Hot: Online gym is password protected.

It's an older case study, kept here for anyone curious.

No password? Ask me for access, or go back to the homepage.

Amirhossein Shamlou

Mobile appiOS and Android2019 to 2021

HotOnline gym

Hot is a home workout app, made to get people in the Gulf who never go to a gym training regularly.

Market
The Gulf (GCC), starting with Kuwait
My role
Sole designer
Result
Top 3 grossing

App Store and Google Play links are available on request.

Hot, the online gym, on a phone
Problem
The subscription was measured on engagement and retention, for people who didn't train regularly, and we started with very little information.
What I owned
All of the design, as the only designer from 2019, with five developers and the CEO: research, the flow, wireframes and tests, the design system and the interface.
What changed
We launched in Kuwait, and Apple featured it as App of the Day a couple of months later. Until my last days with the team we kept tuning it with heatmaps, analytics and NPS scores.
Android
Sadegh Saati, Arash Azizi
iOS
Hamidreza Vakilian, Mahan Mirshafa
CEO
Nezar Al Saleh
Company
Nizek.com, Ahotlife.com
Top 3grossing in the Fitness category in Kuwait, on the App Store and Google Play
+35%engagement with class sessions after the scroll and button placement changes
2%user error rate when choosing a category, after tree testing and card sorting

The ranking is from the App Store and Google Play charts in Kuwait after launch; the other two figures are from our own tests and analytics.

The store rankings and dashboard
Store rankings and the dashboard after launch
Store rankings and the dashboard after launch

What to launch with

We started with live classes, programs and workout sessions, and left leaderboards and achievements for the next major release.

A gym owner helped me find the audience. From their data, benchmarks and the rival businesses I studied, I wrote three personas, from a housewife to a fitness enthusiast.

To see how customers would find us and use the product, I made a simple 5×5 table, explained it to the head of each department, and they filled it in from their own knowledge of the business. I merged their answers into one customer journey: a shared vision for the roadmap.

The research boards

I ran test sessions in Iran and remotely in Kuwait. Our focus groups used their own fitness app, mostly Peloton, and told me what they loved, what they hated and what they wished it had.

An isometric diagram of the product flow
An isometric diagram of the product flow
A persona card: Shima Tabatabaei
A persona card: Shima Tabatabaei
The customer journey table
The customer journey table
A study of emotional responses in the focus groups
Emotional responses in the focus groups

The tab bar we dropped

Testing the wireframes with fitness groups and our personas changed the flow a lot.

A very low-fidelity wireframe went to Kuwait first, then a mid-fidelity mockup once we agreed on the pages.

At first I put the activities (Live, Programs, Workouts) in a bottom nav bar, but a tab bar stacked on a nav bar never felt right. Moving them to the top of the page made switching between activities feel natural to the test group.

A/B tests of early prototypes, with horizontal and vertical scrolling and different button placements, showed me where people actually touch the app while they work out.

Wireframes of the key screens
Wireframes of the key screens

The video player

Most content ended in videos played in rotation, so the player was probably the most important feature.

Iterations of the video player
Iterations of the video player

It had to let people skip a session, show what was going on, and never pull attention from the workout. I built one variation on top of another until I reached one I really liked. Last came the micro-interactions, to give the app a lively, playful feel.

Micro-interactions and transitions in context

Theming without an update

As a team we came up with a way to change the live app's look from a dashboard.

NCS was a cloud system that held the color palette, videos, images, icons, fonts and sizes. Change any of them from a dashboard and every installed app picked it up without an update, so a badly performing icon or image could be swapped at once. Credit to Payam Chamani and Hamidreza Vakilian for building it.

I also kept a design system artboard with every asset as a master component, which made revisions and experiments far easier.

The app's look
#EB0100
#B9B9B9
#1F1F1F
#000000
Primary font
Montserrat
Secondary font
Source Sans Pro
Type scale
Base 16px, major third, weights 300, 600 and 900
Hi-fi screens
Hi-fi screens
I'm glad I got to work with this team.

Questions about this work? Write to me.