vagor.one — Just a Curious Mind

This post was written with direction from me and drafted with assistance from Claude. The data, the links, and the conclusion are ones I agree with.

Someone asked me — actually I asked myself, which is how these things tend to go — whether vagor.one should be translated into other languages. French. Japanese. Italian. Languages I have some connection to and markets I find interesting.

The honest answer is no. At least not now. Here is why, with the numbers to back it up.

The assumption most people make

The common assumption is that English is the internet's universal language and therefore anyone worth reaching already reads it. This is partially true and partially wrong in ways that matter.

English accounts for around 49.6% of all website content among the top ten million sites on the internet. That is dominant. Nothing else comes close — Spanish and German are tied at around 6% each.

But here is the number that complicates the assumption: approximately 75% of internet users are non-English speakers. Half the content, a quarter of the users. The gap is real.

And yet — 1.53 billion people speak English as a first or second language as of 2026, with non-native speakers outnumbering native speakers by nearly three to one. English is not just the language of English-speaking countries. It is the language of anyone who went to school anywhere in the world for the last thirty years.

What the translation data actually shows

The strongest argument for translation comes from ecommerce. 65% of people prefer content in their own language, while 40% categorically state they would not buy products in foreign languages. And over 60% of consumers in key markets like France and Japan shop only from stores in their native language.

Those are real numbers and they should matter to anyone selling something.

vagor.one is not selling anything. Yet.

The distinction is critical. Buying behaviour and reading behaviour are not the same thing. People who will not purchase in a foreign language will absolutely read in one — especially if the content is genuinely interesting and their English is sufficient. The friction of reading is considerably lower than the friction of entering payment details.

The EF English Proficiency Index 2025 ranked 123 countries — and European countries dominated the top tier, with the Netherlands holding the top spot for the sixth consecutive year. Singapore ranked third globally. These are not markets that need translation to access English content. They are markets where English is a working language.

The content-specific argument

Here is where I land on this personally.

vagor.one is not a generic information site. It is a specific voice, a specific perspective, a specific way of looking at things. The dry humour. The sarcasm that lands without explaining itself. The "funny how that works." These are not things that translate cleanly. Not because the ideas are untranslatable — they are not — but because the register requires a reader who is comfortable enough in English to feel the gap between what is said and what is meant.

A technically accurate French translation of this site would lose most of what makes it worth reading. A French adaptation — genuinely rewritten in French with the same voice at the same register — would require someone who both understands the English original deeply and writes French with that kind of precision. That person is not me, and AI translation does not produce that result yet.

The people most likely to find value in vagor.one are curious, technically engaged, internationally aware readers who encounter English content daily and move through it comfortably. That describes a large audience. It does not require translation to reach them.

The practical argument

The digital English language learning market reached $15.98 billion in 2026 and is projected to hit $31.62 billion by 2031. The world is not moving away from English. It is moving toward it. The investment in building multilingual content on a personal site at launch is a significant ongoing commitment — every new post in English needs a translated version, the voice needs to be adapted not just converted, and the maintenance never ends.

The technical work is manageable — Next.js has built-in internationalisation support and Claude Code can set it up in one session. The content work is the problem. Sixteen posts times three languages is forty-eight additional pieces of content before launch. And that number grows with every new post forever.

The return on that investment for a personal site with a specific voice writing about power tools, B-movies, AI music, and WTF moments is unclear at best.

When the answer might change

Six months after launch, real analytics will show where the audience actually is. If a significant portion of traffic comes from non-English-speaking countries — France, Japan, Italy, Germany — that is a data point worth acting on. Not with full translation, but perhaps with a single language experiment. One language. One market. See if the voice translates and whether the audience grows.

The decision to translate should come from evidence, not assumption. Build the English site properly. Measure what happens. Then decide.

The honest summary

The data supports English-only for vagor.one at launch. Not because other languages do not matter — they do — but because:

The content is voice-dependent in ways that do not survive machine translation
The likely audience is English-comfortable by default
The content work of maintaining multiple languages is significant and ongoing
The ecommerce translation argument does not apply to a content-first personal site
Real audience data does not exist yet to make a language decision from

Come back and ask me again after six months of live traffic. The answer might be different then.

Sources used in this post: