The client
PulseFit is a fitness channel — 400,000 subscribers, weekly uploads, and a library of professionally-shot footage undermined by inconsistent editing. Every video looked like it came from a different channel, and the analytics showed it: 62% of viewers left before the 30-second mark.
The challenge
The retention graph told a precise story:
- The first 15 seconds were wasted on slow brand animations viewers skipped anyway.
- Pacing was flat — no pattern interrupts, no visual payoff rhythm, dead air between sets.
- Zero brand memory. Nothing in the visual language repeated: colors, type and stinger styling changed with each editor's mood.
PulseFit didn't need prettier videos. It needed a system — a repeatable motion language any editor could execute.
Our approach
Audit (week 1). We annotated the retention graphs of 30 videos frame-by-frame against edit decisions. Every drop-off correlated with something specific: a 4-second logo sting, a pace sag at 0:45, a text overlay nobody could read on mobile.
Motion identity (weeks 2–3). A motion system built on PulseFit's existing brand: kinetic type scale, energy-based color states for work/rest segments, a 0.8-second signature sting (down from 4), and lower-third rules that stay legible at 320px wide. Every element specified in a motion guideline document — the channel's design system, but for time.
Pacing system (weeks 2–4). Edit templates built around retention psychology: hook-first cold opens, chapter rhythm with visual payoffs every 20–30 seconds, and breathing room engineered into rest segments so intensity segments hit harder.
Handover (week 5). Twelve edit templates, a Premiere/After Effects project kit, and two training sessions with PulseFit's in-house editor. The system had to survive without us.
Results
The first six rebranded videos against the prior six:
- Watch time 4.8× higher per video in the series.
- Retention at 30 seconds up 41% — the intro is now the hook, not the toll booth.
- Subscriber growth rate doubled in the following quarter, with the rebranded series driving the majority of new subs.
- In-house editing time cut ~50% — templates answer the pacing questions editors used to solve from scratch.
Our watch time nearly quintupled after the rebrand. People don't skip the intro anymore — they wait for it. — Marcus Bell, Channel Owner, PulseFit
What we'd tell the next team
Retention is an editing decision made every ten seconds. The rebrand worked because the system makes the right decision the default one — for every editor, on every upload.