Training section
How it was:

you can do one workout per day if it's scheduled in the schedule;

the user doesn't know which training will be next;

you can't choose a workout by yourself, you have to do the workout offered by the platform;

change of training programs is limited (no more than five times a day);

there is additional content on the platform that has low activity.
Research
Besides the analysis of support requests and compare competitors, we did about 20+ interviews. They confirmed our main hypotheses about the training section:

people want to review the workout before taking it and choose it according to how they feel;

but not everyone has the time or desire to choose their workout, they trust what the platform offers;

new users need to evaluate the quantity and quality of content offered.
The positive thing was that the company and users are close and loyal to each other. We heard from people a sincere desire to help develop the product and almost no one was late for an interview!
The minus was that analytics was very poorly set up at this stage and we couldn't see all the data we wanted before we started working on hypotheses.
Tight restrictions in select of workouts forced our users to «cheat the system». They passed a workout, changed the schedule, and passed the next workout.
Why did we start with training?
This is the main functionality of the platform. According to analytics, users mostly go for training, and after analyzing the complaints and requests, we noticed the need to update this functionality. Also at the level of competitors, the content (workouts) holds the bar, but the functionality of the platform loses.
How it became:

after building the user flow and designing prototypes, we ran tests, made edits, and there are the changes we made to the platform;

we added a workout section;

in addition to workouts in the section appeared exercises that were previously hidden in the Knowledge Base;

we added Сhosen workouts and exercises;

descriptions to workouts and focus areas.
Results
Three months after the updates, we measured the effectiveness of the innovations:

the number of workouts completed had not changed. It shows that people were also exercising on the platform as before;

the change in training program decreased by more than three times! That is, users stopped hacking the platform to complete another or more than one workout per day;

users found the workouts because their browsing increased more than three times!