Some people own empty parking lots or land. We made it possible to publish multiple parking spots at once.
We saw huge improvements. Our publish rate for unpaved land increased 13x per host. This boosted Neighbor's overall publish rate by 4x per host! Conversion rate for unpaved land dropped 9.5%. In light of the huge wins, however, we planned to recover that drop and bring the tool to other space types.
My process was not linear, but this case study is.
Alex (PM) analyzed user data and conducted user interviews. I used Google Maps to analyze unpaved lots belonging to thirteen of our hosts.
Some of our findings…
Some hosts store mostly RVs or tractor trailers. Others, cars.
Our highest earning hosts publish dirt or unpaved parking lots.
Large vehicles are very difficult to maneuver in tight spaces, such as RVs or tractor trailers.
Often, the host lives on the lot. They only publish part of their land. The rest they want to keep clear.
Some areas in Google Maps can be hard to read! For example, they can be dark with awful glares and other imperfections.
A host may recognize their property in person, but a top-down map view can be difficult to recognize.
Unpaved lots come in all shapes, sizes, and angles.
Many unpaved lots are visually obstructed on Google Maps, such as by trees.
Design progression
Explanation
We wanted to allow users to lay down multiple parking spots at once. Users would be tasked with doing so in an already lengthy listing creation flow.
I wanted the app to automatically lay out parking spots (see slides 2-6). This relied too much on engineering, however, and gave users too little control.
The VP of Product helped me change my thinking, and we came up with an intuitive tool that empowers users (see slide LAST).
Here are a few ways that we incorporated research
Hosts can click on the vehicles to resize them.
We designed the tool to accommodate land of all shapes and sizes, not just rectangles.
We created visual guards to allow for driving in between vehicles.
We disclose the interface progressively at the start to draw attention to important tools.
We start with a mixture of small and large vehicles to encourage resizing.
We allow adjustment of widths of parking spots.
User testing
We tested the design with three of our top hosts.
Quotes from users
“I didn’t notice the button to add another row. Even if I had, I wouldn’t know what it would do.”
“To resize the vehicles, could you click on them? That’s what I would do to resize them.”
“Can you change the width of each parking space? Sometimes vehicles are just bigger and don’t fit in a 10 foot space.”
“When will I be able to determine the size of the vehicles? My parking spots are large. You’re showing me cars, but the reality is different.”
“I see RVs, cars, trailers, a boat. If someone is storing a car, for example, how would the app determine what size parking spot that renter should get?”
“This is definitely something that could help me visualize how things could be set up on my lot. That’s an awesome tool!”
Final designs
We gave people a way to arrange and publish many parking spots at one time.
They can see potential earnings, and ths encourages them to maximize their space.
We disclose the interface progressively to make it easy to learn the tool.
User context This host is creating a listing. They are looking at a map of their property after having given their address.
We hint to users that vehicles can be angled on a row.
I made an optional tutorial video to educate users
Users can change the lengths of parking spots and set prices for each size. When the user is done, we show an estimation of what they can earn.
Results
4x Listings published per host
13x per host of unpaved lots
-9.5% conversion publishing of unpaved land
Leadership planned to showcase the tool to attract prospective hosts, to draw them in.
-9.5% is no small drop. However, in light of the huge wins, we planned to recover it and bring the tool to other space types.
Reflection
This project helped me to learn that software should empower users. Automation can be good, but so can intuitive tools.
I was reminded how valuable stakeholder feedback is. I got stuck iterating on different ideas with the same fundament. Our VP of product helped me break out of that loop.
It's important to move past my biases. I can grow attached to a design or idea. It can be better to just let go. Try something new.
I'm happy with the results, but I learned that a rich user experience like this needs a lot of user testing and feedback. I wish we had time for more.
Along those same lines, our first release had a lot of bugs. It's important to release quickly and often, but I wish the build had been tested more thoroughly first.
