Michael Shin

Case Study / 01

UTern. Making the internship hunt feel less like begging and more like being found.

UTern logo
Role

CMO · Brand, Content & Growth

Timeline

May 2026 – Present

Team

7 college students across the country

Scope

Brand voice, product surfaces, launch, content & growth

Result

150+ signups & counting

01 / Context

UTern is an AI-powered internship discovery platform for college students. When I joined, the product worked. The story didn't. In a category where every competitor sounds like a career-services office, UTern needed to sound like the one company that actually remembers being 20 and unemployed in April.

02 / The Problem

Students don't trust career platforms because career platforms talk like LinkedIn wrote them. The task wasn't awareness. It was believability. How does an AI recruiting tool earn the trust of an audience that's allergic to corporate speak and pattern-matched to skip anything that smells like an ad?

03 / The Thinking

The insight: the internship hunt is an emotional experience. Rejection emails at 2am. "We've decided to move forward with other candidates." The group chat panic. And no one in the category was speaking to the feeling, only the function.

So the voice strategy was simple: UTern talks like the friend who already got the offer. Direct, funny about the pain, dead serious about the outcome. Every piece of content had to pass one test: would a student send this to their group chat?

04 / The Work

The positioning, made real

The whole platform hangs on one line: an agent on both sides of the table. The voice had to earn trust the category never does, and it had to live on every surface, not just the ads. Copy that sounds like a friend with the offer, right down to "a match, with the receipts."

UTern landing: an agent on both sides of the table
How it works: Bob makes the case, Barb checks the fit, people make the call
A match, with the receipts

Bob & Barb

Two agents, one on each side. Bob works for the student; Barb works for the recruiter. Giving the AI a face and a name is what turns it from surveillance into help, and it gives the whole product a consistent voice, from onboarding to the shortlist.

Bob, the student's agent

Bob · the student's agent

Barb, the recruiter's agent

Barb · the recruiter's agent

Hi, I'm Bob
Your agent works the pipeline, the recruiter side

The student's side

From "Hi, I'm Bob" to a ranked shortlist with receipts: the search, the chat, the matches, and the moment it hands back roles worth applying to. Every screen carries the same voice.

Your agent works the search for you
Chatting with Bob
Ranked matches
A shortlisted role with the receipts
Your deck's tuned, here's your strongest match
Roles are saved, time to start applying
The internship list
Killed Direction

We started as "Tinder for internships." Investors didn't like the sound of it: too bubbly, too easy to dismiss. So we repositioned around what the product actually does for students, a platform for self-discovery, where the agent doesn't just match you, it helps you figure out what you're looking for.

05 / Results

In five weeks, UTern went from an idea to a working platform with a voice students actually trust, all at zero paid spend.

+50%

Engagement vs. baseline

3k+

Students reached, organic only

150+

Signups & counting

06 / What I'd Do Next

Build a student ambassador engine on the top 10 campuses, and turn the highest-performing organic formats into a repeatable weekly content system that runs without me touching it.

Next Project APEX STUDENT →