From Bumble to breakthroughs: Lea Samrani on product growth, monetization & scaling success

Bumble was a well-oiled machine when it came to growth. Every decision was structured around clear goals, and growth was embedded across every team, not just within a dedicated growth function. The way I see it, in a truly great company, everyone is responsible for growth. You do not have a siloed growth team working in isolation. Instead, you have specialized teams focusing on specific areas like retention, monetization, or acquisition, but all contributing to the same North Star.

The most successful companies I’ve worked with all shared common traits. Each of them was deeply data-driven, constantly running A/B tests, optimizing flows, and making incremental improvements. There was no guessing or gut-feeling decision-making. Every experiment was backed by numbers, and we had the scale to test changes rapidly. But what made the biggest difference was the ability to execute at speed. Many companies get stuck in endless debates over experiments. Successful teams move quickly. If something works, we scale it. If it fails, we move on.

Another key advantage is the structured approach to user behavior. Deeply understanding the different stages of the funnel and knowing where to focus at any given time. Growth is not just about acquisition, it’s about activation and retention. Many companies think that increasing top-of-funnel traffic is the solution, but if your activation rate is low, you are just pouring users into a leaky bucket. We spent a lot of time improving those early interactions to ensure users stuck around.

Another major factor was the brand positioning. Bumble was not just another dating app. It had a clear identity, which helped drive organic adoption. It stood for something, and that message resonated. The market timing also played a role. Online dating was already growing fast, and Bumble captured a cultural shift where women wanted more control over their interactions. While some companies rely only on product-led growth, Bumble had a strong mix of brand, positioning, and product optimization. I believe that combination gave it an edge.

One thing I always tell teams now is that growth is not about finding one big hack. It is about consistent, compounding improvements over time. The companies that get this right do not just chase viral moments. They build solid foundations, optimize continuously, and execute faster than their competitors. 

It was a frustrating but important lesson. I used to believe that better-looking designs would naturally perform better. Cleaner layouts, smoother animations, and polished interfaces should, in theory, create a better experience. But time and time again, the numbers told a different story. Some of the most visually appealing redesigns failed when put in front of real users.

One example that stuck with me was an experiment we ran on a dating app. We redesigned the user profiles to look more modern and aesthetically pleasing. The images were larger, the layout was cleaner, and everything felt more premium. We were convinced it would improve engagement. Instead, the new version performed worse. Users were less likely to interact with profiles and spent less time browsing.

At first, it was hard to understand why. The new design was objectively better looking. But when we dug into user behavior, we found that the old version had small but important details that encouraged interaction. The previous layout had more visible call-to-action buttons, clearer differentiation between sections, and subtle visual cues that guided users toward the next step. The new design, while beautiful, was less effective at nudging users toward action.

Another time, we optimized a landing flow for an app. The original version was basic, with minimal design. The new version was visually stunning, with animations and a more refined aesthetic. Again, we expected a conversion lift. Instead, the new version crashed performance. The friction created by animations, increased load times, and an overdesigned layout made it harder for users to move through the flow. The original, “ugly” version won because it was direct, fast, and to the point.

What I took away from these experiences is that good design is not about making something look nice, it is about making something work. Aesthetic improvements that do not support usability are just distractions. This is why companies like Amazon and Booking.com stick with interfaces that look outdated by modern design standards. They prioritize function over form, because function is what drives revenue.

It can be tricky, especially in companies that value brand aesthetics. Designers naturally want to create something beautiful, and marketing teams often push for polished visuals to maintain brand consistency. But if you only optimize for aesthetics, you risk hurting performance.

The key is to separate design improvements into two categories. There are changes that are purely visual, like typography adjustments or updated color schemes. Then there are functional design changes, where the goal is business-led. The mistake many teams make is treating both types the same. If a change is purely visual, it should not be assumed that it will improve conversion.

What I recommend is always tying design changes to a measurable outcome. Before making a redesign, ask: what is the specific problem we are solving? Are we addressing a usability issue, or are we just refreshing the look? If it is the latter, you need to be very careful, because visual changes can introduce unintended friction.

When testing design updates, it is also important to focus on the right metrics. Many teams only track top-level numbers like click-through rates or bounce rates. But the real impact of a design change is often deeper. Are users getting to their goal faster? Are they engaging more with the key features? Are they dropping off at unexpected points? A design that looks better but slows users down is a failed experiment, even if engagement initially looks positive.

This is also why I push for incremental design changes instead of full redesigns whenever possible. If you change too many things at once, it becomes impossible to know what actually impacted performance. Instead of overhauling an entire user flow, test smaller elements one at a time. See what works, then layer improvements on top of that.

What I see over and over again is companies falling into the same trap. They invest heavily in a visual overhaul, only to realize too late that they have removed key UX elements that helped conversion. At that point, they have to scramble to reintroduce those elements. It is why so many major redesigns fail and why teams should always test in small steps rather than go all-in on a new look.

It starts with having a clear model. You need to define your North Star metric and your growth model to understand the mechanics of your business. What’s your rate limiting step? At what stage are you losing users? What prevents people from achieving their “Jobs”? Without answering these questions, you are just making random optimizations instead of focusing on what truly moves the needle.

The first step is mapping out the funnel and once you map your growth model, map out your customer journey, and understand where the two match. You can then start identifying the biggest drop-off point and focus on them one by one using insights, and data to feed your hypothesis and quick iteration as execution. Then you look at benchmarks. These are not universal because every industry has different norms, but you need a reference point. If you are working on an app that requires habituation user activation, and your day-one retention is below 20 percent, you have a serious problem. You are losing too many users too early. If you are somewhere between 20 and 40 percent, there might still be room to optimize. If you are at 90 percent, then it is time to shift focus further down the funnel.

Another way to think about it is in terms of leverage. If you are seeing strong retention but only a small percentage of users are making it past onboarding  activation, your biggest opportunity is that top of funnel improving activation. On the other hand, if activation is solid but long-term retention is weak, that is where you need to focus. The key is to be systematic about it.

A mistake I often see is teams getting distracted by feature requests from users. You will always have people asking for new features, but if 90 percent of your users are not even making it past day one, none of those features will matter. The priority should always be fixing the biggest leaks first.

First, you need a clear understanding of what game you are playing. Different products have different dynamics, and that affects what metric matters most. Amplitude’s former CEO wrote a great piece on this. Are you building a transactional product such as a marketplace, or a habituation product that requires a high level of recurring engagement? Once that’s done and you pick  your North Star, that should reflect the user value you provide and predicts revenue.

For example, if you are building an AI-powered chat conversational product that is monetized per usage, retention might not be the most important metric. Engagement could be a better indicator of success, looking at how many chats a user sends per day session for instance. On the other hand, if you are working on a consumer subscription app, that requires user habituation, such as a learning app or a fitness app retention is crucial. If users are not coming back, they are not going to convert from that trial or renew their subscription.

Understanding what early indication of success means for your business model is key to decide if you need to optimise or pivot.

Once you define the right North Star, you need to break down your user funnel. A lot of teams assume that activation ends when a user completes onboarding. That is not true. Activation is when a user reaches the point where they are likely to stick around. You need to figure out what that moment is for your product. Maybe it is sending seven messages in the first four days. Maybe it is completing a certain number of actions within the first week, maybe it takes months. If you do not define that, you will not know what to optimize.

If the funnel is weak across the board but you have found some level of PMF, you start at the top and work your way down. If everything is below market benchmarks (not just activation, but retention and monetization as well) it is a sign that the product itself may not be working, you don’t have PMF. That is when a pivot might make sense.

But teams need to be careful about pivoting too soon. Many companies try to fix retention when their real problem is activation. If 90 percent of users never get past the first experience, they never had a chance to become retained users. A lot of the time, what looks like a retention problem is actually an activation problem.

The other mistake is teams chasing growth tactics before they have a product that retains users. I see this a lot with product-led growth. Startups get excited about building referral programs and virality loops before they even have strong retention. If users do not love your product, they will not refer anyone, no matter how many rewards you offer. The priority should always be getting the basics right before layering on additional growth levers.

When you approach growth in a structured way, it becomes much easier to know where to focus. The mistake many companies make is optimizing blindly, running random tests without a clear framework. The best teams have a disciplined process. They know what they are measuring, they set clear benchmarks, and they optimize systematically.

The market is definitely shifting. Subscription has been the dominant model for years, but we are seeing cracks in it. Paid acquisition is getting more expensive, user behavior is changing, and a lot of consumers are feeling subscription fatigue. People are more conscious of recurring charges, which makes retention harder.

That being said, I do not think subscription is going away. It still makes sense for certain types of products, especially those that build habits over time. But for many apps, the traditional model of locking users into a monthly fee is getting harder to sustain. More companies are experimenting with credit-based and usage-based pricing, where users pay based on how much they engage with the product.

We have already seen this model work in AI-powered tools. Many AI products charge per action, like generating an image or running a query. It makes sense because these tools have a clear cost per use. Users understand that every action consumes resources, so they are more willing to pay incrementally rather than commit to a fixed subscription.

In the consumer space, we are also seeing experiments with pay-as-you-go models. A good example is the rise of microtransactions in digital content. Some apps break up content into small segments and let users unlock pieces individually rather than paying for full access. We saw this with a category of apps that offer short-form video storytelling, where users pay for each chapter. These models have scaled incredibly fast, though they are not for every type of product.

The biggest challenge is predictability. Subscription models give companies steady, recurring revenue. That makes financial planning and growth forecasting much easier. When you move to a pay-as-you-go model, revenue becomes more volatile. You depend on users continuing to engage, which adds a new layer of complexity.

Another issue is user psychology. Many people prefer a fixed cost because it gives them certainty. With a subscription, you know what you are paying each month. With a usage-based model, the cost fluctuates. Some users might spend more than they expected, which can create friction. Others might hesitate to use the product as much because they are constantly thinking about the cost.

That is why execution is critical. If a company wants to introduce pay-as-you-go pricing, they need to communicate the value clearly. Users need to feel like they are paying for something tangible. If the pricing structure is too complex or feels unpredictable, it will create hesitation.

Another risk is that companies underestimate how difficult it is to build retention in a usage-based model. Many subscription apps get most of their revenue from users who are not fully engaged but keep the subscription running. If you switch to pay-per-use, those casual users might disappear entirely. That is why I think hybrid models will become more common. Some companies will offer a low-cost subscription with the option to buy credits for premium features. Others might mix free access with a usage-based component for advanced functionality.

We are also seeing some categories returning to ad-supported models. Platforms like Netflix and TikTok have introduced ads into their revenue mix instead of relying entirely on subscriptions or transactions. That is another potential direction, especially for apps that struggle with retention but see high traffic.

There is no universal answer. The right model depends on the product, the market, and user behavior. Some apps will continue to thrive with subscriptions, others will experiment with credits, and some will blend different approaches. What is clear is that companies can no longer rely on the old playbook. The monetization landscape is shifting, and the companies that adapt quickly will have an advantage.

I would take an AI-powered builder tool to speed up development. Then I would take Slack to communicate with the rest of the world. For the last one, I would probably go with Amplitude because having solid analytics is critical.

Right now, I am listening to The Financial Feminist and DOC Lipstick Jungle, which is a French podcast. 

The easiest way to turn an opponent into an ally is to ask them for advice. That has been a game-changer for me. When I have had disagreements with people, instead of fighting, I have asked for their perspective. It turns the relationship into a collaboration instead of a conflict.

Do not be your own blocker. Apply for that job. Build that product. The world will say no to you many times, but it is not your job to reject yourself. Just go for it and let others decide.

 I have too many, but right now I would say The Breakfast, which helps people meet over breakfast. I also really liked using Airalo while traveling. It made managing eSIMs so much easier.