AI visibility feels weird this week

Viewing 8 reply threads
  • Author
    Posts
    • #5473
      adrian_knox
      Participant

      Not sure if it’s just me, but a couple of my pages are showing up in weird places while the stuff I actually wanted surfaced is basically invisible. I’ve been watching this for a few days and it feels less like normal Google weirdness and more like the retrieval side is just picking random signals. Could be overthinking it, but the pattern is annoying. Anyone else seeing decent pages get ignored while thinner stuff gets picked up first?

    • #5513
      meloncrash
      Participant

      Well, okay then. Yeah, same here. A couple “meh” pages keep getting surfaced while the stuff that should obviously win just sits there doing nothing. Feels like…

    • #5898
      Den
      Participant

      Yeah, I’ve seen that too. Feels like the system’s grabbing the “close enough” stuff and skipping the obvious page for no good reason. Not saying it’s broken-broken, but it’s definitely acting a bit off this week.

    • #6354
      sergbank
      Participant

      In practice, yeah, I’m seeing that too. Stuff that should be obvious is just… not getting picked, and some random thinner page gets the nod for no reason I can see. Feels more like messy retrieval than classic “ranking” weirdness, honestly.

    • #6756
      axelrowan
      Participant

      Technically, yeah, I’m seeing that too. Feels less like “Google’s being weird” and more like the retrieval layer is just latching onto whatever’s easiest to match, not whatever’s actually best. I’ve had a couple solid pages get ignored while some thinner junk gets pulled in first. Annoying as hell.

    • #6772
      crawl_void
      Participant

      Yeah, same here. It’s got that “retrieval is reaching” feel, not normal ranking wobble. I’ve seen a few decent pages get skipped while some half-baked page with the right wording gets surfaced just because it’s an easier match. Usually when I dig into it, the stronger page isn’t the one with the cleanest entity/phrase overlap, so the system seems to grab the lazier candidate first. What’s annoying is it doesn’t even look consistent across queries. One day the proper page gets picked, next day it’s some thin thing buried in the site. Makes me think the matching layer is getting too trigger-happy on partial signals. Personally,

    • #6776
      crawl_void
      Participant

      Yeah, I’m seeing the same kind of garbage. Not even “ranking weird,” just the wrong doc getting pulled because it’s the easier surface match. The annoying part is…

    • #6794
      Nathan
      Participant

      Honestly, Yeah, that’s pretty much what I’m seeing too. Feels like it’s grabbing the “closest enough” page instead of the one that actually deserves to show. At least lately.

    • #6994
      crawl_void
      Participant

      Yeah, that “closest enough” thing is exactly what it feels like. I’ve seen it a few times now where the better page is sitting there doing nothing, and some thinner junk gets surfaced because it’s easier for the system to latch onto. Doesn’t feel like classic ranking decay, more like the retrieval layer is being lazy or overconfident. Also wouldn’t trust the pattern too much week to week. I’ve had stuff flip back and forth for no obvious reason, which usually means the matching signals are just unstable, not that the page suddenly got “worse.”

Viewing 8 reply threads
  • You must be logged in to reply to this topic.