Architecture Tours, through the lens of UX Research

What started as a personal project became an opportunity to practice UX Research methods.

How It All Started

In spring 2023, I took a step away from my PM job at Microsoft. As my free time expanded, I found myself taking long walks around Seattle and snapping photos of interesting buildings I noticed along the way. When I returned home from my walks, I began searching for more details on the buildings I had noticed. I started with Google Maps and their street view photos, and eventually stumbled upon the Pacific Coast Architectural Database (PCAD) and a social media account called Buildings of Seattle. Before I knew it, I had built a mental map with hundreds of buildings and the stories behind them.

One day, I mapped a route to look at a few dozen notable buildings which had held my interest. The route to see all of the buildings was 11 miles, much farther than I wanted to walk in the Seattle rain! I planned to ride the route on my bicycle. While sharing this plan with a friend, they asked if they could join.

And that’s when the bicycle tour was born.

V1: Starting Small

Route map for the V1 Tour. Each “i” is a stop to view a building or structure.

The first architecture tour had two attendees: myself, and my friend. I stuck to the 11-mile, 24-building route that was optimized to minimize distance. At the time, I thought of the tour as a collection of buildings, and I didn’t have a heuristic for the order or priority of the buildings.

The tour felt like a slog. At each stop, I had to pull out my notes to rattle off facts and details about the building. After 4 hours and 19 buildings, we both turned home, even though there were more buildings to see.

But while my eyes were focused on my notes, my friend was able to focus on the buildings. They began noticing themes and motifs that were common between buildings.

I received a lot of interest from friends and social media to lead another tour. Before I scheduled one, I conducted some small-scale research to improve the user experience. If I was going to have an audience, I wanted the tour to be good!

1. I developed user personas — except I called them rider personas. I knew different styles of bicyclists would show up and I wanted to ensure the tour met their needs. See Appendix for table of Rider Personas
2. I conducted an adaptive interview with my friend regarding their experience on the route — also known as getting coffee and asking for their input. We reviewed a map of the route to refresh their memory of the experience, and we discussed the following topics:
   - route and ride experience
   - stops and points of interest
   - distance and duration
   - narrative and story

Based on my friend’s feedback and the needs of the rider personas, I made the following changes to the tour:

1. I modified the route to start and end at a bicycle rental shop. This added a 3-mile segment on a protected bike path at the beginning of the tour, which gave me a chance to assess the riding comfort and speed of the group before we began navigating the hills and neighborhood streets.
2. I reduced the number of stops, in order to reduce the tour duration and preserve attendee focus. 9 buildings were converted to “fly-by” stops, which I would call out as we passed (noted with a camera icon on the map).
3. I created a storyline to frame the tour, centering around an innovative Seattle architect, Robert Reichert. The tour was bookended with Reichert buildings and I called out which buildings were designed by his contemporaries.

V2: Expanding the Audience

Route map for V2 of the tour. Each “i” represents a stop, and each Camera icon represents a “fly by”.

I posted the tour on a local bicyclist forum and invited riders to join the tour on three separate occasions.

During the tour, I focused on getting behavioral and attitudinal insights from attendees. In order to gather data of diverse types, I used contextual inquiry and direct observation to gather insights. This meant chatting with riders during the ride, and observing the dynamics of the group while riding with everyone.

1. A tour group of many strangers is much less connected than other social bike rides, and riders are less likely to wait up for each other. Riders became more scattered along the riding portions of the tour, especially uphills.
2. Riders consistently lose engagement around the 2 hour mark. Starting and ending the tour near a bicycle rental shop adds additional time to the tour and further reduces the amount of time for engaged tour attendees.

Based on these insights, I knew further revisions were needed. I made the following changes:

1. I struck the northernmost cluster of stops from the route. This was a hard decision to make, because it meant removing the final stop, a private residence designed by Robert Reichert. However, I knew the extra miles and steep hills to get to the home weren’t worth the loss in rider interest.
2. I reversed the route, for a variety of reasons. First, the reversal moved a major hill from the beginning of the tour to the end. Second, it allowed the tour to end with the magnum opus building of Robert Reichert’s — his personal home. It also alleviated some of my heartache from removing the northernmost section of the tour.
3. I revised the stops for a few major buildings to improve the vantage point. When I began the tours in the winter, many trees were bare — as plants grew in, it became harder to observe!
4. I added an informal game called “count the lions”, since many buildings contained lion statues and carvings.

V3: Going Big by Getting Small

V3 map of the tour.

The resulting route was 40% shorter, and consistently concluded within 2 hours. I’ve also begun adding a mid-ride coffee stop to boost morale and break up the rhythm.

I’ve given the tour in its current form 4 or 5 times, to small groups (8 riders maximum). Since the evolution of the route has slowed, my research methods have changed to match its lifecycle. While I still conduct some qualitative research through chatting and observing, I have transitioned to quantitative methods for the long-run, by issuing post-tour surveys. See the Appendix for the questions attendees receive following their tour.

Expanding Tour Routes

To date, over 50 attendees have joined me for bicycle tours. Riders often have questions or suggestions of buildings — whether it’s a building we pass along the route, or a building they’ve seen on social media or in their everyday route. While it’s challenging to work the building into a tour in realtime, I take the time to research and visit every building. Some of these buildings have sparked new narratives, and I’ve developed two additional routes alongside the Robert Reichert route.

These days, I give three tours regularly:

1. Robert Reichert and the Northwest Modern Style
2. Black Architects of Seattle
3. W. M. Somervell and the Seattle Library System

I’ve stopped short of establishing a tour company and establishing an LLC, which has put a ceiling on the growth of these tours. To be honest, I don’t mind. They are a fun way to interact with my city, and a great way to satisfy intellectual curiosity — plus, a change to be intentional with research practices in a non-traditional context.

Appendix: Rider Personas

Appendix: Post-Ride Surveys

Appendix: Key Buildings from Tour

Egan House, a textbook example of PNW Modern style. Designed by Robert Reichert.

Robert Reichert’s personal home, Queen Anne.

Alden Mason’s house, designed by Reichert’s contemporary Victor Steinbrueck.

Home in Queen Anne designed by Bruce Goff.

St. Demetrios, demonstrating innovation in thin shell conrete.