G-FLEX DELIVERY
SCHEDULING APP
G-Flex is a 3rd party app designed to help people who work as an independent contractor for Amazon to deliver their packages, to schedule their delivery blocks without any struggle and frustrations.
Role
UX designer
Timeframe
Oct - Nov/2023
Collaboration
Product manager, Full stack developer
PROJECT OVERVIEW
Most independent contractors for the Amazon Flex program don’t get a chance to schedule well paid blocks. This automated scheduling app will help these people to schedule higher paid blocks without their effort. This app will take account of every possible aspect from users’ specific needs and automatically schedule worthy delivery blocks on behalf of them.
Here you can see few problems:
It’s 2am midnight (when drivers try to schedule surge delivery blocks)
When they try to schedule a block, it becomes unavailable,
When they choose a block it requires image captcha to proceed, during they try to pass the captcha, block itself goes away.
The Solution
G-Flex offers drivers the possibility of multiple settings according to their own availability, requiring the least amount of user’s contribution to schedule a delivery block.
What I was responsible for
User journey
User flow
Wireframes
It is hard to schedule an Amazon delivery block for drivers, even if they want to have base paid ones.
G-Flex is more than a simple automated scheduling app, it’s a daily assistant that helps people schedule a block in their convenience, pick up without any trouble, and deliver their blocks frustration free while helping them to preserve their mental and physical wellness by doing the most time consuming efforts on behalf of them.
What is new?
The Problem
UX research
Empathy map
Sitemap
User testing
Design iteration
Design system
Prototype
User persona
RESEARCH
Focus group interview
Warehouse selection
Manual ETA setting
Warehouse hourly wage setting
Weekly schedule setting
Block duration setting
Random speed search function
Manual settings on refresh speed
Manual settings on Amazon soft block delay
Ignore reserved blocks setting
Amazon Flex Puzzle alarm setting
Manually Solving function on Amazon Flex Puzzle
Hypothesis statement
Weak points
Assorted block duration setting on warehouses
Multiple block scheduling on days
Live ETA to stations function
I’ve arrived assist and Selfie bypass assist
No smart alarm for upcoming blocks
No upcoming scheduled blocks information
The main goal of design research is to inform the design process from the perspective of the end user.
It is research that prevents us from designing for one user: ourselves. I’ve conducted thorough focus group interviews with 25 active Amazon Flex program drivers who are delivering or want to earn more with the program. Before getting started with the interview, I’ve pictured the current amazon flex’s user flow and crafted the questions along with it.
Click here to see research questions.
Interview results
The Offers page doesn’t usually show any available delivery blocks.
Delivery blocks usually appear to be unavailable, it’s highly unlikely to see a delivery block when a driver opens the Amazon Flex app’s offers page.
Surge blocks don't always show up.
It requires constant refreshing on the offers page in a specific time frame when Amazon Flex has less available drivers in the area.
It’s not easy to find even a base pay delivery block.
The Amazon Flex has become extremely competent to find a delivery block due to their active hiring of contractors. In short they have hired more than enough drivers to deliver the packages.
When a driver sees a surge block, it’s not easy to select to schedule that block.
When they see a surge block after frequent refresh on the offers page and try to tap to select that block as soon as it shows up, it goes away.
Captcha solution kills the offers.
Amazon Flex program introduced a captcha solution test before drivers schedule a block in order to keep drivers competent. It happened because some drivers were already using other automated apps to schedule a delivery block. And passing that captcha solution test takes several seconds. In the meantime the delivery block they intended to schedule goes away to another driver.
It takes time to review a delivery block offer.Before selecting a surge block, a driver needs to scan brief information about the delivery block, including the station name, delivery start time, delivery length, and the pay for it. And the driver needs to calculate their estimated time of arrival to that particular station. This process takes several seconds and during that time frame someone else snatches the block away.
Most delivery blocks require high mileage travel to finish it.
Only grocery delivery stations including Amazon Fresh and Whole foods stations take lower miles (utmost 25 miles in total) to drive. Rest of the stations require significantly higher miles (at least 65 miles) to drive. In general, working with the Amazon Flex program puts a heavy toll on drivers’ cars. Including gas, wear and tear. In addition, driving longer means increasing risks.
It’s not easy to calculate travel time to stations.
Traffic conditions change depending on the driver's current location and time of the day. So it requires them to check with it frequently before they schedule a block.
It feels disappointing when you are more than 5 minutes late.
Amazon Flex allows drivers to pick their route up within a window of 15 minutes before, 5 minutes after delivery time starts. When the driver gets late for 5 minutes and a single second, they won’t be able to pick up their route and they won’t get paid for it. That makes their travel back and forth for nothing.
It requires a specific time frame to find a surge block.
Mostly, surge blocks appear early in the morning around 2:30 am - 5:00 am. Sometimes it appears late in the afternoon. Since early mornings have a higher chance to find one, some drivers need to cut their sleep and wake up early. It causes them health issues.
Early morning deliveries are not that easy to complete.
Some customers live in apartment complexes. When a driver tries to deliver their packages early in the morning and if they don’t have access to the building they would need to try to reach out to customers by themselves or through support. Since most customers are asleep early in the morning, drivers are highly likely to return their packages to the station after they complete the rest of the packages. It requires extra time and commuting for drivers.
Forgetting start time is frustrating.
Sometimes drivers have a chance to schedule a block that would start in a few hours or so. It gives them some time for their personal errands and might cause them to forget their next delivery block start time or miscalculate the real time commuting to the station.
Problem statement
Jose is an Independent contractor for the Amazon Flex program, who needs to schedule higher paying blocks automatically, because it requires him too much effort to do so.
Goal statement
Pain points
EMPATHY MAP
Features
If an app that helps to find and schedule a surge block, then drivers would save some time to do it on their own and they would have an extra income out of it.
Our G-Flex app will let users schedule delivery blocks from their phone which will affect drivers who have to schedule delivery blocks spending a long time to find one by subscribing to our scheduling service. We will measure effectiveness by analyzing the successful scheduling rate of total user trials.
Financial – It’s not easy to find surge blocks, delivering for the Amazon Flex program requires a personal car, which tolls extra wear and tear on their cars. And drivers themselves responsible for gas and maintainance.
Product – Scheduling a block has become way more difficult since the Amazon keeps hiring drivers as independent contractors, it requires frequent check and refresh on app to get a block,
Process – It’s not easy to schedule a surged block when the app asks to pass captcha requirement to prove that they’re not a robot, it’s not easy to figure out real-time travel time into the station when it comes to scheduling, getting late to miss a block is one of the nightmares,
Support – when a package becomes undeliverable, it’s not easy to deal with.
Health - Surge blocks usually appear early in the morning, especially night times when there is a slight drivers shortage. It’s not easy to schedule a surge block even when a driver gets up around 2 am cutting their sleep to schedule.
USER PERSONA
USER STORY
As an independent contractor, Jose wants to schedule a surge block with minimum effort, so that he can organize his time efficiently and earn enough to pay his bills.
Happy path
Edge cases
No internet connection or bad services
Not to find a delivery block
Found delivery block doesn’t match with their availability
Delivery block pay doesn’t match with their expected pay
Can not schedule the found block
USER JOURNEY MAP
Traffic congestion
Can not check in within given time frame
Wait too long to pick up a route
Can not complete all the delivery packages
Paid less
DESIGN GOAL
When it comes to scheduling a delivery block, it takes too much time for drivers. So minimizing their interaction would help them save time and effort.
Competitive analysis on Flexer app
SAVES TIME
IDEATION
How might we?
Warehouse selection
Availability selection
Block forfeiting
Block duration selection
Decline unlisted station
Decline unlisted timeframes
SITEMAP
G-FLEX USER FLOW
G-FLEX USER JOURNEY
Remove the bad
Searching function
I’ve arrived feature
Captcha solution assist
Other
Hourly pay rate setting
Refresh speed setting
Soft block delay setting
Multiple block per day scheduling
Auto route estimate to block
Ignore & alert reserved block
Smart alarm
Invoicing detail
Payment info solution
Tips
Feedback
Amp up the goods
TYPOGRAPHY
DESIGN
EARLY SKETCHES
23 margin
WIREFRAMING
DESIGN GUIDELINES
COLOR PALETTE
ICONOGRAPHY
14 gutters
3 columns
COMPONENTS
GRID SYSTEM
4 columns
14 gutters
23 margin
before
USABILITY TESTING
I conducted moderated usability testing after finishing with the Low Fidelity prototype and the results helped me to figure out valuable insights from users. In every design, human error is expected and I was expecting little less than what showed up from the testing.
Task
Log in or sign up into the app
Select your intended delivery stations
Set up your availability
Try to schedule a delivery block.
Result
Average number of tap mistake: 8
Average number of trouble tapping on a button: 13
Average time spent on successful search activity: 430 seconds
Users weren’t able to schedule second shift
Some users were concerned about how they would be notified after successful scheduling.
DESIGN ITERATION
AVAILABILITY
SEARCH SETTING
Most tap mistakes happened due to over clustered home page. It wasn’t easy for users to figure out which buttons to tap for a specific action. So the goal was to simplify the Home page. Used a different approach on Home page layout and transferred some minor features into the slide-in menu.
Button tapping errors happened because some of them were way too small or weren't using any auto layout around the object itself. Another mistake was users weren’t sure if they tapped on the button. That’s because some buttons were not giving any feedback. So, I employed auto layouts on the buttons that were appearing small to make them larger and developed some interaction on buttons to make them give feedback to users.
I added a second shift scheduling feature on the availability tab by improving it as a component.
Added a smart alarm setting feature on the search settings tab to notify users in order to prevent them from getting late from the delivery blocks they successfully scheduled through the app.
HOMEPAGE
after
2nd usability test result after design iteration
Tap mistake: 6
Trouble tapping on buttons: 2
Time spent on successful search activity decreased from 430 seconds to 260 seconds by 40%.
before
after
before
after
HIGH FIDELITY MOCKUPS
Key takeaways
The entire purpose of building this app was to help save time for drivers. In order to achieve this goal, we needed to consider that not all the drivers are familiar with technology. Creating simple to use, easy to navigate through the app was the main priority of the design process.
In addition, finding out every aspect of problems drivers facing during the scheduling and picking up their route was an extremely helpful approach to build the app.
THANK YOU FOR SCROLLING
I appreciate you for taking your time to check out my case study!
I hope you liked it and would be grateful to hear your feedback.