A comprehensive filtering system to help narrow down drivers who fit your criteria.
2. A method to convey the human side of the driver beyond simple numerical metrics.
3. A way to highlight reliability across previous results to remove the anxiety around hiring a driver.
4. Users must have a way to contact each other.

"How might we give users the confidence of a personal introduction for drivers they've never met — through verified data, not word-of-mouth?"

Decision #1

Grid and List Layouts To Assess Initial Fit

From the interviews, I found that users vet in two stages: initially, they eliminate based on raw numbers, then assess on human fit. The grid layout supports rapid scanning of non-negotiables — language, time zone, iRating, and car class. The list layout puts biography and human context first. The three-column cap came from what worked in testing.

Decision #2

List View as Default — Slower, but Users Made Better and More Informed Decisions

Both layouts were tested with all five participants. The grid was faster — participants said they'd only use it for last-minute emergencies when any available driver beats a forfeit. For considered team building, the list took longer, but every participant felt more confident in their final choice after reading bios, before even opening a full profile.

Decision #3

The Profile Layout Follows Manager Logic: Identity & Endorsements → Performance → Discovery

Interviews kept surfacing the same manager logic: who is this person, how have they performed, who else is worth a look. The profile follows that sequence — bio and endorsements first, performance data second, similar drivers last. Community-verified quotes appear above the fold before any numbers. The four metrics — Race Pace, Qualifying Pace, Consistency, Tyre Management — came directly from interviews. Those were what participants named when asked what separates a fast driver from a good teammate.

Decision #4

Reliability Gets Its Own Dashboard, Above Raw Pace

The dashboard combines iRating, average finish, safety rating, win rate, and incidents per corner. The focus is who you'd actually want as a teammate, which typically isn't a person who can post one fast lap but crash on every other.

Decision #5

Strength of Field Score Adds Context to Every Race Result

iRating says nothing about the quality of competition a driver actually faced. Showing Strength of Field — the average iRating of all competitors — alongside every result gives each finish meaning. A hard-fought P5 in a high-SoF race ranks above a comfortable win against weak opposition, which directly addresses Insight #3.

Decision #6

Promoting Discovery of Similar Drivers

I wasn't sure how many similar drivers to show, but in usability testing with five participants, five was the limit. They proved that instead of exploring, they scanned back and forth when the count was above five.

Decision #7

Driver Event Interest

Drivers can add their event interest at the bottom of their profile page — a quick reference for other drivers and managers checking requirements. By this point, unsuitable drivers have likely been filtered out during the search phase already, so this is more of a confirmation than a discovery tool. In testing, users responded well to the notes field, valuing how it let them pass on a driver without needing to message them first if non-negotiables were present in the notes.

Decision #8

One-Click Seat Offer To Streamline Communication

Once a manager confirms a driver meets their criteria, messaging can optionally attach a Team Card with combined roster stats — a structured seat offer in a single click, replacing the usual chain of DMs and voice calls. Users can access chats and messages any time via the floating messaging tray at the bottom of any page.

Initial Scrapped Designs: Lo-Fi Testing

First, the filtering layout failed. A LinkedIn-style filter bar fixed to the top frustrated participants — they had to scroll back up to refine searches after scanning results. A sticky side panel updating results in real time fixed it.

Second, the grid-only layout hid the most important information. Participants had to open a profile just to read a bio — which led to the dual-view system, with list as the default and grid available for rapid scanning.