About the Author
Vladyslav Hrytsenko is the Chief Technology Officer of Mighty Digital, a best-in-class digital data agency focusing on data-informed growth for digital consumer products. He is an engineer at heart, and he has assisted many customers in areas such as finding growth opportunities and working with data to improve it, enhance it, process it, and analyze it.
Growth engineering is about speed, high-quality experimentation, and successful innovation through data. If data becomes your fuel for growth and analytics becomes your engine, essentially, you need a sports car—a modern data stack to drive value fast.
The art of asking “why” is a critical skill for growth engineers, who identify and act on opportunities to maximize business value.
The work of Facebook’s growth team would confirm.
In 2007, when Facebook was struggling to scale, Zuckerberg prioritized growth engineering above all else to boost the platform’s user numbers. Zuckerberg and his handpicked crew pioneered innovative growth experiments to multiply the company’s north star metric– monthly active users (MAU). At that time, Facebook was still competing with MySpace, Bebo, and HighFive, and they needed to switch gears to lure in more users across the world.
Asking “why” has been really powerful for us to leverage the platform’s vast size and expertise in data and provide valuable experiences to our users.
Alex Schultz, Vice President and CMO of Meta.
Growth engineers, who are vital levers for transformational success and growth, boost customer acquisition, retention, and revenue by introducing core product value to the most significant percentage of the target audience as quickly as possible. They are the experimentalists who determine which strategies are most effective in achieving the overall goal.
In the case of Facebook, growth engineers attracted more users and more of their time by generating growth experiment ideas, conducting systematic testing of hypotheses, and continuously breaking out of the local maximum of product performance. They adopted a scientific approach to growth instead of a gut-driven process, where they would:
- Step 1: Formulate a hypothesis.
- Step 2: Prioritize ideas based on scores.
- Step 3: Experiment by executing those ideas.
- Step 4: Conduct analysis and learn from their tests.
Within a decade, Facebook increased its MAU from 50 million to 2 billion, a whopping swell of 4000%.
Exploiting Growth Potential With Growth Engineers
Good growth engineers are driven by the desire to make an impact on the company’s bottom line, and having full stack capability is certainly beneficial for them.
The ability to span across front end and back end middleware, be proficient in extracting and analyzing data, and operate across the layers increases their scope of impact. This gives growth engineers more avenues to test their hypotheses and build more confidence, which increases the likelihood that their product will transform into something that customers will genuinely value. They can then look at correlations in the data and align their growth efforts towards one common goal.
Facebook’s meteoric rise followed a similar path. During Facebook’s expansion, growth engineers adopted the “move fast and break things” philosophy, a developmental mantra coined by Mark Zuckerberg. Instead of planning and debating new features, growth engineers along with product designers would ship features frequently with an aim to bring consistent 1% improvements. By making small but confident improvements, growth engineers created long-term knock-on effects—a compound effect that helped Facebook morph into the world’s most preferred social media platform faster.
Following Facebook’s lead, global unicorns like Uber and Pinterest demonstrated stark results by investing in growth teams and focusing on acquisition, retention, and monetization of more customers. Also called growth levers, these three metrics are strong indicators of sustainability and predictability in the long term, as well as tight focal points for growth engineers.
Maneuvering these growth levers and implementing growth experiments requires growth engineers to have the right data, methodology, and experimental framework in place. It provides them with the needed resources to perform comprehensive analyses of user behavior and iterate products quickly. They can also build reports and testing tools to help other teams like product, sales, marketing, operations, engineering, and analytics understand the customer better.
Challenges Faced By Growth Engineers
Growth engineering teams are nowhere short of data challenges. Just like your product, marketing, and sales teams, they also have lots of data issues to address. These teams are responsible for filling the gaps between different departments and chasing measurable goals for incredible results. Obviously, they deal with a lot of scattered data.
In our interactions with customers, we find that most companies have data quality problems. The frustration of inconsistent integrations or incorrect configurations on the data collection side, or any combination thereof, is a real pain that growth engineers have to face. Incomplete datasets lead to mistaken hypotheses and erroneous experiments, and ultimately wrong product conclusions or actions.
Another big challenge for growth engineers is the ability to cut through data. Even if they acquire the data—how should they use it and understand it? And how do they arrive at the correct conclusions? More data doesn’t necessarily equal better outcomes. It just means that you’ve increased the probability of finding something useful.
This is a challenge that Gustaf Alströmer, a group partner at Y Combinator, summarizes:
“If you worked at Facebook, then you will be, like, dreaming of Daily Active Users (DAU) and Monthly Active Users (MAU), and those kinds of metrics, which actually don’t apply to Airbnb at all. Those are not the kind of metrics we cared about; so if you’re looking at these metrics while evaluating a part of Airbnb, then you weren’t thinking hard enough about the type of product that Airbnb was.”
Central point: Different growth teams have separate accountability metrics that are prioritized based on data. Growth engineers examine the funnel’s most important steps, drop-off rates, lowest percentages, and benchmark each of these steps. They determine the impact they can have on a project, as well as the actual chance they believe they have, such as the likelihood that they will actually hit it—using data.
Every business is different, has a distinct product, and has a varying degree of use. The way you grow a social network like Facebook differs greatly from how you grow a travel business like Airbnb. Unfettered access to quality data is very important because among the sea of intuitions, it is the only way to prove out what is actually driving your growth. Growth engineers enable other teams to avoid the correlation trap and correctly identify the causation.
“The scariest day is when your metrics go down, but the second scariest day is when your metrics go up and you don’t know why.” Growth engineers really have to have the mindset and technology of what is really driving each metric that day and truly understand the behavior behind it. George Lee, head of growth at Instagram
With the abundance of data, it’s highly possible that you can choose data and go wrong in making conclusions. Growth engineers, therefore, need not just the curiosity to ask questions, but the right infrastructure to find answers and try iterative experiments.
The correct data infrastructure enables growth engineers to gain meaningful insights and offers them data-backed confidence when executing A/B testing, copy adjustments, and modifying onboarding methods. Shipping impactful wins are increasingly dependent on how growth knowledge and effect are distributed within the team. Unless growth engineers have the right systems in place to course-correct their actions and help them learn quickly, they will always struggle to create progressive results.
What Makes Modern Data Stack So Crucial for Growth Engineers?
The right data stack serves as the foundation of high-quality experimentation, idea prioritization, and effective decision-making. To deliver value iteratively, growth engineers must have the necessary infrastructure and data integration in place.
Asking the right questions is important, but if your data infrastructure can’t help your growth engineers understand and see the right signals, it’s a waste of their efforts. Vladyslav Hrytsenko, CTO at Mighty Digital
Running growth teams successfully requires a diversity of ideas and hypotheses. Your business really can’t scale if you rely on the product manager to be the source of ideas. Rather, you want that source of ideas to be everyone on that team. Every member should feel like they are a part of the ideation process. And that doesn’t happen by accident. You have to facilitate the maximum generation of high-quality ideas within a team—through equitable access to quality data.
A well-orchestrated modern data stack reduces complexity by providing self-service and agile data management. Growth engineers can work with data by combining various input sources into one clear format and brainstorm pragmatic solutions.
When it comes to benefits, MDS has a lot to offer growth engineers. It helps in:
- Consolidation of Information: Modern data stacks integrate and transform data from discrete sources like Facebook Ads, Google Analytics, Amplitude, and Mixpanel into one common repository like a data warehouse. This helps growth engineers gain a fundamental and common understanding of the product data, user behavior, web traffic conversions, product signups, and growth metrics for which they are accountable.
- Better Prioritization of Ideas: Growth engineering teams constantly have ideas floating around in their heads. How do you choose between those sets of ideas? Which one should you start with first? Most teams use the ICE framework to assign a score to each idea and MDS supports efficient idea prioritization by providing comprehensive and quality data for precise scoring.
- Substantial Time Savings: Growth engineering teams test a variety of ideas and hypotheses to find those that have the greatest impact. Hence, the velocity of testing and iteration becomes very important. By employing MDS, growth teams can reorganize their data funnel and transformation systems in a way to improve data ingestion, processing, and analysis, ultimately reducing time to insight and action.
- Automating Repetitive Work: Modern data stacks using high-performance ETL tools like Hevo support automated data replication from source to target in the case of schema drift. They reduce “grunt work” for growth teams by automating repetitive but necessary tasks. Hevo Pipelines provide a no-code platform with Change Data Capture and Log-based Replication to fast-track source changes. Growth engineers can also make use of self-service analytics in MDS to speed up data analysis and visualization.
- Scaling: An inevitable part of growth is the proliferation of new data sources to capture and analyze diverse data. While your growth teams are busy hyper-fueling growth, MDS takes care of the increased integration needs by providing horizontal scalability in the cloud without incurring expensive or time-consuming downtime.
In all, the tools in the modern data stack empower growth engineers to gain in-depth knowledge about current user behavior and uncover areas for product improvement.
“Growth experimentation improves our ability to make key decisions by refining our intuition with objective customer data.”
Having a reliable data stack to back up their ideas or decisions eliminates subjectivity from experimentation and introduces objectivity. It helps growth engineers correctly map correlation to causation and identify areas where their intuition is playing wrong.
“The Right Approach” to Adopt a Modern Data Stack
Building the right data stack means focusing on each stage of the process and filling it out with tools that match your customer’s needs and long-term business goals. It also helps to identify tools that integrate seamlessly, as this will simplify your workflow tremendously.
If you hear us, our advice for most businesses remains the same—develop a solid understanding of the various tools for use in different situations before you build or modify your existing data stack. If you don’t have time to go through that, Mighty Digital growth experts can help you identify the best available platforms to assure accuracy and consistency across the stack and route data through different stages efficiently.
Chase the “Who, What, and How”
If your growth engineers are creating a modern data stack from the ground up, we recommend that you first understand who your customers are, what their preferences are, and how they use your product. Knowing what to look for and what benefits you want from a modern data stack is far more important than going blindly and arranging various components in disarray.
When you build your modern data stack around questions that you actually need answers to, you create a cost-effective, product- and customer-focused tech stack. Vladyslav Hrytsenko, CTO at Mighty Digital
For a majority of businesses, customers are and will continue to be the primary motivators of business success, so investing in a modern data stack that does not improve the operational teams’ ability to serve customers is a complete waste of money and resources.
A modern data stack is more than just a change of tools; it is a shift in mindset as well. Unless your people desire to change the way they operate, a modern data stack can’t be a success.
Choose the Best of the Breed
MDS is a union of various configurable components: data sources, data ingestion and transformation tools, a data warehouse (or lake), a BI and analytics platform, and a reverse ETL solution in the cloud. Autonomous elements mean greater flexibility; you get to choose the best tools with the best defined core competencies.
As your business teams add new tools, growth engineers can plug them into your data stack quickly, allowing richer insights to be discovered across teams. With the advent of fault-tolerant data pipeline management tools like Hevo, growth engineers can quickly pull in and manage data at any size and scale from different sources (e.g., CRM, billing systems, inbound marketing platforms, etc.), run transformations for analytics, and deliver operational intelligence to business tools without getting bogged down by vendor lock-in.
Using this best-of-breed strategy, growth engineers can create a data stack that works as a synergy of all systems without introducing any inconsistent user operations or mediocre functionalities across the board.
Start Now And Build Incrementally
Paving a path to a successful modern data stack requires patience and deliberate implementation.
For the majority of growth engineering teams, it is crucial to start collecting data as soon as it becomes available.
Asking questions about the customer early on is advantageous because it provides insight. If you begin asking questions too late, you may lose a significant number of opportunities.Vladyslav Hrytsenko, CTO at Mighty Digital
It is only important to remember that as long as you are able to derive actionable value from your tech stack that captures a sufficiently large volume of customer data, you are on the right track to champion your product growth. As long as you are doing so, you’ll always have opportunities to hack your way to sustainable growth.
In sum, a prioritized canon of experiments and confidence for interpreting user experience can only come from an organized data infrastructure.
MDS enables the creative conjuration of new ideas from growth engineers by providing quality and comprehensive data for strategic and objective decisions. If yours is a growth team trying to create a flywheel of satisfied customers and increased revenue along the way, a modern data stack is the McLaren you should be riding on.
Get in touch with our growth experts right away if you’re curious about how we can help your business adopt cutting-edge, industry-leading data practices. We at Mighty Digital are committed to help you make the most out of your data so that you can always make smarter, more informed choices.