Case 05 · Basic-Fit · Product design · Behavioral design · Adaptive workout flows

For a beginner, the gym is a psychological space before it's a physical one.

The Basic-Fit app had a design language for performance. It had none for confidence. This is a redesign of the session experience so a beginner can start, adapt, and finish a workout in a real, crowded gym without feeling watched.

Role: Product Designer: research, UX/UI
Scope: Rework on the workout experience
Type: Self-directed case study
Platform: Mobile · iOS

UX designUI designMobile app

Context

An app millions open, that most beginners never really use.

Basic-Fit is a high-volume, low-cost gym chain serving millions of members across Europe. Its app showcases many exercises and routines but for someone in their first weeks, it works as little more than a QR code to open the door.

For that beginner, the gym is less a place than a feeling: unfamiliar machines, the sense of being watched, and no idea what to do first. The app meets them at exactly that moment, nothing more. I redesigned the session experience to speak to the beginner, not the athlete the product was built for.

The problem

New users were not failing to train. They were failing to feel like they belonged.

Motivation was not the gap because beginners arrived ready. What pushed them back out was structural: the product only knew how to talk about performance.

01 · Exposure

Exposed on the floor

The app suggested floor and free-weight exercises with no sense of the room, pushing self-conscious users into moments they felt too watched to attempt.

02 · Rigidity

Plans that broke against the real gym

A single occupied machine left no fallback, so a beginner's session collapsed the moment reality did not match the plan.

03 · Friction

Preparing cost more than doing

Unfamiliar machine names and generic programs made building or adapting a workout harder than completing one.

Validated through interviews with beginner and intermediate members. Individual frustrations were grouped into shared patterns.

What happens today

Two moments break a beginner's session. The current app sees neither.

A beginner's session has a shape: they arrive motivated, hit a first break when the app exposes them on the floor, push on, then hit a second break when a machine is taken with no fallback. Today's app registers none of it.

The beginner journey timeline: arrives, first break (exposed), workout, second break (machine taken), stuck, exits.

Arrives motivated but unsure → the app prompts a floor exercise → Break 1: feels watched, momentum breaks → follows the plan → Break 2: next machine taken, no fallback → stuck → leaves, no progress recorded.

What the gym forced

Every decision here answers a constraint, not a preference.

The real gym sets hard limits: it's crowded, public, and blind to which machines are free. I let those limits drive the design instead of designing around them.

Constraint 01

Any machine can be taken at any time

The floor is shared and unpredictable. The design can never assume a planned exercise will be free.

Decision

A free-order checklist, never a fixed sequence

Pick any machine to start, in any order. A session survives a crowded floor because nothing depends on going next.

Trade-off: gives up the optimal workout for one that does not collapse.

Constraint 02

Everything happens in front of others

Floor work and free weights carry a social cost the interface ignores.

Decision

A machines-only filter that removes exposing exercises

One toggle strips out the moves a self-conscious beginner would not attempt in public.

Trade-off: a less complete workout, in exchange for one they will actually finish.

Constraint 03

No live data on which machines are free

There are no sensors. Adaptation has to be driven by the user reporting reality.

Decision

One-tap skip-and-swap on every exercise

Report a machine as taken, get same-muscle alternatives that are free. An occupied machine becomes a detour, not a dead end.

Trade-off: puts the user in control instead of waiting and breaking their flow.

The session, step by step

One flow that carries you from first workout to last.

Start without deciding

The session picks itself

One entry point on the home screen, one checklist inside. No program to build, no order to follow. A beginner opens the app and is already moving. The hardest decision, what do I do today, is already made, with easy exercises.

Home screen: a single 'today's session' pick for beginners.

Home: one "today's session" pick

The free-order checklist: pick any exercise to start.

Free-order checklist, start anywhere

Adapt to a busy floor

A taken machine is a one-tap detour

If a machine's taken, one tap opens a swap sheet targeting the same muscles. It's framed so swapping feels like a normal choice rather than a setback and the session count never resets.

Swap sheet: same muscle alternatives.

Stay inside the set

Nothing to remember, nothing to calculate

Weight is anchored to last time, so there is nothing to recall. Reps, sets and rest are one tap each, and adding a set is deliberate rather than assumed.

Reps, kg and rest anchored to last time.

Reps · kg · rest, anchored to last time

Adjusting the weight, with last time shown.

Adjust metrics easily

Add another set, on your terms

Leave with a reason to return

Small wins, made visible

The end screen counts what a beginner can control: you showed up, you added a set, you went up 5 kg from last time. Then it does the fun math: "2.5 tonnes lifted this week." Small wins made to feel as big as they are, and a quiet reason to open the app again on Sunday. A missed week is never a failure.

The end screen: session complete with sets, reps, volume, progress this time, and what you did.

Small rundown of exercises done, to feel proud.

One system, two modes

The same structure carries cardio.

No reps or weights. Timers instead, and a dark UI that reads at a glance mid-effort. Strength and cardio share their structure, their swap logic, and their voice.

Beginner-focused, with details that reassure

Same free-order checklist

Dark UI like Basic-Fit to separate active exercise

Congratulate with an easy option to make it simpler

Where the thesis shows most clearly. The end screen does not ask how hard you pushed. It asks what would make next time easier and if none of the three comfort filters are active yet, it offers them: stick to machines, stay off the open floor, keep it short. Turn any on and the app remembers, quietly shaping the next session around your choice.

Beyond the beginner flow

I explored further, then chose to stop.

Explored, not shipped

Three directions I wireframed

Past the core flow, I kept sketching. A "watch while you workout" widget that pulls up a YouTube video for the long machine cardio. A "build your own session" flow for members ready to shape their own workout. And a "profile page" to give the comfort filters a real home. All three wireframed, none shipped.

Use your phone without losing track of your workout

Duplicate an existing session to modify

A profile page with filters to make the workout smoother

Where I stopped

Restraint was the design decision

I stopped there on purpose. Chasing all three would have turned a tight beginner experience into a full app redesign and lost the one thing this project was about: getting someone new through their first weeks without friction.

An early PRD kept that boundary in view, so I could tell the explorations worth finishing from the ones worth stopping. The discipline was the design decision.

How I worked with AI

AI to explore, build, and refocus. Not to decide.

AI only ran alongside the work. It sped the process and stretched what I could try, but it was never the one steering. The judgment stayed with me at every step.

Synthesis

Claude AIChatGPT

AI did

Clustered raw interview notes into candidate patterns in minutes.

I decided

Which clusters were real, and cut the ones that only showed up because of how I'd phrased a question.

Focus on the purpose

Claude AI

AI did

Wrote up a PRD with my input, giving me a clear scope to pin down the beginner-first goal.

I decided

Kept every choice inside that scope, with no full app redesign.

The confidence voice

Claude AI

AI did

Drafted microcopy variations to help me find the right tone and UX wording.

I decided

Narrowed to one tone. AI gave the options; the voice was my decision.

Explore, then refine

Claude DesignFigma MakeCursor

AI did

Spun up flow and interaction variations to test, faster than I could sketch them.

I decided

Kept what worked, refined the rest, and made every detail mine.

The output is mine. AI made the good ideas faster to reach, but it never did the deciding for me.

Validation · usability test

Tested with 10 people and let it change the design.

Progress

What testing showed

People wanted proof they'd moved forward. A done screen was not enough to feel like the session counted.

What I changed

Added an end-of-session recap that shows their progress building over time, not just today's completion.

What's next

What testing showed

Seeing the next exercise mattered more than I expected. It kept people oriented and gave the flow momentum.

What I changed

Surfaced the upcoming exercise inside the flow, so you always know what is coming, not only what you are on.

The test earned its place by reshaping the design, not by scoring it.

The takeaway

Built so no one feels slow.

You cannot design confidence directly. What you can do is remove the moments where a beginner feels behind. The blank start, the taken machine. Clear enough of them, and what is left is someone who feels like they belong.