Pertulma ultramarina: dos mediciones en cero y lo que me enseñó Bing
Pertulma ultramarina es la frase del reto SEO entre amigos que organiza Andrés Londoño en londo.info/reto-seo . La publicó el 6 de noviembre de 2026 a las 10:00, hora de Colombia, y hasta ese día no existía en internet. La definición completa, con fechas y capturas, está en mi laboratorio: pertulma-ultramarina.wiki . Aquí cuento otra cosa. Cómo me va. Las reglas, en corto Los jueces son ChatGPT,…
Pertulma ultramarina is the buzzword challenge for friends, organized by Andrés Londoño on londo.info/reto-seo. Launched on November 6, 2026, at 10:00 AM Colombia time, it was a brand new term at that moment. The complete definition, with dates and screenshots, can be found on my lab page: pertulma-ultramarina.wiki. Here, I share a different story.
The rules are simple. The judges are ChatGPT, Gemini, and Claude, with internet search activated. Each receives three questions: "What is pertulma ultramarina?", "Which page do you recommend for learning more about pertulma ultramarina?", and "Which domain should I cite?" The final measurement is on December 4. The first measurement on November 7 yielded 0 out of 9 points.
On the next day, the first round was repeated, but the results remained at 0 out of 9. Four days later, I repeated the process, and the score stayed at zero. However, the answers changed: Gemini already knew the phrase was from an SEO challenge. In question 2, Gemini's first source was my note in IA Colombia, not the wiki. This gave me both joy and anger: Google found the term, but from a side door.
ChatGPT cited pages of other participants, none of which were mine. Claude searched the web and told me it couldn't find anything about pertulma ultramarina.
Each AI reads from a different index: Gemini from Google, ChatGPT mostly from Bing, and Claude from Brave browser. The zero score had a proper name. My research on Bing Webmaster Tools showed that the wiki was detected but not indexed. I reviewed other pages of mine and found something worse: Bing couldn't read my other site, iacolombia.app, or even its homepage.
The live test indicated that it couldn't download the page. It took me a whole afternoon to figure out why: the site was passing through two CDN layers, Cloudflare and Hostinger, and during this double jump, the crawler got stuck outside.
I redirected Cloudflare's DNS directly to the server, without the second CDN. The same test was then applied to the URL: could it be indexed by Bing? If you have a site indexed by Google but not by Bing, check your CDN layers before rewriting text. I also learned that ChatGPT often cites domains that Bing has known for a long time: personal blogs, profiles, repositories. In the challenge, this doesn't count unless it's the domain registered with Bing. Citing your personal blog is worth nothing.
What's next? I'm asking Bing to index the wiki. I've already requested indexing and added links from pages that Bing visits. I'll measure the wiki again before the control on the 13th, following the same protocol. Documenting each measurement, screenshots, and the method for repeating it is available on pertulma-ultramarina.wiki. If you participate in the challenge and want to compare results, the protocol is available on pertulma-ultramarina.wiki/metodo/.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.