The attributes of high-growth teams, with Ruslan Nazarenko, former growth lead at Scale AI, Vimeo, Bird

I started in Ukraine’s tech scene during the early days of growth hacking. I was reading Intercom’s books before they were even published, running courses, and working with companies to implement these strategies. However, I noticed a significant disconnect between what I was reading about and the reality on the ground. While viral growth stories and YouTube success narratives were inspiring, Ukrainian startups weren’t achieving the same results by following these playbooks.

I realized that to truly understand how rapid growth worked, I needed to see it from the inside. I wanted to get a front-row seat at companies that were actually achieving the kind of hypergrowth I’d been studying. My first real exposure to this was at  Bird, and it was eye-opening. The pace was completely different from what I’d experienced before – everything moved incredibly fast. We had a mindset of moving quickly, being willing to break things, and then fixing them as we went.

The fundamental difference was in the operating culture. Through my experiences at Bird, and later at companies like Scale, I began to understand that successful hypergrowth companies share certain cultural traits that aren’t immediately obvious from the outside. It’s not just about having the right strategies or tactics – it’s about creating an environment where rapid growth is possible. This was a major revelation that shaped my approach to growth leadership.

Through my experiences, I’ve identified three critical components that set these companies apart. First, there’s an orientation toward what I call “unreasonable goals.” We set targets that everyone knows we probably won’t achieve – like tripling our revenue in a short timeframe. It’s not about hitting every goal, but about creating a mindset where extraordinary growth feels possible.

Second, and this is crucial, every successful high-growth company develops a common language around their core metrics. At Scale, we speak in hours – how many hours we need to deliver, how many hours we actually delivered. This metric alignment runs through every department. Even our support managers discuss their impact in terms of platform hours facilitated. At Bird, it was rides. At Uber, it’s rides too. This shared understanding creates alignment across the entire organization.

The third element is speed of execution. I’ve seen some companies in the Valley moving away from traditional planning cycles. They don’t plan in months or even weeks – sometimes it can be the next day. While they maintain high-level roadmaps, they’re intentionally fluid. This is dramatically different from traditional companies where you might hear, “Eleven months ago we approved this roadmap, let’s keep going,” even when market realities have changed completely.

This is where system design becomes crucial. As a growth function, our role is to drive growth, and often the most efficient way to grow is to be less efficient with resources. The key is having proper counterbalances in the system. You can’t have one person or team trying to optimize for both growth and efficiency – they’ll fail at both.

What works better is having separate functions providing different perspectives. Typically, finance serves as the counterbalance to growth initiatives. For example, we might run aggressive promotions to hit growth targets, and then finance will challenge us on the spend. 

The goal isn’t to prevent all risks but to create a framework where calculated risks can be taken thoughtfully. This balance is especially important now, as the market has shifted from the 2021 “growth at all costs” mentality to demanding more sustainable growth trajectories.

When I joined Vimeo, we were running maybe one experiment every two months. The key lesson I learned was that you shouldn’t start by being overly scientific about it. If you get too caught up in perfect methodology at the beginning, you’ll never launch anything because you’re spending three months developing hypotheses, then discovering you’ve messed up the bucketing, then realizing you don’t know how to run power analysis.

My approach is much more practical: just get started. Run three, five, ten experiments to figure out your tooling and analysis capabilities. Early on, you’ll probably have a lot of winners because product teams generally know what’s wrong – designers and product managers can look at an onboarding flow and identify obvious improvements based on experience.

There’s a pragmatic way to discover activation metrics that many companies overlook. At Bird, for example, we found that users needed to take four rides within seven days to really activate. The methodology was straightforward – looking at retention rates across different usage levels to find the biggest gaps.

The process is actually quite simple: build a retention chart in your analytics tool and create multiple segments – users who did the action once or more, twice or more, three times or more, and so on. When you spot the biggest gap in retention between segments, that’s often your activation threshold.

This is where many teams get stuck. Take Vimeo as an example – people come to the platform for various reasons, from hosting videos to livestreaming. The key is not to optimize for an average that doesn’t represent anyone. Instead, start by tracking where users came from. If they clicked a livestreaming ad, send them straight to livestreaming and build activation around that use case.

For mobile products where attribution might be harder, use the signup flow to simply ask users why they’re signing up. Then build different activation journeys based on their stated needs. Once users achieve their primary “aha moment,” you can thoughtfully introduce them to additional features based on testing different offers and observing which ones resonate.

The real challenge comes when you start having more losing experiments. That’s when you need to get more rigorous about hypothesis development and combining qualitative and quantitative data to increase your win rate. At Vimeo, we evolved to running 20-30 experiments per month. While that’s not enormous compared to larger companies, it was appropriate for our team size and resources.

One of our most interesting discoveries came from our work on screen recording functionality. Initially, we thought our activation moment was when users recorded four distinct videos. But through deeper analysis of usage patterns and timing, we discovered something crucial: it wasn’t four different videos – many users took four attempts to upload one video they felt comfortable sharing!

This insight came from looking at the distribution of time between recordings. We noticed a large cohort of users who were recording multiple times within a couple of hours – technically different sessions, but clearly working on the same content. When we talked to these users, they confirmed they were simply re-recording until they got it right.

This completely changed our optimization approach. Instead of pushing users to create more videos, we focused on helping them create better videos faster. We introduced features like editing capabilities right after recording and teleprompter functionality for preparation. It’s a perfect example of how the right analysis can lead to entirely different product solutions.

While many teams get caught up in small optimizations like button placement or copy changes, I’ve found that the biggest impacts often come from fundamental product shifts. These aren’t simple A/B tests – they’re larger experiments that challenge core assumptions about how users interact with your product.

For instance, at Vimeo, we moved beyond just testing individual features to experimenting with entire user journeys. This meant not just optimizing individual screens, but rethinking the entire flow based on user intent. Sometimes this requires more sophisticated experimental design, but the potential impact is much greater.

The key is to maintain scientific rigor while testing these bigger changes. You can still experiment with major product shifts, but you need to be thoughtful about how you measure success and account for longer-term impacts on user behavior.

The win rate of your experiments naturally decreases over time as you tackle more complex challenges. At Vimeo, we maintained momentum by balancing different types of experiments. We’d run some quick wins to keep the team motivated while simultaneously working on more complex, potentially higher-impact tests.

A stable win rate isn’t necessarily your goal – it might actually indicate you’re not pushing boundaries enough. The key is to create a culture where failed experiments are seen as valuable learning opportunities. When we started having more losing experiments, we didn’t see it as a problem. Instead, we used it as a signal to improve our hypothesis development process and get more sophisticated about combining qualitative and quantitative data.

The AI industry operates at a completely different pace and scale than traditional tech. At Scale, we’re helping companies improve their AI models, which requires rapidly scaling specialized talent – from PhDs in biology to experts across various fields. When a client needs to scale from 500 hours to 20,000 hours in days, we have to deliver.

The interesting challenge is that automation isn’t always possible because requirements change constantly. One day a client needs molecular biologists, the next day they need a different specialization. Even after significant investment, clients might completely change direction based on their model development needs. In AI right now, money isn’t the primary constraint – speed is.

The traditional playbook of gradual scaling and methodical growth doesn’t apply in the same way. AI companies are in a race to reach significant breakthroughs first, which creates intense pressure to scale quickly. At Scale, we don’t have the luxury of lengthy planning cycles. We need to be able to pivot our entire operation within days to meet changing client demands.

This intensity is reflected in the work culture. AI companies often require significantly more hours than traditional tech companies. The compensation reflects this demand, but it’s a different model of work than most tech companies are used to. It’s about being available when breakthrough moments happen and being able to scale resources instantly.

The landscape has shifted significantly from the 2021 “growth at all costs” mentality, but this doesn’t mean companies should move slowly. Instead, successful growth leaders are building systems that enable rapid, sustainable scaling. The key is creating frameworks that allow for quick decision-making while maintaining appropriate checks and balances.

Start by establishing a common language around your core metrics. Whether it’s hours, rides, or another metric, everyone in the organization needs to understand and align around these key indicators. This creates the foundation for faster, more cohesive execution.

Remember that sustainable growth requires both aggressive execution and strong controls. Don’t try to optimize for both growth and efficiency within the same function – it rarely works. Instead, build systems that allow for fast movement while maintaining appropriate checks and balances. The companies that succeed in today’s environment will be those that can most efficiently deploy resources to meet rapidly changing market demands while maintaining sustainable unit economics.

Most importantly, start small but think big. You don’t need perfect systems to begin experimenting and driving growth. Start with simple experiments, build momentum through early wins, and gradually add more sophistication to your process. Focus on building the muscle for systematic growth while maintaining the speed and agility that today’s market demands.

Connect with Ruslan on LinkedIn to follow his insights on growth, AI, and building high-performance teams.