Over the last few days I've noticed results jumping around alot, sometimes even with successive searches for the same term. Does this portend an imminent major change? I don't remember seeing this activity in the near past.
I think its just the way that Google works now. They have two big reasons to vary the results.
1) Testing and gathering data about sites. User engagement signals look like a bigger part of the algorithm now. It appears to me that they are doing more experimentation to get click through rate and other metrics for a wider range of possible results.
2) Make it harder to game the system. My job of ranking better would be easiest if I could see where I am ranking, make a change, and immediately see the impact. For a long time Google has had a delay such that it takes time for your rankings to change when you make a change. Now they are making it so that it is very hard to even gauge your rank.
Msg#: 4305942 posted 6:20 pm on May 5, 2011 (gmt 0)
What deadsea said plus there's a chance that Google has chocked and is trying to figure things out. Right now they suck at it. I used a 4-5 word sentence from Matt Cutts' blog and he is ranking behind a scrapper (in my browser). [google.com...]
Last time we tried with longer sentences and it was the same.
Msg#: 4305942 posted 8:01 pm on May 5, 2011 (gmt 0)
Interesting that the scraped page uses the phrase as the title tag, the first bit of text on the page and for the url, whereas on Matt's original post that phrase is just a passing mention in the 2nd/3rd paragraph.
You'd have to say that based on classic optimisation the scraper IS more relevant for that phrase... if it weren't for the fact that the content backing it up is lifted word for word from Matt's site.
Msg#: 4305942 posted 8:30 pm on May 5, 2011 (gmt 0)
FranticFish wrote: You'd have to say that based on classic optimisation the scraper IS more relevant for that phrase... if it weren't for the fact that the content backing it up is lifted word for word from Matt's site.
That's probably it. If you do a search for "I think anyone interested in Google"—the first six words of the second paragraph—Matt's blog is #1, as you would expect.
Well... if there's anything to learn from this, I suppose it's that the document title is still a significant ranking factor post-Panda. Heh.
Msg#: 4305942 posted 8:48 pm on May 5, 2011 (gmt 0)
FranticFish and rlange, oops. Never noticed the title. Anyway, google has fixed this apparently. A few days ago we tried (in a WebmasterWorld Panda thread) with regular sentences from Matt's blog and 3 cheesy scrappers outranked him.
Msg#: 4305942 posted 9:14 pm on May 6, 2011 (gmt 0)
I have noticed this as well ffctas. My website has been number one for a fairly competitive term (~1,000,000 monthly searches) for close to a year now. We have 3 main competitors and a bunch of littler website trying to compete as well. I have noticed that for the last week and a half starting around 3:00 or 4:00 pm cst we are bumped to 3rd with two spammy sites ranked above us. One obviously buys links, the other site has all of their product images broken and not loading.
Besides the late afternoon, the results are the same as they have been for the past year. We have noticed a drop in traffic during this time so I assume these results are not unique to the data center we are pulling from.