Digital technology has become an integral part of many aspects of modern human life, and education is no exception. Given the challenges the world has faced in recent years, it has become a key tool in ensuring the accessibility and quality of education. It is no surprise that educational technology (EdTech) is gaining increasing popularity. The latest report from Straits Research shows that the global EdTech market size is expected to grow from USD 169.37 billion in 2024 to USD 598.82 billion in 2032, at a CAGR of 17.10%. Against the backdrop of such growth, competition among EdTech companies is becoming increasingly fierce. Data analytics is one of the key factors that companies need to consider to stay ahead of their competitors. We spoke to Svetlana Repina, Senior Data Analyst at leading Russian EdTech company Skyeng and IEEE Senior Member, about the analytical tools used by high-performing EdTech companies.
Svetlana, thanks for taking the time! Data and personalized learning approaches are becoming essential to the EdTech industry. In your view, what is driving this trend?
In recent years, educational technology has expanded rapidly, especially in the online learning sector. This growth is largely due to a demand for more accessible and personalized learning experiences. Educational platforms leverage data to better understand user needs and deliver solutions that align closely with their goals. For instance, at Skyeng, I am leading a large user segmentation project for Skysmart brand to craft targeted offers and optimize user acquisition funnels, ultimately boosting sales and reducing user acquisition costs. By collecting data through quizzes during the application process, we identify behavioral segments and tailor unique offers to increase conversion to payment.
What valuable insights have emerged for you from this project so far?
The quizzes have given us a clearer view of what drives our users. For example, we found that users new to online education are less likely to convert to paying customers. Additionally, we saw that a strong sense of urgency and a readiness to begin classes immediately are key factors influencing user decisions. Using this information, we developed behavioral segments and are testing tailored offers for each group to identify the most effective strategy for driving conversions.
You have optimized the A/B testing process at Skyeng, resulting in a nearly 50% reduction in the time needed to design, summarize, and dashboard an experiment. Can you tell us more about that?
Absolutely! Before standardizing our analytics approach, our teams spent a lot of time preparing A/B tests and generating reports. I streamlined the process by creating templates and unified dashboards, which cut preparation time by almost 50%. Now, analysts can move more quickly from data analysis to strategic insights. Standardization has also made our results more transparent and comparable across teams, enhancing both our analysis’s productivity and quality. This shift has brought significant value to the business.
Apart from leading data analysis at Skyeng, you are also actively sharing your knowledge through academic publications. Could you tell us more about your recent studies?
Certainly! Two of my most interesting papers focus on topics I find personally fascinating. In my article “Evidence-based Practice in Data Analytics,” I explore the importance of transparency, reproducibility, and objectivity in analytics. It’s somewhat analogous to legal principles that help establish trust in data and make conclusions more reliable. In “Duality of Data,” I discuss how cognitive biases and emotions can influence data interpretation. The same data set can lead to different conclusions depending on the presentation. These topics are particularly fascinating because they highlight the need for responsible analytics.
I’ve also published two other articles based on our segmentation project at Skyeng. The project is still ongoing, and I’m looking forward to sharing more insights as we complete the testing phase.
The depth and scope of your work are impressive! Recently, you judged a Humanize AI Text Hackathon organized by Hackathon Raptors, an international association of IT professionals – were there any interesting, innovative developments applicable to the EdTech world?
It was an inspiring experience! The hackathon was focused on how technology can make digital interactions more human-centered. As a judge, I evaluated projects for both their technical merit and their impact on users and scalability. It was exciting to see participants integrate empathy into tech solutions. This experience reinforced my belief that understanding human behavior is crucial for deriving meaningful insights in data analytics.
Although the hackathon wasn’t explicitly aimed at EdTech, some solutions had clear potential for educational technology. For instance, many teams developed approaches to make AI-generated text sound more natural and relatable, which could be incredibly valuable in creating more engaging, conversational learning content. Additionally, a few projects focused on emotional intelligence in AI, equipping systems to respond to user needs with more empathy and adaptability. Bringing these elements into EdTech could lead to more personalized, student-centered learning experiences. Adopting such innovations could help create tools that don’t just deliver content but also support and motivate learners throughout their journey.
To conclude, what do you think about the development of the EdTech market? In your opinion, what are the most prominent trends?
Personalization will remain a central trend. Platforms will increasingly use data analytics to refine learning programs and produce adaptive content that aligns with each student’s knowledge level and interests. For example, user segmentation, like in our project, allows us to build strategies that attract new users and improve retention by offering a tailored and meaningful experience.
