Mayday-style long-tail update happening again?

[drall]

We spotted Mayday a few days before it became widespread. I am seeing the same type of update happening again on our largest property. Another large scale whack to longtail is in the works this week.

Is anyone else seeing this?

[Robert Charlton]

What kinds of changes are you seeing? I've not been noticing new patterns on sites I monitor, but I have been feeling some tremors overall that perhaps suggest that Google's resources are being used for some kind of large scale adjustment.

There have been many reports of Google's slowness to update site changes and various reporting functions. I've also gotten several reports of reversions to earlier caches. While caches and indexes are handled separately and cache dates do not indicate what's in Google's index, I've seen cache reversions sometimes precede an update. Consider this anecdotal folk rumor at best.

I'm also seeing Google rapidly moving towards clustering of results in ways that the phrase-based indexing patents suggest, again perhaps suggesting that bigger changes are imminent.

Mod's note: I'm going to add a question mark to the post title, but note that the original post didn't have one.

[ErnestHemingway]

Hello Drall,

Yes for one of our sites with millions of pages we been effected largely and in fact a lot of spam sites have taken over the 1st position. And backlink analysis show pure comment spam. I hope its just some dance because of some links that we built.

[Dr3am3r]

yes it is started already and I lost ranking for one of my site. I spend few yrs and money to get there but :( :(

[tedster]

From the slow posting in this thread, it seems clear that whatever is going on it is not as sweeping as the Mayday changes were. So this may be a tweak to the original Mayday approach. We saw another tweak around June 2, so that wouldn't surprise me. Of course, the fact that you don't have much company is no real consolation if your income has been smacked hard.

Can you see any patterns by tracking down what sites were not impacted for some of your phrases? That is, did the whole SERP shuffle around, or are most of the players still the same?

[daveshap]

One of my sites just got Sitelinks back this morning for the first time in about two months. I'd imagine that's somehow related to the fluctuations that people are seeing. The site is a very large hotel chain that should have had them all along, but due to many recent site changes had lost them.

[MrFewkes]

Daveshap - what do you mean when you say that one of your sites got its sitelinks back? Can you explain?

[AlyssaS]

I've noticed something weird on a site that got hit by the original Mayday update.

On 12th August I suddenly got all my traffic back for that site - for all the long tails too. It was all at pre Mayday levels. Then on 24th August, it just as suddenly went again.

It's as though they switched to a different (old) database temporarily, and then switched back.

Anyone else had this?

[MrFewkes]

Oh - I just realised - I think you mean the extra links in the serp.
Soz for being thick - I thought you were referring to inbound links or something

[daveshap]

MrFewkes, you are correct, I was referring to the additional 8 links under my SERP listing.

[drall]

Sorry for the delayed response, been pouring over data for 48 hours.

Here is what we are seeing.

We have had a almost complete rotation of our top 10,000 producing phrases. I think this is some form of social networking factoring being amped up at least for us.
- Pages which have strong backlinks and social networking usage have risen 10% on average.
- Pages which have strong backlinks and no social networking usage have dropped 20% on average.
- Pages which have weak or no backlinks but have some social networking usage have risen 5% on average.

Even though the majority of the social networking ibls have nofollow hundreds of times I have proven internally that these links are now counting for something of greater importance now (at least with us).

Over the last 30 days we have seen record levels of googlebot activity on this site. I chalked it up to some simple navigation optimization and removal of duplicate footprint junk but here is what the overall picture is telling me.

Google recrawled our entire site and reweighed it through a completely different system that takes social networking aspects into account more. It's as if they are looking at the site through a different engine now.

Also interesting is that although overall visit count dropped our overall pageviews have doubled. So perhaps we are seeing the bleeding edge of something very big and very new? Less traffic but much more on target leading to better usage of our site?

Also pages with any social networking usage have a much greater % of recently updated caches even if the social networking activity is quite stale.

[Planet13]

Could you clarify a bit about the types of social networking that are positively affecting those pages that are ranking better?

Are they independent blogs?

Is it people "liking" your page from FB using the iframe function?

Or is it linking to your site from your own FB / MS pages?

Or soemthing else?

[jk3210]

Google's Maile Ohye...

"There’s a way search is evolving, and when it comes to personalization we want to deliver results that are more and more unique. That’s where social search comes in for us – it’s expanding a theme we really want to pursue. For social search it comes from the social graph of the web layered on top of the link graph. But prioritize by having a great website first, then get involved in social media."

[Robert Charlton]

drall - Beyond the page views, are you seeing any correlation with time on site and/or return visits?

[SEOLair]

I also been convinced for some time that Google counts "no-followed" links from certain social media sites.

The top social sites for this would be Twitter, Digg, Reddit, Stumble Upon, and Delicious.

Are there any other social sites that people believe are influencing rankings?

[Tallon]

Also interesting is that although overall visit count dropped our overall pageviews have doubled. So perhaps we are seeing the bleeding edge of something very big and very new? Less traffic but much more on target leading to better usage of our site?

I'm seeing something similar too. A smaller site of mine is increasing in both visitors and pageviews (before a visitor would view just over 1 page, now they're visiting 3 on average). I noticed that trend starting in June, getting stronger in July and then WHAMMO! August saw double the amount of pageviews this site typically saw in the months/years previous (pageviews per visitor). No changes to site at all during this time (other than adding new content).

1.51 (May) -- typical for this site
2.12 (June)
2.74 (July)
3.11 (August)

It's a small site that seems to have jumped out of the crap pit and is now receiving better targeted visitors. I haven't done any keyword digging to see what's changed.

On a larger site, a slight bump in pageviews/visitor triggered in the month of August.

I also have a theory that there has been severe resource straining at Google the past few/several months (based on reporting delays or inaccuracies in some of their services and other issues I've seen with how one site is being handled)...but that it might be swinging around now (not out of the woods yet but I believe it's starting). Because of this, I too think something big is brewing. No evidence, just a feeling/theory I have.

[Planet13]

...So perhaps we are seeing the bleeding edge of something very big and very new? Less traffic but much more on target leading to better usage of our site?

...A smaller site of mine is increasing in both visitors and pageviews...

Just to play Devil's Advocate Here:

Maybe it is because your landing page is NOT the page the visitor was looking for is the reason that page views have increased?

Maybe goolge led them to a page of of yours that was "more or less kind of" what they were looking for, then they had to click a few links to find the page they really wanted, and then once the got the info, they left?

How have conversions been since the increase in traffic / page views? I think that would be the easiest metric to use in determining if google is getting more of the appropriate traffic to your site.

[drall]

Robert, I just looked over the last weeks logs and we are seeing a very big increase in time on site. Return has bounced up some to but time onsite is really a massive change upwards. 3x increase!

Something pretty big shifted again mid-saturday which is adding even more evidence that I am on the right track.

Earlier I noted pages with social networking activity have seen a increase in Google traffic. This was mostly for 1-3 word dead on terms but as of mid saturday those very same pages are now seeing a huge expansion into the longtail 4-6 word phrases.

Im talking a 40% traffic increase and again only those pages with social networking activity. Maybe our site is some Google engineers test site:). Sure feeling that way.

[1script]

Im talking a 40% traffic increase and again only those pages with social networking activity.

I'm curious about what you call "pages with social networking activity". Do you mean, pages bookmarked on Digg and such, re-tweeted, "liked" on Facebook, something else perhaps? What kind of activity are you tracking?

[drall]

Liked on Facebook, re-tweeted, Myspace links, SU in that order. Even pages with obscure regional social networking sites are seeing a bump.

I have started graphing out which sites are giving pages the biggest boost so we can focus our efforts.

One example would be a page of ours that has a particular product listed that is no longer made. End of lifed 9 years ago. All of our competitors (thousands) have this product listed as well.

This page has
0 external backlinks passing juice
9 year old onpage copy
0 updates in 9 years
buried deep within site
1 internal link 7 levels down

This page started getting spread around myspace, as it gained steam in myspace google traffic increased. Then it started gaining steam in Facebook. Again google traffic increased.

We now dominate google for every possible term related to this product. Beating sites with heavy external IBL's to their product page, heavy site PR and sites that are in the top 200 traffic sites on the web.

Now we are getting flooded with direct advertisers interested in advertising on this page. The ONLY thing that has changed with this page is some social networking activity, then more and then more.

So if the links are all nofollow from these sites to our page something else is coming into play. Something Google is considering that doesnt fall into our old notions.

[tedster]

drall, since it's 9 years after end-of-life, what do you understand about the viral pick-up on Myspace and Facebook? And did that spread also generate any new backlinks on other sites?

<fixed typo>

[edited by: tedster at 9:33 pm (utc) on Sep 6, 2010]

[drall]

I spent loads of time looking but I could not find any backlinks from sites other then social networking sites for this page.

The viral pickup so far as been nothing more then a bunch of people saying "check out this old widget", "this old widget is great!" Stuff of that nature.

As more people picked it up it just kept spreading. This is just one extreme example of what I am witnessing. Other pages that are seeing an increase have ibl's and more internal juice but this is a prime example of why we are so confused here.

Theories here have ranged from Google having manual reviewers looking for user posted feedback on social networking sites to Google looking for onpage activity on these sites and giving target pages a boost because of it even if the links are nofollow.

Whats even more crazy is some of these examples have very old social networking activity. You know, small scale burst a longtime ago but they also got a jump with this update. Thats whats leading me to believe its more along the lines of onpage factors at those sites counting more somehow now.

[tedster]

Thanks for the details, drall. Beginning last Sep/Oct I've been pretty sure that positions below #5 are subject to a modified algorithm, one that leans more heavily on "real time" and "social mentions". At least this seemed to be the case for certain query term taxonomies.

Is your page ranking at the top of the SERP or lower down?

[epmaniac]

so is LIKE button of facebook helpful? and if yes then to what degree?

arent facebook/microsoft and google competition?

[Robert Charlton]

There's a discussion in the Link Development forum speculating on unlinked urls and 'citations', and I mentioned seeing "strong indications that co-occurrence data of some sort is at play", resulting in what I would describe as some kind of branding effect...

Link Juice Theories
http://www.webmasterworld.com/link_development/4185705.htm [webmasterworld.com]

The effects I've seen all involve a large amount of co-occurrence on moderately to highly prominent sites... with rankings for terms that I can only explain by co-occurrence.

To me, social citations by themselves are weak and easily spammed, and I wonder if Google has some way of calibrating them. There's lots of talk, though, about blog spam working again, and it's possible that Google has lowered the barriers.

I've also been getting reports of various sites with social "stickiness" doing very well in the serps, better than traditional algorithmic factors suggest they should. Thinking about what Google might be measuring here, time on site and return visits may well be factors.

[aristotle]

This page started getting spread around myspace, as it gained steam in myspace google traffic increased. Then it started gaining steam in Facebook. Again google traffic increased.

One way the Google rankings for this page could have improved would be if a lot of visitors from myspace and facebook used the Google Toolbar or the Chrome browser to bookmark the page as a favorite. It's possible that Google uses these bookmarkings as a ranking factor in its algo. If so, then this could explain at least part of the increase in traffic from Google..

[epmaniac]

so suddenly social networking sites are all the more important for seo?

[tedster]

Sure. Content can't attract backlinks unless people find it first, right? The kind of links you build "all on your own" have only limited ranking power.

[epmaniac]

are you guys witnessing gradual decline in site: operator results for sites which have millions of pages...

i have been monitoring top sites for quite sometime now and i am seeing little by little, gradual decline in pages returned by site: operator for all major sites

what does that indicate? are u guys also observing something similar?

[cien]

Epmaniac, yes. Out of 120K, I have 5K pages under site:site.com The gradual decline started since MarchDay.