I rebuilt my body with the app I built—but I still haven’t figured out how to grow the app

Over the past 90 days, I’ve used the fitness app I built to completely change how I train, eat, and track my progress.

The app worked for me. The marketing didn’t.

Reset90 AI is now live on the App Store and it totally free. And here is the link. But after nearly a month, it has only attracted a small number of downloads and almost no active users.

So this is not a success story.

At least, not yet.

01Why I Built Reset90 AI

The idea for Reset90 AI came from a YouTube transformation video.

A coach guided someone through 90 days of training, nutrition, and accountability—and by the end, the transformation was incredible.

Watching that made me wonder:

What if an AI could become that coach?

Not just telling you what workout to do, but guiding your entire journey.

What to eat.

How to train.

When to adjust the plan.

And encouraging you to stay consistent when motivation fades.

I’ve been going to the gym for nearly years, yet I was never truly satisfied with my own physique.

So before trying to build it for anyone else…

I decided to build the coach I wished I had—and test it on myself first.

02Where the App Is Today

The app is live and it is free.

But being live and being used are two completely different things.

At the time of writing:

• Downloads: 23
• users: 20
• Revenue: $0 (free app)
• Marketing experiments: more than I expected
• Confidence: changing every day

03Why I’m Sharing Before the App Succeed

Once someone succeeds, every decision they made can start to look like the right decision.

They used Strategy A, B, and C—and because the outcome was successful, it is easy to assume those strategies caused the success.

But success is not always repeatable.

Timing, luck, audience, product, execution, and dozens of invisible factors all matter.

I haven’t grown Reset90 AI successfully yet, so I’m not writing this as an expert.

I’m writing it as someone running the experiments in real time.

Success stories can inspire you.

Failed experiments can sometimes teach you exactly what to avoid.

I’m not building in public because I have the answers. I’m learning in public because I don’t.

04My Growth Experiments So Far

Before diving into the different channels I’ve tested, here’s one principle that has guided almost every experiment I’ve run:

Educate people or entertain them. Ideally, do both.

People don’t open TikTok to watch an ad.

They open it to learn something, laugh at something, or be inspired.

If my content doesn’t do at least one of those things, it probably won’t get their attention.

That philosophy shaped every growth experiment I’ve tried so far.

Strategy 1 — Running My Own TikTok Accounts

Experiment A. Green Guy Fitness

My first approach was to build a recognizable fitness character instead of relying on my face.

I created Green Guy—an original fitness character designed to teach simple workout lessons to busy people.

Every day, Green Guy shares simple workout tips, fitness knowledge, and practical advice that people can learn from—even if they’ve never heard of Reset90 AI.

So far, the account has been growing by around 10 new followers a day, with many posts reaching 1,000+ organic views.

A well-designed character alone doesn’t become an IP—people have to keep coming back because the content is valuable.

Feel free to check the Green Guy fitness if you are interested.

Experiment B.Exploring AI-Generated Content

Another area I’m currently exploring is AI-generated content.

I’ve been amazed by some of the AI-generated fitness videos I’ve seen online. The quality is getting incredibly realistic, and it’s made me wonder how AI could change the way fitness content is created.

Strategy 2 — TikTok Influencer Outreach

Another strategy I’ve been testing is working with TikTok creators.

Like many founders, I thought finding creators would be one of the fastest ways to reach new users.

It turned out to be much harder than I expected.

Some of the challenges I’ve run into so far:

• Many outreach accounts became restricted, making it difficult (or impossible) to send DMs consistently.
• Even when creators replied and we had positive calls, some eventually disappeared or stopped responding.
• Finding creators who genuinely fit the product has been much harder than simply finding creators with followers.

Strategy 3 — Reddit

Reddit was another channel I experimented with.

I tried writing organic posts—asking for fitness advice, sharing parts of my own transformation journey, and discussing AI prompts—without directly promoting Reset90 AI.

Some of those posts actually performed much better than I expected. A few reached 10,000–20,000 views organically.

However, views didn’t translate into downloads.

The moment I mentioned my app or included a link, the post was often removed. In some cases, my account was also restricted, making future posting even more difficult.

My biggest takeaway is that Reddit rewards people who are already part of the community, not new accounts trying to promote a product.

I’m not saying Reddit doesn’t work—it clearly works for many founders.

But in my case, it hasn’t been an effective acquisition channel so far.

05What’s Next

One thing these experiments have taught me is that building a product and building an audience are two completely different skills.

TikTok is where I’m experimenting with short-form educational content.

But I want X to be something different.

Instead of trying to teach fitness in 30-second videos, I want to document what it’s actually like to build an AI fitness startup while rebuilding my own body.

Not as an expert.

Not because I’ve figured everything out.

But because I’m figuring it out one experiment at a time.

What I’ll Be Sharing

Going forward, this account will mainly document two ongoing journeys.

1. Rebuilding My Body

• My workouts
• Nutrition and meal tracking
• Weekly progress
• What actually worked for me
• The habits that helped me stay consistent

Not generic fitness advice—just honest lessons from my own journey.

2. Building Reset90 AI

• Product decisions
• Marketing experiments
• Influencer outreach
• App Store optimization
• AI content creation
• Wins, failures, and everything in between

If something works, I’ll share it.

If something completely fails, I’ll share that too.

06Final Thoughts

I don’t know whether Reset90 AI will become a successful business.

I do know that building it has already changed the way I train, eat, and think about consistency.

Now I need to learn whether I can turn something that worked for me into something useful for other people.

I’ll share the downloads.

The experiments.

The failures.

The creators I reached out.

And hopefully, over time, the things that finally work.

I’m building Reset90 AI in public.

But more importantly—

I’m learning how to grow it in public.

I’ll keep sharing the real numbers, the experiments, and the lessons.

If you’re on a similar journey, leave a comment and let’s learn together.