Measuring the impact of Nijsanji EN speaking other languages via Super Chats, pt. 1

VTuber Project NIJISANJI EN of ANYCOLOR Inc. is notable for “using mostly English” for streaming activities.

As that implies, members of Nijsaniji EN are known to use multiple languages. The first three waves of talents under management appear to hold near-native if not native Japanese, Tagalog, French, Spanish, and Russian language skills. Some of them know a bit of Hmong and Cantonese too.

As such, it would not be unreasonable to intuit that Nijsanji EN’s members may have a larger addressable market size as multilingual streamers than if they were English-only speakers.

Let’s find out if that’s true!

Using tuber.report’s super chat and live comment dataset, we can answer two research questions:

1. If a Nijsanij EN talent speaks a non-English language, is the monetary value or volume of super chats from countries associated with that language significantly higher?
2. Are any of the super chat senders from those countries unique to that talent? In other words, can we correlate the language spoken to a unique income source — a source of super chat?

All data mentioned below was pulled on 2021/11/03. And due to article length, we’ll only cover question #1.

We’ve mapped livers to languages spoken and mapped that to the Super Chat-eligible countries as of October 2021. Note that many regions associated with Hmong and Russian are unable to send super chats.

Talent -> Language -> Country/Currency mapping

In tuber.report’s BigQuery dataset, that means running this SQL query. The query averages out counts of super chats by video to roughly control for channel lifespan.

the actual BigQuery query used below

Immediately throwing the result straight into Google Data Studio to visualize, we find that a comparison between the channels queried already gives us some fun insights.

Count of Super Chats by the top 20 ISO 4217 currency codes by total count

The visual outliers here already answer the first research question:

• Nina gets just about as many super chats in Russian Rubles (RUB) as US Dollars (USD).
• Millie has a very clear outlier in the amount of super chats sent in Philippine Pesos (PHP).
• Reimu’s super chat currencies are spread evenly, with the USD taking a very small majority over other countries. Of particular note is that most of these countries are in South America. (and the Norwegian Krone too, I guess?)
• Petra gets more super chats in Japanese Yen (JPY) than USD.
• There are no visual outliers for Rosemi and Selen, so we will ignore them for now.

Let’s focus in on Reimu first — because it only includes top 20 currencies by count, this 100% bar chart doesn’t accurately capture the diversity of her super chat population.

Top 20 of Reimu’s super chats by currency as of 2021/11/03

This pie chart contains the top 20 currencies by volume for her channel. Cross-referencing the Spanish countries table above, at least 41.2% of Reimu’s super chats were from Spanish-speaking countries.

Top 20 of Nina’s super chats by currency as of 2021/11/03

Nina’s chart requires less math. Nina’s top two super chat currencies — RUB and USD — have approximately equal contributions at ~24% each.

For Millie, we already know that PHP composes the second-largest slice of super chats — so let’s do something a little different. Let’s compare her super chat count in PHP to every other VTuber that has gotten a super chat in a Philippine Peso.

Average % composition of super chats in PHP, per video, per VTuber — for VTubers that average at least 100 Super Chats per video

At ~25%, Millie appears to have the second-highest average % composition of super chats in PHP sent to VTubers that average ≥100 total super chats per video. Interestingly, her % composition is only 5% larger than fellow Nijisanji talent Joe Rikiichi, who holds 3rd place.

Somewhat hilariously, the Nijisanji EN Official channel holds 4th place at 13%.

Data appears to validate our intuitive assumption around uniqueness. Research question #1 is half answered: If a Nijsanij EN talent speaks a non-English language, the volume of super chats from countries associated with that language is significantly higher.

The next post in this series will cover the second half of that question: monetary value. Afterwards we’ll address the research question around uniqueness, which might help us correlate a talent to the growth or exclusivity of a specific market segment.

tuber.report is a project that aims to answer engagement-related questions, mostly as they pertain to Virtual YouTubers — VTubers. For more info and for questions about commercial/academic access to the data used in this article, please see https://tuber.report/about