The live class ended, and then it was gone.
TRX was not storing its live classes. We paid to produce the workout, streamed it once, and lost the video after that. Customers were also telling us the class times did not fit their schedules.
The first product brief was basically empty. It had headings for the problem, goals, solution, open questions, and success metrics, but no actual answers.
Two days later, the answer was getting specific. Record the video stream, trainer audio stream, and timer stream. Store them. Add the finished class to the on-demand catalog. Measure whether more people started and finished workouts when the schedule stopped being the gate.
I was writing this from a barely furnished apartment in Newport Beach. The floor was still covered in boxes. I had also bought a black Mini that came from the dealership with a giant pink bow. I had chosen an automatic with Amanda in mind, even though she and Eva were still home in New York.
I texted Amanda that the apartment kind of sucked, so maybe making it eccentric would be fun.
The product list was getting eccentric too. That same afternoon, my Drive had briefs for connecting completed TRX classes to Strava, redesigning the product category pages, and building an Army Combat Fitness Test registration flow.
The stored live class became Replay.
Replay was not supposed to be another ordinary on-demand video. It would remain part of the live subscription. A replay could appear inside the live schedule as a session that people booked normally, then unlocked at the assigned class time. It could also live in its own filterable gallery.
That created a longer list of ordinary product questions. Where would Replay appear? How would people book it? Which classes were worth keeping? How could a recorded workout avoid feeling obviously old?
Nothing had shipped yet. The next job was turning the brief into software.