Role: Research, strategy, design
Project duration: 1 month (11.11.22 - 12.15.22)
Primary stakeholders: CTO (Adrian), software developers, Sphere learners
Results: 250% increase in the number of learners who posted about Sphere courses on LinkedIn
About Sphere: At the time of this project, Sphere was a corporate education startup where vetted leading experts ran live, cohort-based courses for professionals around the world. Sphere is backed by Felicis Ventures and Y Combinator.
Sphere pivoted out of corporate education in March 2023, but that is a different story. Learn more here.
This project: This case study focuses on the referrals feature I designed for the corporate education product, from user research to ideation, wireframing, iteration, implementation, results, and learnings.
Target User
The referrals feature targeted users who have enrolled in a Sphere course, whom we referred to as "learners".
Occupation: Primarily tech roles, varying in seniority. Examples include Head of Data Engineering, Senior Product Manager, Software Engineer.
Company: Primarily software companies, varying in size. Stripe, Flipkart, Tinder, and Accenture are examples of companies Sphere had annual contracts with.
Primary goal: Upskill in topics covered by Sphere's courses. Examples include data science, machine learning, and sales.
Project Background
January 2022
Sphere implemented a referral system that encouraged learners to share their referral link with friends and colleagues during registration, onboarding, and after completing a course. We offered a $50 Amazon gift card as an incentive for people to send and accept referrals.
With this system, the overall referral conversion rate was 21%.
The average number of paid seats from referrals was 18 per quarter.
September 2022
I joined Sphere as Design Lead.
The Goal
November 11, 2022
I was given the goal of improving the referral system in order to increasing that number to 30 paid seats from referrals for the next quarter.
Research & Outreach
November 14-16, 2022
I did background research on referral program best practices, prepared discussion guides, and reached out to Sphere users who had either sent or accepted referrals previously.
First Iteration
November 14-16, 2022
Because we didn't want to waste time while I waited for users to respond to my outreach, we got started on the project with the following hypotheses:
1. Offering $100 credit to use towards future Sphere course is a more valuable incentive than a $50 Amazon gift card
2. Users will be encouraged to send more referrals if they can see how close they are to earning enough credits to cover a free Sphere course
Based on these hypotheses, I wireframed a Balance page to help learners keep track of their referrals and credit balance while I waited for users to respond to my outreach.
I iterated on the designs with Adrian and handed them over to our offshore developers.
November 17-21, 2022
Developers released the Balance page to staging, and I provided feedback.
I created a planning sheet for event tracking and the devs implemented it (we used Amplitude).
Developers released Balance page to production (see prototype below).
November 22-23, 2022
I worked on a different project (Sphere Zoom App).
November 24-27, 2022
Thanksgiving break
User Research
November 28 - December 6, 2022
Next, I conducted Zoom meetings with 5 Sphere users: 2 referral senders and 3 referral receivers.
I asked them about their experiences with Sphere's existing referral system to find out what the friction points were and how they could be improved.
I directed them to the new Balance page, observed how they interacted with it, and gathered feedback.
I also asked them about their experiences with other referral systems to find out what generally motivates them to send or accept referrals.
Affinity Mapping
November 28 - December 6, 2022
After these interviews, I organized my notes and created an affinity map to extract the following insights:
1. Learners find out about Sphere primarily through LinkedIn + word-of-mouth
2. Learners decide to enroll based on interesting/useful topic + reputable instructor + positive experiences within network
3. Learners are happy to refer others to Sphere if they feel the course is relevant to them
4. Learners care more about sharing valuable knowledge than getting small monetary rewards
5. Sphere’s existing referral system is unclear/inconsistent
6. Learners usually don’t think that much about referral systems/rewards
7. Learners don’t care too much about tracking their referrals in a dashboard
Turns out our hypothesis about $100 Sphere credit being more valuable to learners than a $50 Amazon gift card was correct, but our hypothesis about users wanting to keep track of their credit was not entirely correct. Based on these learnings, I began the next iteration.
Second Iteration
December 7-9, 2022
Because my research showed that:
1. Learners find out about Sphere primarily through LinkedIn and word-of-mouth and
2. Course instructors and learnings are the most important factors when sending a referral
I proposed a solution that would:
1. Leverage social referrals
2. Highlight course instructors and learnings
as opposed to focusing on monetary rewards.
Competitive Research
December 7-9, 2022
Once I decided to focus on social referrals, I conducted further research:
1. Researched how competitors such as Section4 and IdeoU do social referrals
2. Spoke to our advisor Jerllin Cheng (ex-VP of Product Development at Section4) about social referral best practices
3. Examined how past Sphere learners posted about Sphere on LinkedIn
4. Tested out the technical limitations of link and image sharing via LinkedIn, Twitter, and email
Low-Fidelity Designs
December 7-9, 2022
I created some low-fidelity designs of an offboarding page that would appear at the end of the course with certificate images that the learners could share along with a referral link.
These designs aligned with the solutions I proposed based on my research because:
1. Encouraging users to share their experience with Sphere courses on social media
2. The certificates would be designed to highlight course instructors and learnings
After sharing these designs with Adrian, he suggested trying out a few more variations in the UX.
Key Decisions
December 7-9, 2022
I created 3 versions of high-fidelity designs of pages that would appear at the end of the course:
1. Offboarding page with shareable certificates and referral link
2. Offboarding dashboard displaying all the actions the user can take
3. Offboarding checklist walking the user through tasks to complete
Of the 3 iterations, Adrian and I decided the offboarding page was the best option because the other 2 options had intermediary steps that would decrease the likelihood of a user sharing their certificate and referral code.
If we had more time or were in a bigger organization, we would have conducted usability testing on these options, but the nature of an early-stage startup is moving very fast!
High-Fidelity Designs
December 7-9, 2022
Below are the high-fidelity desktop and mobile designs that I passed to the developers.
Offboarding page
This page appears in Sphere's learning portal once the learner has finished a course, where users can view, expand, and select certificates that are personalized to them and the course they took. See the Graphics section below for more details about these certificates.
The primary action button is downloading and sharing the certificate in order to maximize the chances of the user doing so. The secondary action is filling out the end-of-cohort survey, which is also sent to them via email.
Download and share modal
After selecting the desired ticket and clicking the "Download and share" button, a modal opens explaining the referral credit system and showing the different sharing options. The primary action button is "Share on LinkedIn" because my research showed that LinkedIn is where learners are most likely to share a Sphere course to their network.
In order to alleviate the pain point of writing a LinkedIn caption, we provided learners with the option to copy a pre-written caption telling their network about the course they completed and the opportunity to earn $100 towards Sphere courses.
We also provided an option to copy the link only if learners wanted to write their own caption.
Graphics
My user research showed that the caliber of the instructors and value of the course content were the most important considerations when referring courses to others, so I designed the following graphics:
1. Certificate of Completion that we always give to learners upon completing a course
2. Graphic with course information, instructor image, and learner's LinkedIn image
3. Graphic with instructor information and guest speaker information
4. Graphic with instructor information and key learnings
Our developers were able to automate the generation of these graphics which was very helpful considering Applied Causal Inference had 79 learners in it!
Results
December 12-15, 2022 - Implementation
After releasing the offboarding feature in December 2022, we saw a 250% increase in the number of learners who posted about Sphere courses on LinkedIn.
However, only 36% of learners who shared the graphics on LinkedIn included the referral link. The other 64% deleted the referral link from the pre-written caption, or wrote their own caption without including the link.
We hypothesized that this may have been because the language in the pre-written caption about the $100 credit was too salesy. We knew from user interviews that learners did not want to come off as salesy on LinkedIn, so we experimented by removing the language about the $100 credit from the caption. After that, the amount of learners who included the referral link when sharing the graphics increased by 10%.
Reflections
Because Sphere pivoted out of corporate education not long after, we were not able to track the referrals for long enough to determine how much of an effect the offboarding feature had on paid seats from referrals per quarter.
Given more time, I also would have liked to conduct more thorough usability testing. However, I felt quite proud of the 250% increase in social posts about Sphere!
If you have any questions, feel free to reach out at ashley.zhang@columbia.edu. Thanks for reading!
