Can we build this video in parts?

Android
iOS
Project lead
Shipped

Freeletics is a Munich-based fitness app that delivers AI-powered, personalized training programs that fit your goals.

No gym, no equipment required. With over 54 million registered users worldwide, it's one of Europe's most recognizable fitness platforms.

“This is not going to scale if we keep doing it the same way. We need to break it down.”

[I kept saying]

Problem

The license on nine foundational exercises expired, and every video had its instructions burned in.

Every new exercise or language meant a new shoot, with no fixed studio and no consistency between shoots
New exercises cost too much to make, so new programs didn’t ship
A new language took up to two months, too slow to keep up with fitness trends

The old experience consisted of a video with instructions — each video was translated into 11 languages.

The new experience lets users view exercises from different angles and control the video playback speed.

Key contributions

Led the video system end to end, with 20+ people across coaching, engineering, product, marketing, localization and support, around one direction: build every video in parts.
Started from evidence nobody had read. I went through years of support feedback in email, Facebook and Instagram, then ran 12+ interviews myself on lighting, sound, angle and whether the movement reads, all before we designed anything.
Took the instructions out of the video. Angle and speed became player controls, so a new language stopped meaning a new production. Localization went from 2 months to 4 days.

Impact

Measured three months after launch, against the metrics we agreed with Product before the final iteration.

80%: lower total video production budget
+19.4%: weekly training retention
2×: more training programs offered
4 days: to localize a new language, down from 2 months
< 2 weeks: to produce a new video, down from a 4–5 month production cycle

Retention and program growth are shared results of the video system and Freeletics’ modular training platform, rebuilt by the wider team at the same time.

User impact

I started this project by reading years of complaints in our support inbox and social channels. After launch, the same channels started saying thank you. People named exactly what they had asked for: clearer form to avoid injuries, multiple angles, no music cutting into their playlist, and women performing the hard workouts too. The 80% saving and the happier community came from the same decision.