CarpoolWorld Mobile App

Client Project — Designing for better engagement & retention

Duration: Jun — Aug 2018
Tools: Paper & Pencil, Adobe XD & InVision


Overview

CarpoolWorld does real ride-share matching since 2000. It connects employees to find colleagues to carpool with.

I worked with CarpoolWorld over 3 months to understand their business and product to redesign their mobile app. We collaborated with their team and focused on key areas to improve that would better demonstrate their value to users.

Business Needs:

  1. Mobile App
  2. Increase User Retention Rate
  3. Improve data collection (surveys)

Team
Damini Bhatt, Malvi Shah, Suleiman (Me)

My Role

  • Understanding user journey to improve onboarding
  • Streamline data collection
  • Lead efforts to redesign essential parts of the app

Strategy

Evaluating The Existing Web App

Using this feedback, we categorized issues based on severity on a board. Looking at the ‘high severity’ issues gave us a strong indication of the pain points and hence a direction for our design which we eventually opted for.

Proposal — Redesign Goals

Ease & Shorten Onboarding

The current onboarding was excessively long and time-consuming. It asks for too much information upfront that can lead to increased cognitive load.

Before
Users need to manually fill a lot of fields, just to see potential matches. If a match isn’t found, then all that data entry effort would be for nothing!

Data inputs required by user to find Carpool Matches — Onboarding

Re-visualizing this onboarding journey as a flowchart reveals 6 extra steps (of data input) required by users, before they can ‘hopefully’ find a carpool match.

Onboarding visualized as a flowchart

After
To avoid wasting user effort, onboarding needed to be short and clear. The aim was to ask users only the bare minimum needed to find matches. Moreover, asking only one question per screen allows users to easily focus on the question at hand (Single Responsibility Principle).

Redesigned & Simplified onboarding for first-time carpoolers

Data Collection

Although onboarding was simplified, most of the data that the business needed from users was now omitted. This data is part of CarpoolWorld’s ‘Commuting Habits’ survey. It helps determine an individual’s contribution to the environment.

How might we collect this survey data, without troubling the user every time?

Before — Explicit data collection
While testing the previous version, almost all participants did NOT want to take the survey every week. There can also be a tendency to enter highly skewed data which can be a serious problem!

After — Blind data collection
All the data required by the ‘Commuting Habits’ survey can be automatically collected in the background, without asking the user. This can be achieved using the Google Maps API. It eliminates the need of explicitly asking users for data such as distance driven, fuel consumption and costs.

The Google Maps API helps increase data quality while decreasing user frustration of repeatedly taking a survey.
Data Collection before & after with the Google Maps API

However, not all information can be auto-collected, we still needed some data that relied on manual user input.

Gradual Data Collection

First, we help provide users value through the app. Then, at key moments, we encourage users to take actions. In other words, we sprinkle the survey over the user journey rather than forcing the user at one moment.

SplitCost (New Feature)

How might we give users something of value before asking for commitment?

SplitCost is a provision to track carpooling expenses for trips and fuel. It aims to create a sense of responsibility and reliability among carpool group members. It also provides an opportunity to gradually engage users.

SplitCost helps address two business needs:

  • Retention rate — SplitCost gives a reason for people to keep coming back to the app
  • Data collection — Prompt users to enter trip expenses (data) because the app is providing value (via SplitCost) by helping them manage their carpool costs

Testing the Prototype

Test Results & Outcome

Feedback and our success metrics
Our final client presentation

Learning

  • Don’t ask users for commitment early on when you haven’t provided any value
  • Don’t let users feel that their efforts were for nothing
  • Design with limited information and legal restrictions

Final Recommendations