| 7:55 pm on Oct 27, 2011 (gmt 0)|
I don't know about any approaching Panda flux - but I also see outrageous changes over short periods of time. I'm beginning to suspect some kind of statistical sampling is at play. There may be several versions of the algo in live use at any time. Some could show more frequently than others - and then the stats for each algo version would allow Google to "learn" more quickly what works better for their user.
| 11:27 pm on Oct 27, 2011 (gmt 0)|
We're in the 9th month of panda. When are you guys going to learn there's a new update on every given wednesday/thursday (with very few exceptions). Even my girlfriend figured that out.
| 1:08 am on Oct 28, 2011 (gmt 0)|
But those regular changes are not always Panda related. There's a lot going on in the algo in addition to the Panda portion. Writing it all off as Panda is not a smart thing to do.
| 10:59 am on Oct 28, 2011 (gmt 0)|
Google is tweaking its algo almost every hour today (Friday) we can see ranking fluctuations on Page 1 for a very competitive keyword, all sites have moved places except the top 1 spot.
| 11:38 am on Oct 28, 2011 (gmt 0)|
a SEO agency checked 100 000 keywords over the course of 24 hours, they measured that about 8% of all results varied in those 24 hours.
in this google youtube video the people of google explain the proces: [youtube.com...]
| 1:25 pm on Oct 28, 2011 (gmt 0)|
may be so, but on the wednesday/thursday time frame there's (almost) always one with a larger impact than the usual daily fluctuations.
| 2:42 pm on Oct 28, 2011 (gmt 0)|
I find it odd that when my forums are most active Google sends the lowest amount of traffic per hour on that day. Right now that's primetime Saturday evenings. You'd think I'd be getting more traffic since it's obviously a busy day/hour for that topic.
| 7:32 pm on Oct 28, 2011 (gmt 0)|
pontifex... I've been seeing for a while now that the serp "refinements" (sitelinks, multiple results for a site, title changes, etc) are getting more and more long tail, with local Place results appearing where I hadn't seen them before. I've been assuming that Google has been using these to test both user satisfaction with serps layouts, and also to test, say, "engagement" with pages made prominent by the extra exposure.
Very recently, I'm also seeing the Place results appear for much longer tail phrases than I've seen previously, and now I'm seeing different sets of pages being tested for Sitelinks. I'm noticing many more title changes too. While I do associate some appearances of these "refinements" as prefaces to a Panda iteration, the refinements have been appearing since well before Panda. If Google has been using them for testing, they were most likely testing other algo changes prior to Panda, and are likely testing all sorts of algo changes now.
With these changes you see, are you seeing any of the kind of traffic throttling as we'd discussed previously? The many current reports of "zombie" traffic also make me think that some kind of bigger changes are on the way... but, again, traffic throttling, and traffic shaping and zombie traffic also had preceded Panda by quite a bit. See...
January 26 2011 Change - Back to "Zombie Traffic"
Are we entering another Panda user-engagement calibration phase?
Google & Traffic Shaping - a hidden method to the quality madness?