| 1:32 am on Aug 10, 2011 (gmt 0)|
Thanks for that report. It's very hopeful... and it's about time! Panda's initial frozen verdicts were really nasty.
| 1:36 am on Aug 10, 2011 (gmt 0)|
Yes, I could not believe it last Friday and today the second recovery, just unbelievable. Awesome content + usage metrics really matter.
| 3:33 am on Aug 10, 2011 (gmt 0)|
I am also seeing traffic bouncing back from 4-Aug-2011 on one of my client's site which was hit by panda2.3. Now the traffic is almost equal to pre-panda 2.3 period.
| 6:52 am on Aug 10, 2011 (gmt 0)|
|...Awesome content + usage metrics really matter |
Care to elaborate with a few crumbs of advice for the remaining Panda leppers?!
| 3:07 pm on Aug 10, 2011 (gmt 0)|
|Care to elaborate with a few crumbs of advice for the remaining Panda leppers?! |
Ditto! Especially the 'usage metrics' part..
| 3:43 pm on Aug 10, 2011 (gmt 0)|
"usage metrics really matter" he is probably referring to the amount of time, page views, and bounce rate of the site.
| 4:00 pm on Aug 10, 2011 (gmt 0)|
|"usage metrics really matter" he is probably referring to the amount of time, page views, and bounce rate of the site. |
Does anyone have any hard evidence of this? I mean real data. Its all the talk, and makes sense, but we have not been able to definitively prove this theory.
| 4:39 pm on Aug 10, 2011 (gmt 0)|
I don't think usage metrics are a cause - they are an effect of the update - my own bounce rate, amount of time etc all changed AFTER the updates, not BECAUSE of them.
In other words, even though my traffic dropped, I think google appears to be serving better (or the right) pages from my sites, thats why bounce right appears to have dropped and time on pages has increased - because the "lower quality" pages aren't getting the traffic they used to.
| 4:44 pm on Aug 10, 2011 (gmt 0)|
It has been discussed at length these metrics are becoming more and more part of the site rankings. If you look at adwords/adsense part of the bidding cost placement in adwords/or cpc in adsense is based on the sites metrics. Google indeed places a weight on this in the formation of the score your site is given.
"So the PR you see publicly is different from the number our algorithm actually uses for ranking. Why bother with a number thatís at best three steps removed from your actual goal, when you could instead directly measure what you want to achieve? Finding metrics that are directly related to your business goals allows you to spend your time furthering those goals."
This is speaking directly about site metrics. Furthering your goals is all about getting visitors, engaging, and converting whatever it maybe.
| 5:00 pm on Aug 10, 2011 (gmt 0)|
|Does anyone have any hard evidence of this? I mean real data. Its all the talk, and makes sense, but we have not been able to definitively prove this theory. |
...scenario, i tell my 35.000 forum members to search on google for "keyword xyz" click on my site that is on page 2, and stay there for a minute and then click to another page and stay for a minute and so forth for three more pages and than only leave the site, will that be UBO, User behavior optimisation?
| 5:01 pm on Aug 10, 2011 (gmt 0)|
I dunno man, but pretty puleez give that a try and see what happens!
| 5:03 pm on Aug 10, 2011 (gmt 0)|
|It has been discussed at length these metrics are becoming more and more part of the site rankings. |
Talk is cheap and everyone knows that much of Google's advice is actually wishful thinking (what they want from us) rather than based on real factors tested within the algo.
I'd love to believe in site metrics, but how the heck can the resilience of dead pages (or plain terrible pages) in plenty of searches be explained? I've seen pages that should have been torched on this basis but with each panda iteration they rise!
| 5:12 pm on Aug 10, 2011 (gmt 0)|
Before I would accept the whole user metrics theory as gospel, I would also look at how well the sites that are linking to me are doing.
Did those sites that have backlinks to your site suddenly regain their rankings, too? In that case, maybe the PR that is flowing from their site into your site is higher.
I say this because if user metrics played a significant role in ranking, then some of my highest ranking pages would have vanished from the SERPs long ago.
| 5:15 pm on Aug 10, 2011 (gmt 0)|
|...scenario, i tell my 35.000 forum members to search on google for "keyword xyz" click on my site that is on page 2, and stay there for a minute and then click to another page and stay for a minute and so forth for three more pages and than only leave the site, will that be UBO, User behavior optimisation? |
Why not find someone on fiverr to do it with bots and proxies? I'm pretty sure any 13-year-old kid from the Ukraine could figure out how to spoof user metrics pretty easily for a few dollars.
| 7:03 pm on Aug 10, 2011 (gmt 0)|
I puzzled now
How do you go from 2/4 sites re ranked to live Panda updates ?
I guess I must have missed some other posts or info
| 8:19 pm on Aug 10, 2011 (gmt 0)|
Google has informed me that there was no Panda update since Panda 2.3 and no, it is not a rolling update as of now.
| 8:22 pm on Aug 10, 2011 (gmt 0)|
|Google has informed me that there was no Panda update since Panda 2.3 and no, it is not a rolling update as of now. |
So what you are saying, in essence, is that the ranking gains the original poster mentioned are unrelated to Panda, right?
| 8:30 pm on Aug 10, 2011 (gmt 0)|
user metrics like bounce rate and time spent on site have been around since the dawn of time -- since the very beginnings of SEO. why would they suddenly start paying them more attention now? it doesnt sound very plausible to me.
| 9:01 pm on Aug 10, 2011 (gmt 0)|
londrum - in my humble and surely non-comprehensive opinion, it seems to me that the previous algorithm heavily favored link metrics (as per the original PageRank), and when Amit Singhal came into the picture, he revamped the algorithm entirely. It took a long time and what we got was panda.
To your point, user engagement metrics have been around, sure, but implementation of them has never been as apparent. As an example is an option to "Block [domain.com]" when you immediately bounce back to a SERP. This is surely a harbinger of a new type of implementation.
I can't seem to locate the link where Amit pretty much states that he was given card blanche to change the algorithm, and he took a different approach than Larry Page did with his initial pagerank algo. If I find it I'll post it.
| 9:20 pm on Aug 10, 2011 (gmt 0)|
The reason Panda updates aren't real time is they require a huge amount of resources.
| 11:05 pm on Aug 10, 2011 (gmt 0)|
I find it unlikely that user metrics play any role in the Panda. However, Google's vague answers likely make you think this is a possibility and do all you can to improve these metrics just out of fear.
Sure they can do it to Adwords, because it's an even playing field, they have the same data on all the web sites. But not all web sites use WMT, analytics, Adwords, etc.
BTW, I am also seeing one of my sites improve today, it is not a total recovery, but it is an improvement. My other main one is still down. This makes me think that there are also various levels/stages of Panda for more than one factor. But I don't think it is user behavior. I am sure it is all down to content.
| 11:35 pm on Aug 10, 2011 (gmt 0)|
Whether user engagement metrics are part of Panda or not - why WOULDN'T a webmaster work to improve things like bounce rate, page views per visit, average time on the site, etc? Surely these have always been a valuable tool in the website success toolbox, right?
| 12:26 am on Aug 11, 2011 (gmt 0)|
There may be many reasons, some obvious ones are time, money, knowledge. It may require a huge redesign, which can be expensive, along with a lot of reading and research. If the site is already performing well enough financially, then they may not bother. However, if Google puts their foot down (or appears to give that impression) that these changes must be done, then it will certainly make webmasters react more to it.
| 5:07 am on Aug 11, 2011 (gmt 0)|
I don't know if more sites will follow but try at least to track your events with a statistics package so you will know the exact metrics of your site, which links your visitors click and so on.
As soon as you run this tracking for 1-2 weeks try to improve these numbers, it won't hurt.
Since I did both, adding great (expensive) content and improving user metrics at the same time I can't isolate a single factor so doing both at the same time is probably the key. The sites have been punished at a Panda date and have been released on a non-panda date, therefore I'm sure it is a Panda release.
I don't care what Google says since I use Bing+Yahoo (my favorite search engines) for my converting traffic so I don't need Google at all at the moment but of course 10k additional visitors daily are nice.
Bingdude has been more than helpful, Thanks a lot! He also exactly answered questions about bounce rate which might help you with goog.
You will never see this clear language from goog. 5 stars for Bingdude!
| 11:11 am on Aug 11, 2011 (gmt 0)|
User metrics is in fact a key factor.
why WOULDN'T a webmaster work to improve things like bounce rate, page views per visit, average time on the site, etc? Surely these have always been a valuable tool in the website success toolbox, right?
Tedster is absolutely right.
Panda 2.4 is not running just yet and would not be until the end of August, but the date is still unknown even to G. staff at this stage, but what is known is that there are few major changes in algo based on the information collected since February up until July.
As we have all already witness the major changes in Page Rank first in June and second in August, this has already provided us with lot more information than we can only think about see the example below based on my main site stats
Page 1 - 150,000 unique visitors/m -P/R 4 - NO user engagement obvious signals - dropped to P/R 2 after 27th of July
Page 2 - 55,000 unique visitors/m -P/R 3 - 25,000 Twitter posts, 8,00 facebook likes, 93 +1s, also Digg and Delicious posts and shares - whentt up to P/R 5 after 27th of July.
Content on both pages is about red widgets on page 1 and about blue widgets on page 2, but red widgets are downloaded lot more frequently than my blue ones.
the question is still being unanswered, but at least stats are going up slowly but surely.
| 11:25 am on Aug 11, 2011 (gmt 0)|
Planet13, honestly, I do not know. Just relaying the message.
| 1:25 pm on Aug 11, 2011 (gmt 0)|
people are saying why wouldn't webmasters try and improve user metrics, whether they are part of panda or not, and okay that is true. but how do you measure an improvement in bounce rate? for a site like wikipedia, success could be a 100% bounce rate -- as the user found the info on the page they were sent to. but on another site success could be a 0% bounce rate, because they followed a purchase all the way through a shopping cart. its not like an exam paper, when everyone knows what is good and what is bad. the truth is, that each site could have a completely different bounce rate and each would be measured as a success. its the same with other user metrics like time spent on site. if you are looking for a fact on wikipedia then you could be in and out in 30 secs, which would be a "success". but on a shopping site 30 secs would be a disaster.
if these things are now part of the algo, then how do we know for sure what google considers a success? working your bounce rate up to 100% a la wikipedia, or down to 0%, might do you more harm than good.
| 2:06 pm on Aug 11, 2011 (gmt 0)|
|User metrics is in fact a key factor. |
It's not just a factor, but a key factor? And that's a FACT?
I don't think you've got the facts to back up that statement.
| 2:06 pm on Aug 11, 2011 (gmt 0)|
|User metrics is in fact a key factor. |
What "facts" do you have to prove this?
I'm not talking about sites that were negatively effected by Panda; if you were hurt by Panda, change is most likely all that will improve the situation (as opposed to doing nothing).
However, statements like the above imply that if you improve the "quality" of your site, and subsequently the "user metrics" data Google collects, your rankings will improve.
Has anyone been able to prove, say for example, that lowering a sites bounce rate has improved it's rankings?
| This 60 message thread spans 2 pages: 60 (  2 ) > > |