Say Hi to Berry · Heiko Polinski

Role: UX Designer
Illustration
Team: CEO
Lead developer
Backend
Frontend
UX (me)
Stack: Adobe XD
Illustrator
After Effects
Lottie
Status: Launched
retention measured 2 months post-launch
Span: Early 2022
10 weeks

The Trigger

A recent visual overhaul had made the student app look better without making anyone come back. Thirty-day retention sat at 8 percent against a 12 to 15 percent category benchmark, and the store rating had stalled at 3.9. The app looked fine, but it gave students no reason to return.

Five screens of the student app before the engagement work: home, tasks, timetable, grades, and careers — A clean redesign that still left retention flat.

Research

The focus groups brought together twelve students aged fourteen to eighteen; a survey reached a hundred and four more. The teardown of Duolingo, Knowunity, and Scoolio mapped what regular-use apps did that berry2b didn't. The through-line was an emotional-engagement gap: students found the app useful but had no relationship with it, and they responded strongly to mascots, achievements, and a visible sense of progress. Because every participant was a minor, the research ran on explicit consent, anonymized data, and the option to drop out at any point.

research timeline · five weeks

Virtual focus groups

1 week

Online surveys

2 weeks

Competitive analysis

1 week

Data analysis

1 week

Mixed methods over roughly five weeks: focus groups, a survey, a competitive teardown, and analysis.

Persona photo: Svea, a graduating student

Svea Heilandt, 18

12th-grade Gymnasium student in Düsseldorf, months from graduation, not actively looking for vocational training

"As graduation approaches, I want to enjoy my time with friends and explore my options before deciding on my next steps."

Goals

An app that helps her manage her academic and personal life efficiently
Career guidance for later that respects her current focus on friends and hobbies

Pain points

Balancing the search with school creates stress; she wants a streamlined solution
No real career orientation at school, so university feels like the only known path
Pressure to plan her future while time is short before graduation

Survey results bar chart skewing toward agreement — Survey responses skewed toward wanting a more engaging, rewarding app.

Bar chart of the competing apps students named most often — The apps students named most often set the teardown shortlist. Duolingo proved the mascot concept; KnowUnity and Scoolio had none, an open gap for berry2b.

Word cloud of features students said would bring them back, dominated by the mascot, daily quests, and daily streak — What students said brings them back to their favorite apps, sized by how often it came up.

The Frame

The brief collapsed into a single question: how might we make berry2b engaging enough that students come back on their own? Research pointed at five directions worth testing, a guiding mascot, a customizable play space, quests, achievement badges, and notifications that read like a person rather than a system.

Prototype flow map wiring the candidate directions into an onboarding and gamification flow — Each direction traced back to something students said, then wired into an early flow.

Co-Design

Rather than design these in isolation, I ran four participatory workshops with twenty-five students who sketched avatar outfits, proposed quests, and reacted to early mascot directions. The students weren't validating finished work, they were generating it. Their pull toward a guiding character is what turned berry from a nice-to-have into the spine of the whole system.

A participatory workshop session with students, faces blurred for privacy — Twenty-five students across four sessions, co-creating rather than reviewing.

Hand-drawn phone wireframes sketched on a notebook page — Screens started on paper. Wireframes were sketched by hand before anything touched XD.

From sketch to mascot

berry began as pencil sketches, several of them drawn by the students themselves. A berry-themed character emerged as the favorite, then went through design iterations, five-second tests, and an Instagram poll before the final concept landed.

Pencil sketches of early berry character ideas next to the final colored berry illustration — First drafts to final concept: the sketches that became berry.

What I Shipped

Five features shipped, each traceable to something students asked for. I tested every direction before it landed, five-second reads on avatar designs, an Instagram poll on the mascot, and prototype rounds returned to the same students who co-designed them.

Berry the mascot in ten different outfits and moods — One character carrying onboarding, celebration, and notifications.

Playzone screen showing Berry, an XP bar, and hunger and future meters — The Playzone, where students dress their avatar with outfits they earn.

Two quest screen iterations listing repeatable, one-off, and done quests with coin and item rewards — Daily, one-time, and done quests, reframing routine tasks as rewarded goals.

Six achievement badge designs with German labels, from Lehrling (Apprentice) to Grossmeister (Grand Master) — Badges tied to real milestones, from Lehrling up to Grossmeister.

Animated Playzone screen where berry eats a snack and reacts — Lottie animations keep berry alive. Feeding him was the students' idea.

The five features

berry, a mascot that guides onboarding, celebrates wins, and fronts every notification
Playzone, a customizable space where students dress their avatar with earned outfits
A quest system that turns routine tasks into short, rewarded goals
Achievement badges that mark real progress against milestones instead of empty rewards
Notifications rewritten in berry's voice so nudges read as friendly rather than nagging

Quiet design choices

Lottie micro-interactions gave berry enough motion to feel alive without bloating the app
Badges attach to milestones students actually care about, so progress stays honest
Every gamified surface was co-designed and tested before it shipped, not added on faith

Handoff

Development received finalized mockups, a gamified icon library, and a refreshed set of reusable UI components, so the new features shipped as one coherent system rather than bolted-on extras.

The gamified icon library that shipped with the handoff.

Result

A student holding a tablet showing the Playzone with Berry — The same app, now with a reason to come back.

A five-star App Store review praising the app — Reviews named berry and the new interactivity as standout improvements.

Defensible numbers

Thirty-day retention rose from 8 percent to 12 percent, measured two months post-launch, landing inside the category benchmark
Store rating rose from 3.9 to 4.3
Reviews named berry and the new interactivity as standout improvements
Grounded in 104 survey respondents, 25 workshop co-designers, and 12 focus-group participants

What I won't overclaim

The research cohort skewed regional and toward students already familiar with the app, so the retention lift is real but measured on a friendly audience.
There was no long-run retention curve past the two-month window at writeup time.
The lift is attributed to the redesign as a whole, not isolated to any single feature.

What I Learned

Engagement wasn't a visual problem, it was a relationship problem. The reframes below carry into any product trying to earn a return visit.

Carried over to the next system

Co-design beats designing-for. The strongest feature in the app was the one students kept reaching for, not the one I planned.
Gamification only works when it's tied to real progress. Badges for nothing get ignored; badges for milestones get chased.
A mascot is an interface, not decoration. Giving nudges a consistent voice changed how students received them.

What I'd do differently

Recruit outside the existing user base to test whether the lift holds for cold users, not just familiar ones.
Instrument the funnel to attribute retention to specific features instead of the redesign as a whole.