Katy Harper Mills
Queer Artist
ghd_nav.png

GHD In-App Navigation

Grubhub for Drivers is the mobile app used by all Grubhub delivery couriers. As part of a larger redesign, adding in-app navigation to our delivery experience was a top priority. This project had one pilot already go live, which meant we had data to make the next release even better.

Clickable prototype here

Role: Product stakeholder, design discovery, visual design, prototyping, dev handoff + QA

As a food delivery driver, you need your phone on and ready when you're ready to work. After accepting an order, it's your responsibility to get to the restaurant, usually with a navigation app. While navigating to a destination, if you need any o

As a food delivery driver, you need your phone on and ready when you're ready to work. After accepting an order, it's your responsibility to get to the restaurant, usually with a navigation app.

While navigating to a destination, if you need any order-related information, you'll need to switch apps.

First pilot went live in January 2020 and as our quant researcher was gathering feedback, she noted 25% said data was inaccurate. Real-time map data is a high value proposition for in-app navigation, which prompted a second pilot project.

First pilot went live in January 2020 and as our quant researcher was gathering feedback, she noted 25% said data was inaccurate. Real-time map data is a high value proposition for in-app navigation, which prompted a second pilot project.

First, I found Uber's navigation documentation that stated the "3 foot-1 second rule" : looking at phone screen from three feet away, for about one second at a time. But I knew i wasn't going to try and redesign Google Maps. Map apps work for a re

First, I found Uber's navigation documentation that stated the "3 foot-1 second rule": looking at phone screen from three feet away, for about one second at a time.

But I knew i wasn't going to try and redesign Google Maps. Map apps work for a reason and I'm not trying to go against driving instinct.

My hypothesis: Show relevant information at certain event markers, creating a seamless and guided experience as you travel and arrive at your destinations.

My hypothesis:

Show relevant information at certain event markers, creating a seamless and guided experience as you travel and arrive at your destinations.

I knew the visual design part was easy - Grubhub has a robust pattern library, Cookbook, that I could utilize and have fun with. Secondly, I worked with Content Strategy and Research to determine what information was important and when, keeping in

I knew the visual design part was easy - Grubhub has a robust pattern library, Cookbook, that I could utilize and have fun with.

Secondly, I worked with Content Strategy and Research to determine what information was important and when, keeping in mind of the 3 food-1 second rule.

Navigation_mockup_2.png
Navigation_mockup_3.png
User testing focused on the arrival milestone and what the mental mindset is when you're approaching. Partnering with Research, we conducted unmoderated testing on usertesting.com. Was a timer anxiety inducing? Did a driver understand what signal w

User testing focused on the arrival milestone and what the mental mindset is when you're approaching. Partnering with Research, we conducted unmoderated testing on usertesting.com.

Was a timer anxiety inducing? Did a driver understand what signal was going where and to whom?

Results? We found that leaving the decision up to the driver (showing 2 buttons) was the most clear and intuitive.

Navigation_7.png
Currently, the 2nd pilot is in development and expected to launch December 2020. Google’s API is for iOS and Mapbox with Android. Would I do anything differently? I’d push for earlier seat at the table, as our research might have been skewed due

Currently, the 2nd pilot is in development and expected to launch December 2020. Google’s API is for iOS and Mapbox with Android.

Would I do anything differently?

I’d push for earlier seat at the table, as our research might have been skewed due to the fact more than half of surveyors didn't realize they were in the pilot.