Nazar Babenko 🇺🇦 retweetledi
Four years since russia’s full-scale invasion of 🇺🇦Ukraine.
Four years of resistance, resilience, and courage.
English
Nazar Babenko 🇺🇦
229 posts
Nazar Babenko 🇺🇦
@wellenborg
Product Manager @StreamsCharts. Helping brands, agencies & creators to get detailed streaming data. Formerly @esportalSE 🔫 Email: [email protected]
Kiev, Ukraine Katılım Ağustos 2014
1.6K Takip Edilen366 Takipçiler
Introducing the first Full-Stack Vibe Coding Platform powered by Claude Code.
Not only can you build a professional mobile app that accepts payments and ship it to the app store... As of today you can build a web app and deploy it to the internet with one click.
In celebration of 5000 vibe coded apps published to the app store, we're giving away a free month of vibe coding for those who reply to this post.
Reply and we'll DM you credits 👇
English
Nazar Babenko 🇺🇦 retweetledi
Have you gotten your 2025 recap, yet? 👀
YouTube: charts.team/nkzv
Twitch: charts.team/w26i
Kick: charts.team/7j83
Rumble: charts.team/g0as
If you do share, don't forget to tag us! 🥰
English
@ostonox 2/2 We see how the wording led to a misunderstanding, so we’ve updated it to make things clearer.
Appreciate you flagging this.
English
@ostonox 1/2 Hey — thanks for pointing this out.
Stake is partnering with Streams Charts on our 2025 YT summary pages, and this does not mean they sponsor the streamers who download the summary. As for Kick, Kick chose to sponsor their own summary, which is why their logo appears there.
English
Finally live!
Streams Charts 🇺🇦@StreamsCharts
🚀 Our whitepaper, The Hidden Cost of Viewbotting, is OUT NOW! What’s inside: 📊Latest data on fake viewership ✍️Actionable insights for marketers 🤬Proven detection methods and business risks 🟣Exclusive statements from @Twitch 📩Download your copy: streamscharts.com/reports/viewbo…
English
Nazar Babenko 🇺🇦 retweetledi
🎤 Managing top creators? Our API gives talent agencies the data edge:
📊 Compare streamer performance
🌎 Track audience by region
💸 Monitor growth, earnings & engagement
📈 Spot rising stars before anyone else
Turn insights into smarter deals.
👉 streamscharts.com/api?utm_campai…
English
Nazar Babenko 🇺🇦 retweetledi
🚨 @Forbes covered Twitch’s crackdown on viewbotting and quoted our PM, @wellenborg
📊 Our upcoming whitepaper with @Audiencly dives deeper into the hidden costs of viewbotting. Stay tuned.
Read their piece 👇
forbes.com/sites/maryroel…
English
I will givecredit where credit is due. @wellenborg shows a much more robust analysis in this post with some interesting results.
Specifically: According to the early data there WAS a significant impact on viewer numbers compared to earlier months.
That being said: The 25% or 47% other outlets are stating are not likely to be correct, erring closer to the 10% mark. We will need more data before we can judge this.
Nazar Babenko 🇺🇦@wellenborg
Thank you for your opinion, Barry, I really appreciate that from someone who understands statistics. It was simply easier to provide specific examples with week-to-week comparison for public interpretation, but I agree that this does not justify my team. And fully agree that week-to-week comparisons alone can be misleading if taken out of broader context — that’s why, behind the scenes, we didn’t limit our analysis to just two weeks. What we actually did in the lab: 1) Looked at the entire year of 2025, comparing week over week trends, not just Aug 14–28. 2) Controlled for natural fluctuations like collabs, events, and seasonal shifts by comparing against monthly minimum baselines. 3) Found that Mon–Thu (Aug 18–21) tracked normally and aligned with prior weeks. 4) But starting Aug 22, trends diverged: the following weekend dropped below the hourly minimum of every month in 2025 (see chart attached). 5) No other week this year registered declines this deep — making the post-Enforcement weekend a clear statistical outlier. So this isn’t about “abusing charts” — it’s about showing that even when measured against a year-long baseline, the last weekend stands out as historically weak. 📉 To me, that strongly suggests the drop wasn’t random noise from a few collabs or special events, but a structural shift specific to the crackdown period.
English
Nazar Babenko 🇺🇦 retweetledi
🚀 With our data, @MovistareSports ranked the fastest growing Hispanic streamers of H1!
All top 10 are on @kick with @mernuel_ leading the pack.
Don't miss the full breakdown with some insights from our PM, @wellenborg 👇
movistaresports.com/los-streamers-…
English
@InvestigatorD7 I’ve checked it, nothing change, still counts
English
Thank you for your opinion, Barry, I really appreciate that from someone who understands statistics.
It was simply easier to provide specific examples with week-to-week comparison for public interpretation, but I agree that this does not justify my team.
And fully agree that week-to-week comparisons alone can be misleading if taken out of broader context — that’s why, behind the scenes, we didn’t limit our analysis to just two weeks.
What we actually did in the lab:
1) Looked at the entire year of 2025, comparing week over week trends, not just Aug 14–28.
2) Controlled for natural fluctuations like collabs, events, and seasonal shifts by comparing against monthly minimum baselines.
3) Found that Mon–Thu (Aug 18–21) tracked normally and aligned with prior weeks.
4) But starting Aug 22, trends diverged: the following weekend dropped below the hourly minimum of every month in 2025 (see chart attached).
5) No other week this year registered declines this deep — making the post-Enforcement weekend a clear statistical outlier.
So this isn’t about “abusing charts” — it’s about showing that even when measured against a year-long baseline, the last weekend stands out as historically weak.
📉 To me, that strongly suggests the drop wasn’t random noise from a few collabs or special events, but a structural shift specific to the crackdown period.
English
While I really appreciate what @StreamsCharts does, the comparison of two weeks is inherently flawed and statistically invalid.
The week to week analysis does not account for natural fluctuations such as special events or a few large streamers collabing.
Abusing the power of Maths and charts like these to paint fallacies as fact has to be something that you, as a Product Manager, have to be against as it damages the reputation of the platform.
English
Reposting our team’s analysis on the recent Twitch viewbotting crackdown 👇
What I find most interesting here is the balance between data-driven insights from third-party analytics and Twitch’s official perspective.
Twitch’s statement is clear (per @MikeMinton for @Dexerto in @DexertoVirginia story):
“Our viewership is not in free fall... We’ve seen some misinformation swirling, and a lot of that misinformation includes data pulled from third party sources. Those numbers are incorrect and are not from Twitch.”
A few thoughts worth highlighting:
✅ Third-party data (like Streams Charts) comes directly from the Twitch API, which makes it a valid lens — but interpretation always matters.
↔️ The reality probably sits between both perspectives: yes, the crackdown exposed how widespread viewbotting had become, but no, Twitch as a whole isn’t “collapsing.”
🔍 For the industry, this raises a bigger question: How do we reconcile official platform statements with independent analytics?
Personally, I think these contrasting narratives are healthy. They push the ecosystem toward more transparency, smarter discussions, and better trust signals for creators, advertisers, and audiences alike.
💡 Curious to hear from others: Do you trust third-party analytics, or do you rely more on official platform reporting?
Streams Charts 🇺🇦@StreamsCharts
📊 After Twitch’s major viewbotting crackdown, real audience numbers shifted fast. Typical hourly drops ranged from 17–21%, but at peak the decline hit -47% compared to the week before. Full breakdown of the ripple effect ➡️ streamscharts.com/news/how-twitc…
English
@InvestigatorD7 Where you saw this information? Any proofs?
English
Spending my Saturday volunteering with the Animal Wellness Foundation helping cats find their forever homes 😸
Each cat is fully spayed / neutered, vaccinated and microchipped!
English
That’s a fair point — fraud and methodology are universal challenges across digital platforms, not just Twitch. But I have to admit, I find the question of “who is validating how many times a user is counted?” a bit surprising. Ultimately, shouldn’t this be exactly what the platform itself controls and transparently communicates?
To give an example from our own side, I’ll share authorized viewers count methodology (those who watched broadcast as a logged-in user):
- We count a viewer as unique if they’ve spent at least one minute on a broadcast.
- That viewer is then tracked consistently across the entire session — we don’t duplicate them during the same stream, instead we record total viewing time.
This avoids inflation and provides a clear measure of both unique reach and depth of engagement.
If third-party analytics can implement and validate transparent counting rules like that, it feels reasonable to expect the same or better from a major platform. After all, creators, advertisers, and brands are making real business decisions based on these numbers.
At the end of the day, the issue isn’t whether concurrent counts are “perfect” — it’s about platforms establishing trust in methodology and giving the ecosystem confidence that the numbers being used to make business-critical decisions are consistent, accurate, and fraud-resistant.
English
@wellenborg @Dexerto In addition to the bot problem who is validating the methodology of how many times a user is counted? How devices are deduped? Etc etc. Not a great source and incentivizes fraud.
English
I completely agree that engagement is the king of metrics, especially on Twitch where live chat activity, emotes, and community interactions define the value creators bring. Bots can’t fake real conversation or authentic communities.
That said, I’d argue viewership is the other side of the medal (and view counts directly affecting overall viewership). While inflated view counts don’t reflect true engagement, they do shape perceptions of platform popularity and momentum — both for creators attracting sponsorships and for advertisers deciding where to invest.
My take:
- Engagement metrics (chat, follows, subs) = quality of the audience
- Viewership metrics = scale and visibility of the platform
The crackdown is helping us separate the two more clearly. Long term, this could actually increase trust in Twitch’s numbers and make both creators and brands more confident in what those stats really mean.
English
@wellenborg @Dexerto View counts are not viewership. The bots that manipulate view counts don’t impact our core engagement and audience metrics. While removing them results in lower view counts it doesn’t mean our actual audience has decreased or is engaging less.
English
Our upcoming whitepaper dives deep into just how advanced viewbotting has become.
In Q2 2025, we found that over 10% of Twitch channels with 50+ average viewers showed persistent signs of viewbotting, and on Kick, nearly one in six streamers in that bracket were confirmed viewbotters.
What’s even more concerning is that many are now using analytics tools to actively refine their tactics, making detection a moving target.
The cost isn’t just inflated numbers—brands are losing millions, and genuine creators are being pushed out as trust in the ecosystem erodes. Looking forward to sharing more data and solutions when the full report drops in September.
#BeatTheBots #viewbotting #livestreaming #twitch #kick #streamscharts
Streams Charts 🇺🇦@StreamsCharts
As @wellenborg said: 🗣 “One in ten streamers who use broadcast analytics tools engage in viewbotting, attempting to manipulate their stats.”
English
I was botted again this stream, this time for 5.5-6k viewers
Raid said 3.7k which means I most likely had 4.2-4.7
The problem is viewbots are starting to get harder to detect and my graph seems normal
Twitch is not fixing this issue that they promised to work on
This will create a lot of fake natural engagement streamers
It’s going to become a bigger requirement to viewbot to succeed on this platform
This is a MASSIVE twitch L
English
I wanted to share some early findings from our upcoming Streams Charts x Audiencly whitepaper (releasing mid-September): in Q2 2025 alone, over 41,000 Twitch channels with 50+ average viewers had at least one suspicious stream, and more than 10% showed clear, persistent signs of viewbotting.
On Kick, the trend is just as alarming—starting in 2025, we began to consistently record over 1,000 blatant viewbotters each month, and in Q2 2025 that number tripled to nearly 3,000. This means that about one in six Kick streamers averaging 50+ viewers are now confirmed viewbotters (there were 18k+ streamers with 50+ average viewers on Kick in Q2), a record high and a sign of rapidly growing fraudulent activity on the platform.
The impact is massive: brands are losing millions to fake audiences, and it’s the small and mid-sized creators who suffer most as companies pull out of the space. Viewbotting doesn’t just hurt sponsors; it undermines trust across the entire ecosystem, making it harder for genuine streamers to get noticed and for the industry to grow sustainably.
I’ll be sharing more details and data in the full whitepaper soon, but wanted to highlight these numbers here because this is a problem we can’t afford to ignore.
xQc@xQc
Twitch has cracked down on bots in the 2-3 days and viewbotters/victims of viewbotting have been exposed. Streamers that are part of groups/orgs are seemingly being botted much more heavily. I don’t want to start witch hunts but the data is interesting. Go see for yourself
English