The grant application process is inconsistent, unapproachable, and difficult to complete

Contextual Inquiry & User Interviews

Completing the Clean Cars for All grant application is mostly completed on a desktop device. We spent the majority of our first week and a half getting to know the grantees of the program and GRID Alternatives case managers. We asked a lot of questions, attended an Oakland Clean Cars for All event, and began to collect information into post-its for eventual affinity mapping.

Usability studies

We created a research plan, conducted two remote usability tests, and a third usability test in person. We tested understandability of the Clean Cars for All website versus two similar websites, we tested three additional goals in the Clean Cars for All website, and conducted usability testing of the Clean Cars for All application in Fluxx.io.

Key observations

• The Clean Cars for All website is an easy site to navigate, however it was difficult for users to quickly understand if they were eligible for the grant and what to expect in the grant application process.

• The third party grant system, Fluxx.io, had significant usability issues, with 100% of testers having non-critical issues, and 33% of testers abandoning the application altogether.

A lesson in what makes us human

We invited our clients, GRID Alternatives, to join our team in a co-design session. In that session, our clients and our team completed an affinity map, empathy map, and journey map. From that meeting, we got a much deeper picture of our target user and began to deeply empathize with them. A memorable quote from our empathy mapping was:

“My mother was diagnosed with cancer and I have to drive her to chemo.” 

This quote became our north star and guided the creation of our archetype user. Our archetype user, the Discouraged Decider, was someone who:

• Is time-constrained

• In need of a vehicle upgrade for security and reliability

• Needs support with involved applications

• Listens to the TV and radio

• Is Goal-directed

• Needs high return on investment on new activities

The result of our meeting was a convergence of information that helped us identify the key pain points that would be the basis of our problem and hypothesis.

Making the Clean Cars for All application a friendlier and approachable experience

Eliminate unnecessary steps, make language approachable, give the user all opportunities to succeed in the grant process

Pragmatic form design and UX writing

We front-loaded the grant application form so users can confirm their eligibility at the beginning of the grant application process, instead of needing to complete five separate actions before confirming eligibility. We redesigned the landing page experience to so that the user feels welcomed and is taught how to use the Fluxx.io application. We also introduced images to several of the application pages so users knew how to complete the application, knew what to expect, and had the resources available to receive help. 

Revising the existing form flow was an excellent immediate solution. A well-researched and well-tested form laid the groundwork for our long-term solution: scaling the application for a growing population of applicants. By re-designing the Fluxx application, we were able to work around limitations that the Fluxx.io third party vendor presented.

Key improvements

• Simplified flow so the user could quickly confirm their eligibility

• Improved experience for users looking to understand what to expect

• Reduction of all critical errors by 100% with the introduction of

• Reduction of non-critical errors by 60%

"We are incredibly impressed. We are going to change the way we work with our clients"

This design project culminated with overwhelmingly positive feedback from our client, GRID Alternatives. They were excited to receive recommendations that could be implemented immediately, and were delighted at the creativity we demonstrated to work around the limitations of the grant system. They also appreciated our engagement with their partnering organization, BAAQMD, and expressed appreciation with being involved in a project that affected them directly. 

What’s next?

We looked at the bigger picture and understood that the major limitation in implementing changes was insufficient funds for BAAQMD’s software engineers. As part of our recommendation, we recommended a partnership with the Code for San Francisco non-profit civic-hacking organization. With Code for San Francisco’s help, we recommended that the following changes be implemented:

• Implement changes to the content to improve engagement and satisfaction from target users

• Introduce imagery and more iconography to improve satisfaction and learnability of the program

• Implement new flow for the Fluxx portal to reduce burden and effort of learning eligibility

• Pair modifications we made with a robust email campaign

Key takeaways

When in doubt, communicate with the client

We hit a few obstacles when it came to understanding our users and the grant application process during the first week of our sprint. Due to the intricacy of the organizations behind the Clean Cars for All grant, we were running in circles and going too deep in the weeds in trying to understand the problem landscape by ourselves. We course-corrected as soon as we involved our clients in the conversation. We gave them room to teach us about what they know best, their business process, the landscape, and their clients. Once we increased communication frequency, we were able to move speedily through the UX process.

Testing early and often

We were, at first, unsure how to test the grant application. It required account creation, email verification, and uploading of personal data. In week two, we finally tested doing the application ourselves and found a few resources that allowed us to conduct baseline usability testing. With grit and persistence, we were able to have three usability testers show us the pain points in the application system.

The higher education space is underserved

The Registrar’s Office plays a critical role in the operations of universities in the United States. It manages and distributes information to people across the university, for a typical week in a Registrar Office with 4 staff members in a medium size university:

• 80-150 messages and tickets from all over campus

• Toward beginning of quarter, additional 50-150 emails / tickets regarding course & schedule,  50-150 emails regarding grad admission, per day

• Toward end the of the quarter 100-200 messages regarding degree progress for graduation

Most of these requests are repetitive. As a result, students, faculty and staff have to wait days to weeks for a response from the Registrar’s Office because of the use of manual solutions and inefficient tools. 

AIUR’s differentiator is that it offers deep specialization in university data

There has been a boost in machine learning, deep learning, and natural language processing (NLP) based solutions in education. The education-based AI market is projected to grow from $537.3 million in 2018 to an expected $3.6 billion by 2023, indicating that stakeholders in the education space are increasingly becoming more open to mining and monetizing big data.

Despite the growing opportunities, an AI-powered intelligent assistant targeted towards the education industry is uncommon. While many of the current players have unique advantages, there is currently no “de facto education registrar assistant”, while other products are either service-based companies which do not specifically target Registrar Offices, or are other companies mainly focused on sales and marketing.

AIUR, therefore, will focus on differentiation by increasing customer willingness to pay relative to rival’s offerings with the deep specialization in understanding university data and the ability to extract immediate value with that data through our AI-powered platform.

AIUR’s edge will principally be vertical, outperforming rivals’ products on dimensions for which customers would agree that “improved understanding of the higher education space is better.” 

What are our hypotheses about taking AIUR to market?

What does AIUR look like under the hood?

AIUR provides building blocks for any university to organize its large volume of unstructured data and deploy a smart assistant that its students, faculty, and staff can interact with at any time. The minimal viable product of AIUR offers personalized, accurate answers to university constituents’ wide range of questions.

The primary goals of AIUR are to:

• Create a product that leverages Natural Language Processing (NLP) and an API platform-as-a-service (PaaS) specifically for the higher education industry.

• Simplify the experience that university data providers undergo to securely and easily access and distribute data for university-related projects

• Provide self-service options for customers

• Provide access to internal university databases

• Expose core data to internal users in a controlled manner

• Expose selective data to external users in a controlled manner

AIUR: Offering personalized, accurate answers to university constituents

AIUR is a fully compliant PaaS solution providing building blocks for any university to organize its large volume of unstructured data and deploy a smart assistant that its students, faculty, and staff can interact with at any time.

The goal of AIUR is to offer personalized, accurate answers to university constituents’ wide range of questions by leveraging cutting edge machine learning and natural language processing.

What is our product release cycle and MVP?

We went through a planning poker exercise, translated effort into time estimates and developed a release plan.