After February 24, 2022, we witnessed a mass shift to the Ukrainian language — on social media, YouTube, and in everyday communication. It wasn’t just a change in linguistic habits; it was part of a broader societal transformation. But later, a new question emerged: are we still moving forward with Ukrainization — or are we stuck in an information bubble, seeing only what we want to see?

At newage., we work with data every day — building analytics-driven campaigns and making decisions based on numbers. So when it came to understanding the state of language use in Ukraine, we approached it the way we know best: using open data and machine learning tools.

We asked ourselves a series of questions:

• Where are we now as a society in terms of language use?
• What language do Ukrainians use to comment, communicate, and publish content online?
• And how can we measure this — not based on assumptions, but on real numbers?

The problem with traditional approaches to measuring language dynamics

The language situation in Ukraine has long been the subject of sociological studies, media debates, and political discussions. However, despite active interest, there is still a lack of objective, large-scale, and regularly updated statistics on linguistic shifts in the country. Sociological surveys remain the primary tool for gauging public sentiment, but they come with limitations: small sample sizes, the influence of question phrasing, and political bias can distort both the results and their interpretation.

In recent years, analyzing search queries has become a common method for identifying social shifts — a step toward a more accurate, data-driven approach. However, such research does not provide insight into user motivations and requires robust infrastructure to be systematized. To gain a more comprehensive picture, it makes sense to examine user behavior in the open digital space (though even search engine data can already serve as a valuable indicator).

Research Methodology

Instead of relying solely on traditional methods, we chose to use what the newage. team knows best — working with data. In the era of open data, it’s possible to measure the language environment more accurately, frequently, and at scale. The key is knowing what exactly to analyze and how to work with it. That’s why we developed our own approach — combining open information sources with the capabilities of machine learning.

At the core of the research lies a simple idea: to use data already available in open sources and process it through modern machine learning tools. The language environment is shaped not only by kitchen-table conversations, but also by YouTube comments, Telegram discussions, and classified ads on websites.

Important disclaimer: the public space — even if it’s just comments under a YouTube video or a Telegram post — doesn’t always fully reflect private life. However, this is where shifts begin. That’s why our task is to capture, analyze, and make these shifts visible.

Data sources we analyzed

To conduct this analysis, we selected several platforms that best reflect the linguistic landscape of the public digital space:

• YouTube comments — one of the most powerful information channels. Comments under videos provide insight into audience reactions and the predominant language of communication.
• Telegram comments — an extremely popular platform, especially in the news and political segments. We analyzed both post comments and public discussions.
• A major e-commerce platform — an example of a service with highly everyday, user-driven content. The language of listings often reflects real-life speech habits.

How the System Works

To process such large volumes of data, we built an automated system consisting of several components:

• Scripts, parsing, aggregation — the system is built on three main scripts. One orchestrates the entire process, while the others are responsible for collecting and processing data from specific platforms.
• Language identification via LangDetect — the collected texts are processed through the LangDetect machine learning model, which detects the language of each comment. We label them as ua, ru, or other (e.g., English or emoji). The model may not always accurately identify the language of short phrases or symbols — but this is taken into account in our analysis.
• Visualization via Google Sheets — a user-friendly interface allows for platform, channel, and time range selection, instantly displaying the number of comments, language distribution in percentages, and dynamics over time.

This approach allowed us to build a flexible tool that can be scaled, updated, and adapted to new data sources.

What we saw in the data

After processing over 300,000 comments and messages across various platforms, we gained a clear picture of the language landscape in the online space. These data not only confirmed our intuitive impressions but also allowed us to observe trends in dynamics. It turns out that the digital environment can very accurately reflect shifts in the linguistic behavior of society.

Percentages and Dynamics Across Platforms

The Ukrainian language confidently holds a leading position in online communication.

On most major platforms, it has become the primary language of interaction.

YouTube

• Among the most popular channels, “Levy na dzhypi” stands out with 81% of comments in Ukrainian. On Fedoriv Vlog — 89%, and Pidpilnyi Standup — 87%. However, on channels with a predominantly Russian-speaking audience, such as “Kvartal 95”, Ukrainian has not yet become dominant — only about 49% of comments are in Ukrainian. This highlights the need to strengthen Ukrainian-language presence and support in certain segments.

These channels are just examples — in total, we analyzed data from over 200 YouTube channels.

Telegram

• Telegram shows a similarly clear trend. The channel of Serhii Sternenko has 83% of comments in Ukrainian. DeepState has 73%, slightly lower due to waves of bots from Russia. Forbes Ukraine holds steady at 83%, and AIN.UA leads with a record-breaking 92%. Even on channels with a diverse, multilingual audience, Ukrainian continues to maintain a confident presence.

E-commerce classifieds platforms reflect real-life language habits

• Everyday language is another key indicator of change. In categories such as Jobs, Transport, and Children’s Goods, Ukrainian dominates consistently, with 80–81% of listings. In Electronics, Rentals, and Services, the share is slightly lower at 60–70%, but steadily growing. Even in informal segments like Free items and Barter, over half of the ads are written in Ukrainian.

These results reflect more than just a shift in language — they represent a conscious choice by millions of Ukrainians to be heard in their native tongue.

Ukrainian Language Across Regions: Moving Toward Unity

Data from Telegram channels across various Ukrainian cities clearly show: the shift toward the Ukrainian language is happening not only in the central and western regions but across the entire country — albeit at different levels of intensity.

In western Ukrainian cities, the trend is especially strong:

• Lviv — 93% of comments are in Ukrainian
• Ivano-Frankivsk — 92%
• Ternopil — 82%
• Rivne, Zhytomyr, Vinnytsia — 87–88%

The central regions of Ukraine also demonstrate high levels of Ukrainian-language use:

• Kyiv — 86%
• Bila Tserkva — 81%
• Poltava — 78%
• Khmelnytskyi — 77%

In Ukraine’s eastern and southern regions, the share of Ukrainian-language communication is gradually increasing — but Ukrainian still does not dominate:

• Kharkiv — 53%
• Mykolaiv — 65%
• Dnipro — 61–62%
• Odesa — 54%
• Occupied territories — predominantly 5–10% of comments in Ukrainian

Important! In many cases, these numbers have already doubled or even tripled since the beginning of the full-scale invasion. This may not be the most optimistic picture, but it continues to improve every day.

Is There a Language Rollback?

This is a question we’ve heard often. Social media users regularly raise concerns that “Ukrainization has slowed down” or that “there’s a reversal.” And indeed, after the wave of Ukrainization in 2022, some may have felt that the pace has decreased. dWe decided not to rely on assumptions — but to turn to the data. And the data gave a clear answer.

Our analysis covered dozens of channels across various platforms — from Telegram and YouTube to major e-commerce sites. On most of them, the Ukrainian language maintains a leading position, and on many, it continues to grow. While growth rates may vary depending on the platform, topic, or region, the trend is clear: Ukrainian is gaining strength.

Even on platforms where bot activity or a foreign audience might have an impact (such as Telegram), Ukrainian still holds the lead. In segments with lower vulnerability — like YouTube or platforms for everyday listings — Ukrainian has already become the norm.

Another important indicator is regional change. In cities where Russian traditionally dominated, we recorded a significant rise in Ukrainian-language content. For example, in Odesa Telegram channels, the share rose from 2–11% to 50%. In Kharkiv, it’s now over 45%, and in Dnipro, consistently above 60%.

This isn’t just a temporary wave. It’s a stable trend. Yes, the transition is uneven, but it is ongoing.

Key Insights from the Research

After processing data from various platforms and channels, we identified several important insights that help better understand the real status of the Ukrainian language in the digital space.

Ukrainian is the new norm

On most of the analyzed platforms, the share of Ukrainian-language comments ranges between 65–85%. The full-scale invasion became a turning point — millions of Ukrainians consciously switched to Ukrainian both in public and private communication.

Bubbles Shape the Language Environment

Ukrainian-speaking bloggers, media, and content creators are building strong communities where Ukrainian is the default language. This reinforces the feeling that “everyone here speaks Ukrainian.” While it may not reflect the full picture across the country, it plays a crucial role in normalizing the language in digital spaces.

The content we consume daily shapes our language habits. Ukrainian-speaking creators define the tone of the online environment — and that’s why it’s essential to support their visibility and growth.

Political Channels Targeted by Bots

On news and politically significant channels, we recorded waves of bot activity — mostly Russian-speaking. Such attacks can distort language statistics, but at the same time, they serve as a clear indicator of the vulnerability of the information space. Ensuring digital hygiene in this area is critically important.

Regional shifts run deeper than they seem

It’s not just cities like Kyiv or Lviv that are showing positive dynamics. According to research by Detector Media, the share of people in the Odesa region who speak Ukrainian at home increased from 26% in 2021 to 42% in 2023. Similar shifts are taking place in Kharkiv, Dnipro, and Kryvyi Rih. Yes, the road to widespread Ukrainian language adoption is still ongoing — but it’s now unstoppable.

These insights aren’t just conclusions of the research — they’re markers of the future. The Ukrainian language is alive, evolving, and gaining strength. And that’s not an abstract idea — it’s a measurable fact.

Language trends in numbers reflect deep societal transformation: if before 2022 Ukrainian-language comments were in the minority, now we observe a clear shift toward Ukrainian. The data shows that language change doesn’t come from mandates — it stems from people’s daily choices. And this shift, although uneven, is steadily cementing Ukrainian as the new norm, even on mass platforms like Telegram.

This is not just about communication — it’s about choice, responsibility, and national identity. We can now confidently say: we speak, write, and think in Ukrainian — and behind these words stand not only convictions but also concrete data that confirms this shift.