Sign in Subscribe
SaaS Growth

Interview with Mauro Botanes, Founder of the $160K/Month Symmetry Gym App

Roger Mendoza

• 9 min read

Interview with Mauro Botanes, Founder of the $160K/Month Symmetry Gym App

Most builders do not have a feature problem. They have a guessing problem dressed up as productivity. A new AI coding tool makes it easier to build almost anything, so naturally the internet responds by building everything. Dark mode. Settings redesigns. A dashboard with seven charts nobody asked for. The app gets busier. The business stays politely unconscious.

Mauro Botanes took a more useful route with Symmetry, a gym tracking app serving Spanish-speaking markets, primarily Spain and Mexico. Within about a year of launch, the app reached roughly $160,000 in monthly revenue and around three million downloads across iOS and Android.

The interesting part is not that Symmetry has workout plans, weight logging, social feeds, streaks, ranks, and exercise videos. Fitness apps are not exactly a fresh pile of untouched snow. The useful part is Mauro's operating system for deciding what deserves to exist at all.

It is a strong model for anyone assembling an AI-powered launch system for creators, shipping a SaaS product, or trying to make a no-code startup less dependent on vibes. Build less. Measure more. Talk to people. Keep what works. Delete the rest without holding a funeral.

Table of Contents

🏋️ The Setup: A Gym App Built for a Specific Market

What is Symmetry, and who is it for?

Mauro describes Symmetry as a simple gym tracking app. Members select a workout plan, preview the workout, log weights and reps, and use exercise videos when they need help with execution.

The product goes beyond a basic lifting notebook by adding social and game-like loops. People can post completed workouts, track streaks, compare ranks by muscle group and exercise, and level up as they progress. It is closer to a training tracker with the motivation mechanics of a social fitness network.

The company started from a personal connection to the problem. Mauro grew up overweight, and training materially improved his life. Before Symmetry, he and his co-founders were fitness YouTubers. They had a community and wanted to help people beyond the boundaries of content. Building an app was the next attempt.

That origin mattered, but it did not magically solve distribution. An existing audience gave the launch a small initial spike in downloads. It did not produce the reliable, scalable acquisition engine the team needed. This is the part people conveniently leave out when saying, “Just build an audience first.” An audience can be useful. It is not a vending machine.

How did Symmetry get to $160,000 per month?

After launching on June 6, the team spent the summer testing user-generated content. Results were initially modest, at roughly 30,000 downloads. A small improvement appeared near the end of the fourth quarter, then demand jumped in January, which makes intuitive sense for a gym product. Growth later stabilized.

The real acquisition engine was a scaled UGC operation. Over one year, the team published nearly 80,000 videos through more than 500 accounts and generated nearly one billion views. That scale is not a cute side project. It is an actual media machine.

More importantly, the team segments content by formats and sub-formats. Rather than calling a video “good” because it got traction, it categorizes what happened in the creative. One effective format showed a phone lock screen with a breakup-style notification followed by a Symmetry reminder to go train. That particular emotional contrast performed extremely well, with one example reaching 17 million views.

Creators receive nearly half of revenue, which gives them a reason to stay, produce more, and care about the long term. It is a costly approach if treated casually. But it is also more honest than expecting creators to deliver an endless stream of profitable content for exposure and an emoji.

🧪 The Magic: Stop Building Features and Start Running Experiments

What were the three product-building phases?

Mauro is unusually blunt about Symmetry's evolution:

  • Phase one: Build what the founders wanted. Terrible idea.
  • Phase two: Build exactly what people requested in feedback. Mediocre.
  • Phase three: Use a methodical, scientific process to find useful product changes.

The first phase is familiar. Founders build the feature they personally want because they are available, emotionally invested, and have access to the codebase. Conveniently, none of those things prove impact.

The second phase is more subtle. User feedback matters, but direct feature requests are often proposed solutions, not clean descriptions of underlying problems. A customer asking for a dark theme may be asking for less eye strain, a better nighttime routine, or simply repeating something they have seen elsewhere. If a team blindly builds the literal request, it can spend weeks carefully solving the wrong problem.

The third phase begins with behavior. Where do people get stuck? What action predicts retention or revenue? What is the actual problem? Only then does a potential feature enter the room.

Why do some screens matter far more than others?

Not all product work has equal leverage. That sounds painfully obvious, yet product roadmaps routinely pretend each item is a contestant in a fairness competition.

For Symmetry, the exercise search screen is a high-impact area because it materially affects activation and retention. If someone cannot quickly find an exercise and begin recording a workout, the product fails at its central job.

By comparison, the screen users see after onboarding, where workouts are displayed, received multiple redesign attempts with no measurable impact. The designs were substantially different. The metrics did not care. That is valuable information. It tells the team not to keep “improving” that screen because a founder is bored on a Tuesday.

One small commitment screen at the end of onboarding performed very differently. It asked people to commit to working out. Adding that single screen increased retention by roughly 3%.

This is the core lesson: an AI funnel builder or product builder can generate fifty variations before lunch. It cannot tell which variation changes behavior unless the system measures behavior. Speed without instrumentation is just faster guessing with better typography.

📈 The Metrics: Pick a Direction Before Picking a Feature

Which metrics does Mauro use to guide product decisions?

Symmetry uses ambitious quarterly OKRs, meaning objectives and key results, to create direction. Mauro calls the 90-day planning cycle a “warm-up,” because the team treats the quarter as a focused campaign. Marketing objectives, product objectives, and core metrics are reviewed weekly and often daily.

His argument is simple: newer products should set more ambitious improvement goals than mature giants. A fresh app may be able to improve conversion by 10% in 90 days because it probably still has obvious friction and unfinished fundamentals. Netflix does not have that luxury. It has likely already tested most of the obvious buttons.

For Symmetry, activation is especially important. The app's activation moment is not merely opening it or liking a workout plan. It requires someone to go to the gym and actually log a workout. That is a high-friction behavior change, whether the person is a beginner unfamiliar with training or an experienced lifter accustomed to another method.

Other apps have different activation moments. A calorie tracker may activate someone when they snap a meal photo. A writing app may activate someone when they publish a first draft. The product's job is to identify the action that predicts a real habit, not celebrate a vanity event because it is easy to count.

The primary metrics Mauro tracks include:

  • Downloads: A core growth input, especially early in an app's life.
  • Activation: The meaningful first action that indicates a person reached value.
  • First-week retention: Whether people return shortly after starting.
  • 30-day retention: Whether the product has started becoming a habit.
  • ARPU: Average revenue per user, or how much revenue each acquired user generates.

ARPU connects product, monetization, and acquisition. It informs how much Symmetry can afford to pay creators for downloads. It also varies meaningfully by country, so tracking it as one global average can hide the actual economics. A growth strategy that works in one market can become a small accounting fire in another.

What should a new app track first?

Track the funnel from the beginning, especially the steps closest to first value. Mauro's rule is practical: the earlier a step occurs in the funnel, the more granular its tracking should be. Onboarding should be measured screen by screen.

Without events, a team cannot answer basic questions like where people drop off. At that point, feature planning becomes an expensive form of fan fiction.

Quantitative analytics reveal what is happening. A funnel might show that many people abandon onboarding on the workout setup step. Qualitative research reveals why. Talking with users, reading posts in communities such as Discord or Reddit, and digging into real complaints can expose the reason: confusing language, missing workouts, unclear value, or an expectation the product failed to meet.

Neither data type is enough alone. Analytics tells a team where the wound is. Conversations help identify whether it is a paper cut, a broken bone, or a feature that has been quietly insulting everyone for months.

🔬 The Experiment Engine: A Repeatable System for Better Decisions

How does Mauro structure an A/B test?

The process begins from the problem space, not the solution space. That distinction protects teams from jumping directly to, “We should add a widget,” before confirming whether the widget-shaped problem exists.

Mauro's A/B testing workflow looks like this:

  1. Define the problem. Identify a specific behavior or funnel drop-off that matters.
  2. Confirm it is real. Use data and user conversations to validate the problem.
  3. Create hypotheses. List possible solutions and predict how each may move the target metric.
  4. Set guardrail metrics. Decide what must not get worse while pursuing the primary result.
  5. Run a control and variants. Compare the existing product experience against new versions.
  6. Run long enough. Symmetry generally lets experiments run for at least two weeks.
  7. Analyze and decide. Keep the winner, reject the loss, or mark the result inconclusive and move on.

The team maintains an internal system called Evelyn, short for Experiment Velocity Engine. It pulls together user interview insights, problem statements, hypotheses, experiments, and results. PostHog tracks the behavior data for tests.

This is worth copying even in a smaller, simpler form. A Notion database, spreadsheet, or plain document is enough to begin. The important thing is creating institutional memory. If a test fails, log it. A failed experiment is only wasted effort if the team forgets it and runs the same idea again six weeks later with a new gradient.

What does an inconclusive experiment tell a team?

It tells the truth, which is already more useful than most roadmap meetings.

Symmetry tested two variants of the screen immediately after onboarding. Despite major design changes, nothing moved. The result was not a reason to panic or endlessly tweak the color palette. It was a signal to shift focus toward higher-leverage parts of the product.

Most experiments do not move the needle. That is normal. It is also the point. Each test improves the team's sense of where leverage actually lives. Over time, a builder starts recognizing which product surfaces are likely to matter and which are mostly decoration wearing a business-casual blazer.

🛠️ The Lean Solo Builder Tech Stack

Which tools does Symmetry use to run experiments and move quickly?

The stack is not complicated for the sake of being complicated. Mauro and the team use AI tools broadly, including Claude and MCP connections to access operational data from tools such as PostHog, Superwall, and RevenueCat.

  • Claude with MCPs: Used to work with connected product and business data.
  • PostHog: Used for analytics and A/B test tracking.
  • Superwall: Used for paywall experiments and experimentation workflows.
  • RevenueCat: Used for subscription and revenue data.
  • WhisperFlow: Used for voice-based computer interaction, which Mauro finds faster than typing.
  • Notion: Used for documentation and experiment tracking.
  • Slack: Used for team communication.

This is a useful reminder for anyone planning to launch with AI tools: the stack should serve a clear loop. Capture evidence, organize decisions, run tests, and learn. Adding six more AI agents does not improve that loop. It mostly improves the odds of receiving a very confident Slack message from something that cannot do push-ups.

🧱 The Real Talk: Long-Term Games Beat Clever Launches

What would Mauro tell his younger self before starting Symmetry?

Play long-term games with long-term people.

Symmetry launched about a year before this conversation, but the company began two years earlier. The founding team was not good at the beginning. They kept going anyway, improved the product, improved their own skills, and kept taking another swing.

That is the unglamorous part of a $160K-per-month app story. Sustainable results came from pairing a scalable acquisition system with a disciplined product-learning loop, then giving both enough time to compound.

An AI-powered launch system for creators should work the same way. Use AI to produce faster, inspect the funnel more closely, and turn research into experiments. Do not use it as an excuse to manufacture features at industrial speed without a reason.

What can a solo builder test tonight?

Start small and make it unavoidably concrete:

  • Write down the one action that proves a new person received value.
  • Map every step required to reach that action.
  • Instrument each early funnel step with events.
  • Look for the biggest drop-off.
  • Speak with a few people who reached that point and ask what happened.
  • Write one hypothesis that could improve the metric.
  • Test the smallest credible version against the current experience.

That is the build. Not a $97 playbook. Not a heroic weekend redesign. One documented problem, one measurable hypothesis, one controlled test. Then repeat until the product gets harder to ignore.

❓ FAQ

What is Symmetry?

Symmetry is a gym tracking app for Spanish-speaking markets that helps people follow workout plans, log weights and reps, learn exercises, maintain streaks, compare ranks, and share workouts socially.

How much revenue does Symmetry make?

Symmetry reported roughly $160,000 in monthly revenue after about a year from launch, with around three million downloads across iOS and Android.

What product metrics should a new app track?

Useful core metrics include downloads, activation, first-week retention, 30-day retention, and average revenue per user. The exact activation metric should reflect the meaningful action that delivers value in that specific product.

How long should an A/B test run?

Mauro generally runs experiments for at least two weeks before analyzing results. The test should compare a control version against one or more variants while monitoring the target metric and guardrail metrics.

Why combine quantitative and qualitative product research?

Analytics identifies where people drop off and what behavior is occurring. Conversations with users and community feedback explain why the behavior is occurring. Together, they create stronger hypotheses for product experiments.

Sign up for more like this.

Enter your email
Subscribe