Background
Vehicle deployments were previously based a mixture of “release anywhere” or a rudimentary nest system with specific locations. Analytics into RpD (rides per deployed vehicle) showed a significant level of cannibalization in rides between Fleet Managers or operators within a given zone, meaning more deployments does not equal more ridership. In order to optimize vehicle supply to match real demand, the product team created a new Nest Deployment system using demand forecasting from Zoba (https://www.zoba.com/) a third party platform for fleet optimization.
Roles
UI/UX Design
User Research
Visual Design
Deliverables
Research
User Flows
Wireframes
High Fidelity Screens
Tools
Figma
The Team
I was the Lead Designer, assisting on research, creating wireframes, high fidelity screens, prototypes etc. Product Managers Michael and Prat were key partners with operations and city compliance support, while VP of Operations Hunter (now CTO) led the direction of the product.
Fleet Managers
We worked directly with several top Fleet Managers to better understand their workflow and common pain points. They ranged from single operators to larger operations with multiple employees and fleets. They include but were not limited to the following:
Moises (Orlando, FL Lexington, KT), Renee (Newark, NJ), Josh (San Jose, CA), Jhari (Atlanta, GA), Conor (Nashville, TN), Casey (Los Angeles, CA)
Existing Flow
The first iteration of the nest release flow used system action sheets and a simple bottom sheet that conveyed basic information about the nest. We decided to work based off this existing system, keeping the general pin and flight sheet design while adding more information to help Fleet Managers understand the demand signals and release requirements.
The old flow required going into a different, nest only map that was separate from the main map that operators use. This meant that important features and functions such as vehicle locations, filters, and zone perimeters were not visible on the nest map. Incorporating nests into the main map was a primary goal of this redesign.
Additionally, there is a validation check at each step: location check, vehicle check, zone check, etc. If any of the validations checks fail, the same generic error is delivered.
User Flow
The new nest system necessitated a rethinking of the standard operating procedure around deploying vehicles. I worked with Michael and Prat to create new rules around claims and releases to standardize operations and incentivize adherence to Zoba demand signals while deterring fraudulent activities by bad actors.
After feedback from Fleet Managers, we also added new features to help them manage and send feedback about the nest locations.
Fleet Manager Feedback
Renee
Demand signals from Zoba don’t seem accurate, and would prefer to be forecasted demand signals rather than real time.
Transportation routes are important in designing the rules around use. Most Fleet Managers will have their warehouses outside their main zones of operation. Strict rules around nests claims makes it difficult to plan for their day.
Flags are a frequently used tool, but requires enough Ops capacity to address in a reasonable amount of time.
Emily
Excessive amount of nests in certain areas is overwhelming, would like better filtering and favoriting features. However, Recommended Nests is helpful for finding new deployment locations in the outer perimeter.
There is a perennial problem of the ice cream truck , the presence of a van of scooters will create demand from riders, but there were no nests around to release easily.
Jhari
Also finds the amount of nests excessive, his employees don’t have the knowledge to know where to drop even with the recommendations. Wants the ability to favorite and filter.
Finds the recommendations to be based on supply rather than demand and prefers the latter. Currently most recommendations are in the outer zones, Atlanta is a very spread out and makes it difficult and expensive to drive around.
Connor
Biggest benefit is the ability to train and direct employees using the nest system rather than a list of instructions and google maps.
However some nests are not good due to size or location and he has had to flag many. Certain nests also seem to be in non-deployment zones or are not present in high demand zones.
Overview
The new nest system is accessible via the main map. Because of the increased amount of pins, we added segmented controllers to allow Fleet Managers to change their map to view vehicles only, nests only, or both.
When an operator selects a nest, a new flight sheet with summary information and common actions is pulled up. From there, operators can release at the nest, or reserve it for releasing later. The first iteration included single claims only, but we later released the ability to claim multiple nests.
Once at the nest location, operators can release as normal, but several new anti-fraud features we released necessitated new error screens.
From the nest flight sheet, operators are also able to flag and favorite certain nests as well as see current vehicles.
Manage Nests
After a few months of Fleet Managers using the new nest product and giving their feedback, we needed to implement a management system for common actions such as flagging, favoriting, changing locations, and suggesting new nests.
Flight Sheets and Icons
The nest and vehicle flight sheets were redesigned to include more information. Here’s a closer look at the design assets such as pins, icons, map filters, sheets and more.
Business Impact
The nest redesign had a large impact on the business in several ways. First, it changed the workflow of the Fleet Managers, some more significantly than others. The rollout and adoption of Zoba and required nest deployment happened over several months, first in a few test cities, and then more broadly to other markets. The transition was not frictionless, with some Fleet Managers expressing more dissatisfaction than others. There were many times that we had to change our plans and pivot the direction based on test results.
Additionally, due to macroeconomic factors, the micromobility industry as a whole saw lower utilization rates than in previous years. However, we saw that cities with Zoba and nests on average had higher RpD than cities with nests and no Zoba requirements or no nests at all. The decrease in ridership was much less pronounced.
However, Zoba had difficulty predicting demand changes from events such as festivals, concerts, and games. The next stage is to build out a way to incorporate the knowledge of local ops and fleet managers about events into the recommended nest system.