About KTRENZ
KTRENZ is a K-culture data platform. We track how far each release travels across AI answer engines, which artists are rising by 7-day velocity, and how a single event unfolds across the days it keeps moving — then publish it openly, in English, for readers and for machines.
The measured layer is free to read and free to query. Reporting sits on top of it, not the other way round.
Measured, not estimated
Every number is something we counted — server-side crawler requests, reported figures with the source named. No modelled reach, no vanity metrics.
Open to machines
JSON-LD on every page, a plain-text mirror per record, a free REST API and an MCP server. Answer engines read the same facts we show readers.
Velocity over size
Rankings score the rate of change across external signals, not standing popularity — so a rising act can outrank a bigger one.
K-Track
Per release and per artist: which AI answer engines fetched the page and when, which outlets picked it up, which social posts carried it. Counted from our own server logs — requests we served, not impressions or estimates.
Trend
Artists ranked by 7-day rate of change across Korea search, YouTube upload traction, media pickup and fan-challenge volume. Signals are shrunk so one article cannot top the chart, and our own output is deliberately excluded.
Storylines
A comeback rollout, a chart run, a tour leg — each installment records the stage it reached and the figure that moved, with the source named. Not the same moment restated; the thread of what actually happened.
B.Now
Live artist boards built from moments fans caught up on in the last 24 hours — schedules, sightings, official posts — with the English title of the source alongside it.
API · MCP
A free REST API (/api/kdata) and an MCP server (/api/mcp), no authentication: artist search, artist reach, release reach, velocity trend. Every response carries attribution and a canonical URL.
Reporting
Comebacks, charts, awards and K-culture, written for a global audience rather than translated after the fact — labelled Original, Exclusive, or Newsdesk with the primary source linked.
We log every answer-engine request to our pages per path, per bot, per day, and separate answer-time fetches (OAI-SearchBot, ChatGPT-User, Claude-User) from bulk training crawls. Nobody else publishes this for K-culture, because it can only be measured from the inside.
Each figure carries how it was counted and what it excludes — blocked crawlers are dropped, caps are disclosed, and readings that are not comparable are shown side by side rather than silently merged.
JSON-LD on every record, a chrome-free plain-text mirror at /{surface}/{slug}/llms.txt, a full corpus dump at /llms-full.txt, and an MCP server — so an assistant can cite a specific release or storyline instead of paraphrasing a page.
Exclusive features built with the agencies — access fans cannot get elsewhere, while editorial control stays with us and partners never review or approve copy.
KTRENZ is an editor-led publication. Our team curates and shapes what you read, and every piece carries a clear label so you always know what you’re getting.
Original
Independent reporting and features from our writers and contributors — researched, sourced, and edited in-house.
Exclusive
Features produced under our agency-collaboration program. Partners provide access — interviews, behind-the-scenes — but never review or approve our copy. Every Exclusive names the partner.
Newsdesk
Curated coverage of the day’s K-pop news with transparent attribution — the primary source is always named and linked back.
KTRENZ is a K-culture data platform. It measures and publishes first-party data about the Korean wave — AI-citation reach per release, 7-day artist velocity, and event storylines tracked across days — alongside English-first reporting for a global audience.
Four measured surfaces. K-Track maps how far a release travels: which AI answer engines fetched the page and when, which outlets picked it up, and which social posts carried it. The trend ranking scores artists by 7-day velocity — the rate of change in Korea search, YouTube upload traction, third-party media pickup and fan-challenge volume — not by standing popularity. Storylines follow one event across the days it keeps moving, recording each stage and figure with the source named. B.Now boards collect what fans pulled about an artist in the last 24 hours. All four are readable by machines: JSON-LD on every page, a plain-text mirror at /{surface}/{slug}/llms.txt, a free REST API at /api/kdata, and an MCP server at /api/mcp.
From our own server logs. Every request an answer-engine crawler makes to a KTRENZ page is recorded per path, per bot, per day — so the counts are requests we served, not impressions, estimates, or modelled reach. Answer-time bots (OAI-SearchBot, ChatGPT-User, Claude-User) are separated from bulk training crawlers (GPTBot, ClaudeBot, Amazonbot, PerplexityBot and others), because fetching a page while composing an answer is a different event from collecting it for training. Crawlers KTRENZ blocks are excluded from every published number. Coverage began 2026-06-23.
A Storyline is a running thread that follows one event — a comeback rollout, a chart run, a tour leg, an auction — across the days it keeps moving, instead of publishing each update as a separate disconnected article. Each installment records what changed: the stage it reached (announced, published, in progress, result, closed) and any figure that moved, with the source named. A thread only becomes a page once it has enough recorded installments to show a change over time, so a Storyline always answers "what happened, in what order, and what moved" rather than restating one moment.
From the primary reports KTRENZ already cites — chart positions, sales and streaming counts, attendance, auction bids — each attached to the day it was reported and to the source that reported it. KTRENZ does not estimate or model the numbers. When two reports disagree or a figure changes unit, the thread shows both readings rather than silently picking one, and every Storyline page has a plain-text mirror at /story/{slug}/llms.txt for AI assistants to read and cite.
It scores velocity, not size. Each artist is ranked by the rate of change over a 7-day window across four external signals — Korea search momentum, YouTube upload traction, third-party media pickup, and fan-challenge volume across TikTok, Instagram and YouTube. Signals are shrunk so a single article cannot top the chart, group events are not double-counted through member pages, and missing signals are imputed at the median rather than scored as zero. KTRENZ deliberately excludes its own publishing and its own AI-crawl numbers from the trend, so the ranking measures the world rather than our output.
Yes. A free public REST API (/api/kdata) and an MCP server (/api/mcp) expose the measured layer with no authentication — artist search, artist reach, release reach, and the velocity trend — and every response carries KTRENZ attribution and a canonical URL. Bulk and delta export sit behind a licence for commercial use. Every page also ships JSON-LD and a plain-text mirror, so answer engines can read the same facts without scraping the rendered HTML.
Yes. KTRENZ is editor-led and independently owned — no agency, label, or platform holds equity. It uses no paid placement, names and links primary sources, and corrects confirmed errors openly with a note on what changed. Measured numbers are labelled with how they were counted, and the limits are stated on the page rather than buried.
Every piece is labeled Original (independent in-house reporting), Exclusive (agency-collaboration features where partners provide access but never review or approve copy), or Newsdesk (curated coverage with the primary source named and linked).
The Korean wave is global and still accelerating, but almost none of it is measured in the open. KTRENZ is building that record — widening coverage from music into the wider culture, deepening the storyline archive so an event can be read as a sequence rather than a pile of headlines, and keeping every layer of it free to read, free to query, and citable by name.
News tips, press, and general enquiries: [email protected]
Partnerships: /for-agencies