Student Homepage
User Testing & Redesign

Old Homepage Areas for Growth

The old designs hadn’t seen a facelift in years so it was time for an interface update. Additionally the engineering team was shifting from Angular to React. The old homepage had too many calls to action for users. With user research we also wanted to identify the best way to meet student user needs. Additionally many students reported not understanding that this platform was uniquely positioned for only current students and alumni of the same alma mater. From a UsabilityHub baseline test 4 of 15 (26%) of students understood the value prop, a third of participants said ‘networking’ but were vague, and the rest were way off in their understanding.

That seemed like a large miss because our value proposition is part of what makes the platform unique. Also because we are a white labeled company we wanted to make sure our clients, universities, could still have some customization without detracting from the overall UI.

Design Principles for New Homepage

1. Have ONE primary call to action and clear path to value (reduce clutter and set students up for success with connecting with alumni)

2. Clearly convey what PeopleGrove is and our value prop so its clear to users why and how they can use the platform to meet their needs

3. Allow for limited, thoughtful administrator customization

4. Consider pre and post connect states so that homepage feels dynamic and relevant after initial connection

User Research 

This was a multiple month project conducted via a two week sprint cadance. At various points stakeholders such as the CEO, CTO, Head of Product, engineer managers were consulted for feedback and buy in.

Identify User Segments

Define:

What are the biggest problems for key user segments that we can and should solve?

Possible segmentations:

By goal: job/internship seekers, explorers / wanderers, career switchers, networkers, wise mentors, etc)

By school year: under / upperclassmen, alumni

Generative Surveys

 Exploratory questions to survey:

• Do users already have goals in mind or do they need heavier guidance?

• When do users’ problems meaningfully differ? Which user segments need to be treated differently? Which the same? 

• What are the most pressing / urgent problems by user segment?

• What user problems can we effectively address on PeopleGrove?

Survey Results

Top student goals, representing ~80% of survey responses, are to: 

• Find an internship or job (26%)

• Network with people in my industry or career path (21%)

• Learn about different career paths (16%)

• Set an academic / career goal  (14%)

Wireframes

Based on survey results, then I proposed a few different wireframes. The concepts ranged from being module or “tile” based, exploring different ways of integrating the value prop/social proof, following a search approach for a mentor similar to “Airbnb”, having a one mentor spotlight, having a carousel based on anticipated student needs such as “preparation for grad school” etc. This is also where the concept of bookmarking an alumni and having them become saved in “your connections” section began and eventually was shipped. That allowed students to have immediate engagement with the homepage even if they weren’t ready to immediately message an alumni.

 Next Round of User Testing

After receiving feedback from stakeholders such as the CTO, Head of Product, and Engineer Managers then I selected the top 3-5 student dashboard wireframes and value proposition copy.

I ran usability tests to see:

• Which design best conveys what is PG + value prop using 5 second test & preference test

• If it is obvious how to connect with an alumni by using first click tests

3 Different Approaches Preference Test

Learning goal: Which of these action types do students prefer:

1) Career search (by city, industry, role)

2) Browse alumni (that have been previewed one at a time)

3) CTA to search the alumni network

Hypothesis: One of the action types will be preferred the most by students.

Experiment: Preference test where students pick which action type they prefer.

Metric: Start with 30 testers. 1 of the 3 options gets +50% of the votes.

Testing audience: 30 college students, aged 18-21

Value Prop 5 Second Click Tests

Learning goal: Which design best conveys our value prop to students?

Hypothesis: One of the value prop designs will effectively convey PG’s value prop

Experiment: Test 3 designs in 5 second tests with a recall question and a short answer. Question: “What do you think you could do on this site? Please be as detailed as possible.”

Metric: 80% of students will understand the value prop for a given design

Testing audience: 15 college students, aged 18-21 x 3 tests

First Click Test

Learning goal: With the new designs will there be a fairly clear call to action that will drive connections?

Hypothesis: Most students will click somewhere within the mentor card

Experiment: Where would you click first? Then the actual click occurs and a second short answer “Why did you want to click there?”

Metric: 70% of students will click within the mentor card or related to sending messages/connections

Testing audience: 30 college students, aged 18-21

Final Designs

After I finished the initial research, wireframes and product design work, another Product Designer polished the designs as seen in the carousel below. These are the designs that shipped. Main additions were an upcoming meetings module, nurture your connections section (which basically are reminders to students about actions they need to take to follow up with alumni) to drive user engagement.

Success Metrics

• 70% increase messages sent by new students in first 7 days

• 50% increase successful connections by new students in first 7 days

• 65% in messages sent by all students in first 30 days

• 20% in successful connections by all students in first 30 days