Here is a fun experiment. Ask three people in three different cities what was trending on X on June 23, 2026. You will get three different answers, and all of them will be right.

I went looking for “the trends” that day and found a mess. In the US, X was wall-to-wall NBA Draft. The top entries were names like Carr, Philon, Koa Peat, Cenac, and Isaiah Evans, plus the team accounts: Lakers, Grizzlies, Sixers, Mavs, Thunder. Meanwhile a separate global feed showed #ENGGHA, #SecretStory, #WUTheSeriesEP8, and Hornets. And the month-level breakout searches looked nothing like either: AI Ethics, AI Video Generator, AI Image Enhancer, Programmatic SEO, and a baby bottle washer.

These are not slightly different lists. They are different planets.

We still talk about trends like there is one shared conversation happening, a big public square where everybody sees the same thing at the same time. That idea is dead. The algorithm slices the audience by location, by time window, by inferred interest, and serves each slice its own private “trending” panel. The basketball fan in Memphis and the AI marketer in Berlin both believe they are watching the pulse of the planet. They are watching a personalized cable channel labeled with the word “public.”

The counterargument is fair: trends were always local. X has shown location-specific trends for over a decade, and Twitter before it. True. But the gap used to be small. A genuinely huge event broke through everywhere. Now the slicing is so fine that even within the United States, different time windows on the same day produced different trend lists. The shared spike is getting rarer.

This matters because we make decisions based on “what people are talking about.” Marketers chase it. Journalists write about it. Politicians fear it. If the trending list is a hall of mirrors built per-viewer, then “the public is outraged” often means “my specific feed is outraged,” which is a much smaller and much weaker claim.

So stop trusting trend panels as a map of reality. They are a map of you. Treat them like a mirror, not a window, and you will be wrong far less often.