Gismart: What Happens After 1 Billion Downloads?

In Retention Podcast #65, we talk with Iryna Moskalenko, CMO at Gismart and an expert in User Acquisition and Product Marketing.

Gismart is one of those companies whose apps have probably lived on more phones than most people realize. Back in 2013, the team started with music products, then grew into Health, Wellness, Utilities, and Lifestyle, building a portfolio that now includes Luvly, Dancebit, FitMe, Famio, Beat Maker Go, and DJ it! Across their ecosystem, Gismart has recently reached 1+ billion downloads worldwide.

Here’s what stood out from our talk.

From Centralized Marketing to Product-Led Focus

Gismart looked very different when the company first started shifting into a more focused product structure. At first, product marketing covered several apps at once: music apps, a GPS utility, and a Health & Fitness product. It was a useful learning ground, especially for building knowledge beyond user acquisition and getting closer to onboarding, paywalls, and product communication.

Over time, the company moved toward a more focused model. The Wellness unit became a separate direction, with Luvly, a face yoga app built around a holistic approach to beauty, becoming one of the key products. Later, Gismart moved through another stage of restructuring and started giving separate products, or product groups, their own CEOs and marketing teams.

Today, Gismart works more like an ecosystem. Each product has more ownership, while the core company provides shared support: billing, infrastructure, recruiting, customer support, and other operational services. It’s a shift from one central team trying to manage many different apps to a model where each product can grow with its own focus, leadership, and priorities.

Why Dedicated Product Ownership Matters

Scaling existing apps is important, but it can’t be the only growth path. At some point, the business also needs to test new directions, launch new products, and look for fresh opportunities inside the categories where it already has expertise.

That thinking led Gismart to R&D apps inside the Health & Fitness ecosystem. The team wanted to build something new, not just keep improving the products they already had. That’s how FitMe appeared: a general fitness product created as a new bet in a familiar market.

The key decision was to give FitMe its own dedicated marketing lead instead of treating it as one more product in a shared portfolio. Since Gismart was still mainly a performance-based business, that role needed someone who deeply understood performance marketing and could build the right team around the product.

The result proved the model. FitMe became Gismart’s top product, showing that dedicated ownership works better both for launch and scale. When one person is fully focused on one product, owns the team, and carries responsibility for growth, the product gets much more attention than it would inside a large mixed portfolio.

Focus Before R&D

New product ideas are tempting, especially when the market is full of visible opportunities. If a company has a walking app and sees another category making money, the natural reaction is to try that too. But this is where focus becomes important. Before moving into new products or R&D, the team needs to understand whether the current use case or value proposition can be developed into something stronger.

It also helps to define success and failure conditions from the start. The team should know what result means the product is ready for the next stage, and what result means it is time to stop. Without that line, it becomes too easy to keep investing because it feels like the product is almost there, even when the numbers do not support it.

For a product to become stable, it needs more than one working angle. Two or three active use cases can give the team more room to keep testing as some ideas fade or stop performing. After that, R&D can mean a new product or a narrower experiment inside the current product, like testing a very specific course or use case once the core product is already stable.

Choosing a Niche That Has Room to Move

R&D at Gismart starts with deep product and marketing research. A crowded niche is not necessarily a bad sign. If there are many competitors, there is probably demand already. After that, the game is execution: who tests faster, uses AI better, automates more, and runs stronger performance marketing.

For new products, existing purchase intent matters. Without a large budget to educate the market, it makes more sense to enter a category people already understand and are ready to pay for.

Another important filter is flexibility. General Fitness works because it can include many smaller use cases, from different workout formats to specific courses. Mental health apps work in a similar way: ADHD, trauma, family issues, body image, parenting, and other angles can all live inside one broader product.

Founder Energy Matters Too

Market research helps narrow down the options, but a founder still needs genuine interest in the idea. A niche can look promising, have demand, and fit the company’s experience, but if the founder has already been there and no longer feels curious about it, building in that space becomes much harder.

For a new product, the topic needs to feel both interesting and promising. It also helps when there are people around who share that energy and want to build from scratch. In this case, the focus is on finding people with strong performance marketing thinking, curiosity, and a real appetite for experiments.

That marketing-first mindset matters because new products still need money to grow. The team first needs to build a base, prove the economics, and earn enough room to keep improving the product.

Paid, Downloaded, Ignored

Inside Gismart’s ecosystem, Luvly still stands out as a personal favorite. It was one of the first Wellness products the team worked on closely, so there is some history there. Dancebit is another strong example: a dance fitness app built around fun formats like high heels and country dances.

Health & Fitness is a tricky category because users want the result, but the result still requires action. People are ready to download apps and pay for subscriptions, but the hard part is getting them to actually do the workout, follow the plan, or build the habit.

That’s why many fitness products now lean into gamification: streaks, leaderboards, challenges, and playful mechanics. The goal is to make the experience feel less like a chore and more like something users want to return to. Gismart’s Health & Fitness apps are also moving in that direction, testing ways to keep users engaged without making the process feel heavy.

Different Products, Different Playbooks

Gismart’s main product groups, Health, Utilities, and Music, are very different from each other. The difference starts with user intent: why someone downloads the app, why they consider paying, and what kind of promise makes the product attractive. It also affects the channels, creative formats, and marketing angles that work best for each category.

That variety makes internal knowledge-sharing more useful. Teams can see top creatives from other product groups, borrow formats, and test whether a working idea from one category can be adapted to another. A strong Utilities creative will not copy-paste into Health & Fitness, but it can still spark a new angle.

Because the products are not too similar, the ecosystem avoids the kind of toxic competition that often appears when teams fight over the same audience or product space. The competition is still there, but it feels healthier: marketers are naturally competitive, and each product team wants to prove its own approach works.

When Search Beats the Quiz

Marketing playbooks change a lot between Health & Fitness and Utilities. Health & Fitness often performs better on Meta, where a user can be pulled in through emotion, aspiration, or a strong creative angle. Utilities are different. For them, Google Search can work especially well because the user already knows what they need.

That difference also changes the funnel. In Health & Fitness, quizzes help warm users up, personalize the promise, and move them toward purchase. In Utilities, that step is not always needed. A person searching for a cleaner, tracker, or safety app often comes with clear intent already.

This also explains why unsubscribe rates can be higher in Utilities. The user arrives with a specific job in mind. If the app does not solve it quickly or well enough, they move on and look for another option.

Learn Fast and Cheap

In user acquisition and funnels, not every failed test should be treated as a failure. Most of that work is built around experiments: testing angles, funnels, creatives, and channels to see what can actually scale.

The real question is how quickly and cheaply the team can learn. If an experiment does not work, but it gives a clear answer without burning too much budget, it still has value.

A real mistake is something more operational, like setting the wrong campaign budget by accident. Testing a funnel that does not convert is part of the process. Launching a campaign with 50,000 instead of 5,000 is a different story.

Good Tests Start With Clear Rules

A strong experiment starts with a validated hypothesis. It can come from competitor research, marketing data, user behavior, or product research. Gut feeling can still play a role, especially in creative work, but it should be the exception, not the whole basis for a test.

The next step is setting the rules before the experiment starts. The team needs to agree on what result will mean further scaling, what result will mean stopping the test, and what result will mean “keep improving.” For example, the decision can be tied to spend, ROI, subscription rate, or creative success rate.

This also keeps marketing, product, and creative teams aligned. If the goal is product net profit, each team owns its part of that result. Creative producers may test a set of 50 concepts, marketing checks spend and ROI, and the team decides whether the topic deserves a bigger funnel or should be closed.

The 178-Pre-Land Lesson

The decision to keep testing depends on the product, its stage, and the team. In R&D, even weak numbers can look promising if everything else performs worse. For example, ROI of -10 may still deserve another round if the earlier baseline was closer to -70.

The middle zone needs context. If the team spent $10,000, got ROI of around zero, and saw positive signs from competitors or user behavior inside the app, the experiment may be worth improving. If the same spend brings ROI of -30 and that was already the agreed limit, it is probably time to stop.

This is why success criteria matter. One earlier Web2App test for Luvly turned into 178 pre-landers over two or three months. Other companies were making the format work, so it felt like one more attempt might crack it. It never did. That experience made the case for written rules even stronger: decide in advance when to scale, when to improve, and when to stop.

The Morning Starts With the Numbers

The day starts with Gismart’s dashboard. Every morning, Slack sends a screenshot with the main view across all products, so the first numbers to check are predicted net profit, actual revenue, and cost. Predicted net profit is the key metric: the team sets goals around how much net profit they expect to get over 365 days, based on LTV and acquisition costs.

After that comes a more detailed daily check: product-level results, sources, and anything that looks inconsistent across channels or platforms. A deeper analysis happens less often, usually weekly, with billing metrics reviewed every couple of weeks. During Q5, the focus shifts to keeping scale after the seasonal spike. 

Bigger spend is not the scary part by itself. If the ROI drop is within an acceptable range and there is a chance to scale, it can be better to spend a little more and see how the product behaves at higher volume.

AI Makes Creative Teams More Marketing-Led

AI has made creative production faster and broader. Teams can produce more creatives, test more angles, and sometimes improve quality at the same time. But it does not remove the need for strong people. AI can help with ideas, drafts, and production, but it still needs someone who can shape the concept, sharpen the angle, and understand why it should work.

The bigger shift is in the role itself. Motion designers and graphic designers can no longer rely only on execution skills. More of the work now requires creative marketing thinking: understanding the product, the user, the hook, and the performance logic behind the creative.

That is why teams are moving toward roles like AI generalist, AI motion designer, or AI marketer. The team still grows, but the skill set changes. Designers need to think less like pure designers and more like creative marketers who can use AI as part of the production process.

AI Moves From Research to Automation

Inside Gismart, AI is already part of the way teams work, not only in creative production. Developers are taking a practical course on building AI agents, with the goal of creating tools that help their own teams work better.

In marketing, AI is used mostly for research. It helps with creative research, YouTube analysis, competitor checks, and exploring new niches. Tools like Perplexity and deep research make it easier to collect signals before testing a new angle or product idea.

The next bigger shift is likely to happen in media buying automation. Right now, a lot of automation still works through rules and algorithms. But the market is moving toward AI-based systems that can recommend what marketers should change, and in some cases even make changes themselves, with the marketer reviewing the result.

There is still some hesitation around giving AI that much control. But the direction is clear: AI is moving from helping teams research and produce faster to becoming part of how marketing decisions are made.

In a Nutshell

Gismart grew from music apps into a much wider ecosystem, and that changed the way the company works. One central team can’t treat a face yoga app, a dance fitness app, a family safety utility, and a music product the same way. Each one has its own user intent, funnel logic, creative formats, and growth rhythm.

R&D only makes sense when there is a real reason to test something new, not just because another niche looks profitable. Experiments need clear rules before the team starts spending money: when to scale, when to improve, and when to stop. AI now helps with research, creatives, and automation, but it still needs marketers who can spot a good angle, question the data, and gently take the laptop away before “just one more test” becomes 178 pre-landers.