Sid, I totally agree with you.
One of my sites, previously at positions 1 - 5 for several years for its major search term has dropped to position 15.
I have looked at all of the sites above mine, many of which are new to the top 50 search results and, for all I know, well beyond this.
All but two of the 14 sites above mine have much higher numbers of links into them as reported on Yahoo - no surprise there, maybe. However, in all cases except one, the top 14 sites are irrelevant, weak or content-thin - even allowing for my interest in the field, I can see the sites are a poor result for the user.
What they have in common is that they all have hundreds or thousands (in one case, 266,000) of links into the domain (example.com).
This is true even where the search results present a subdirectory of the domain (example.com/widgeting/)
and this subdirectory has few if any links directly into it: a fact which testifies to the low quality of the content in question, in my view.
One of the results presented
has in fact no text of kind except for a one word title on the page; no meta tags, nothing else whatsoever except a few gifs which are a comical representation of widgeting. Yet because this site could be seen as interesting, quirky or funny, it has acquired 3,600 links from blogs and such like (most definitely not from widgeting authority sites). Most of these links have obvious but relevant text in the links, so the result is that it's achieved position 8 in the search rankings.
There are also several sites like aboutdotcom which have a subsection of greater or lesser value on widgeting, but with few links to the widgeting: I assume they feature because of the huge number of links into the main domain.
All very odd, and not I think sustainable without a further decrease in the value of Google's results.
Something definitely happened yesterday, but it doesn't seem to be anything major.
I'm in a fairly uncompetitive niche and the results for the two-word phrases that I monitor hardly change at all for months at a time. However, yesterday I moved up one-or two places for some terms and down a few places for others. It's too early to say whether it is having any effect on traffic, but I'd expect it to be fairly neutral.
I have some relevant observations to share. One of my sites is very sensitive to even minor changes in algo as I observed within the last 8 months. I call it "my indicator" site.
It enjoyed #2 for months starting with the Jagger, and all of a sudden (yesterday), it dropped down to 51. So something happened, I thought.
As I said elsvere in WW, what I see seems more like a rollback rather than a tweak in the algo, since I see the same old ranking sites again in the SERPs.
What strikes me in current SERPs is the total absence of on-page factors. An example:
One of the top 10 sites in my niche ranks high solely by wirtue of anchor links. (kw1+kw2). KW1 is located on the upper right of the page (a language name), and kw2 is located elswhere down the page. The only relevance is the keywords in all links pointing to the page.
But there should be other factors that play a role also, since my site appears at #2 with an allinanchor: query.
It was more than just an algo change and it is still going on. I have a site that belongs to a client, sells high dollar widgets. It was just beginning to come out of the sandbox. 3-4 weeks ago after 10 months in.
On the 11th, did a site:mysite check and all the widget pages are deindexed and pr is gone. greybarred. Now, that is a third of the site. Today, the index page appears only when you click the omitted results.
Absolutely no reason for this. Google has twisted some knob that is just hammering this poor site.
If off page factors are involved, this site has a number of quality inbound links but that is far outweighed by the number of ezines and scraper sites that have articles I have written about the subject of these widgets. Plus, the dmoz scraper sites.
I had become complacent with the consistently high placement in the SERPs of a long time client's site. After reading your post I checked that site and found it has been bumped down significantly for all the keywords I looked at.
What also caught my eye was the replacement of page specific snippets with a single snippet for all pages. The snippet is old, very out of date. Haven't seen it since the site was new (eight years ago). It was the DMOZ description that G was using then, if I recall.
For most pages, the site has dropped from #3 of the first page down to the bottom of the first page, and worse.
Overall, (after only 15 minutes of looking at the wreckage), I see no pattern or logic to the changes in the SERPs on these pages. Seems very odd to have such radical changes of sites listed, in SERPs that have included the same basic group of sites for the past five years. Even during the big revisions like Florida, there was not such a wholesale change of familiar faces.
For a moment i was thinking "did i miss something in November" then i realised the posters from the US (where you do everything back to front!)and means 11th April!
Joking apart - i would agree with seeing the adjustment (and still believe Google is updating and not finished anyway, so could change) but im not certain that its to do with backlinks in the way mentioned.
For example, one of our dynamic pages rich in content, with good backlinks from other PR7 authority sites should be in the top five position for a two keyword search term.
On looking at the SERPS i see the site listed in position 6 for that two keyword search term (a bit lower than expected) yet Google has selected another page from the site (still relevent) about the keyword but this page has no backlinks to it at all other than a few internal ones.
My conclusion is that Google is now taking into account links to site pages overall and then looking at delivering what it thinks is the most relevent and current page, or something along these lines.
So if your site is widgets.com and you have say 20 backlinks to the home page saying "Blue Widgets" it will rank your most relevent intenal page on blue widgets in its serps rather than your home page.
In closing, i do think this update is far from over so anything could change yet.
does seem to be a new strengthening of the importance of a dmoz listing. Quite extraordinary given the high level of sites listed simply by a well placed seo/editor. Not to mention the not listing of competitor sites for the same reason.
And, thanks for the due diligence prior to posting.
I ran two, 2-word phrases which have been very good to me through Bourbon, Jagger -- even Big Daddy liked us for them. Results: high up in SERPS and Google Analytics keyword conversions on 10Apr06, all but gone in both on 11Apr06 -- nowhere on 12Apr06.
Researching the new SERPs leaders, I didn't see so much of a backlinks correlation, as I did a strong Google Directory/ODP dependence and somewhat of a commercial-over-informational preference.
Small sample and early days, of course, but something
is definitely afoot.
"..Something very subtle was changed in Google's algo overnight.."
Yes..some subtle changes here as well...I have a two word phrase that I work with in a very competitive niche...the site I work with has a very strong balance of on page / off page factors and has been very stable through all the recent carnage...(and, if you don't mind me correcting you...I believe you mean February 2004 .. as the infamous Florida update was on November 16, 2003 ... from what I remember...this was when Google integrated the word stemming algo from Applied Semantics)...
"..The balance between on page and off page factors and the quality of those off page factors changed.."
Yes...there has been a shift...but seeing some of the other sites that have come into the mix recently for this one sector I watch closely and work with daily...(well known price comparison sites)...I would say that "some of the algo dials" have been tweaked back to some historical setting (florida-esk)...and I would say that for them to arrive and settle in now would point to "off-page" factors...which they have very deep pockets with....
SO I am watching this evolution very closely as well....and certainly feel for the folks who are being heavily affected by recent changes...
(oh .. to sit in the Goolge war room and see the traffic flow ... even for a few minutes..)
After years reading about Google algos, trimming my sites, checking the links, looking the datacenters and so on, I don't just know I know nothing but I know I know less everyday.
Some friends sometimes ask me "How do you do it?" and I just can smile while I think "Gosh... if I knew..."
Some conjecture here. Looking at a few SERPs that shifted, the pages that advanced seem more loosely related to the query, and the pages that declined seem more precisely relevant.
I'm sure that would not be what Google is hoping to achieve, but that's the shift I'm seeing. There is something "brute force" involved in the change -- the sheer number of pages in a domain and the sheer number of backlinks is winning over more precise relevance. It's as if the semantic and content-related factors are dialed down and the mechanics are dialed up -- if that makes any sense.
Said another way, in the past, people have often wondered how a PR0 page could trump a PR7 page. I'm wondering if such examples are currently fewer after this bit of flux.
After Florida (which seems a lot closer than I remember ;) ) one theory I went with was that Google didn't like dead end pages ie those with no external relevant links and so put some links to very relevant sites. For some terms I found that this put a page with a particular term in the visible link text on page one for that term even though the page only contained that term once. Clearly this was very useful.
I also put links using my main target two word term and aimed these at leading directory pages that contained links back to my page. Thereby ensuring that Googlebot knew that I was on authority pages outside.
The page that has popped in above me is a dead end page, it is only on the topic of the two word term and so is relevant but the key reason that I can identify for it being placed so highly are hundreds of links to it, with the two word term visible in link text, but from none relevant pages.
I hope that they roll this back!
yeah a lot of that stuff too. A few 1000 paid links from sites on any topic you can name can push a sub page into competitive areas against sites that are totally devoted to the subject.
I have a property related site which was topping serps for location related keywords, the entire site has taken a smash. over 1k5 pages that were high ranked. All white hat. The the current winners seem to be spam sites? but old sites?
Well, I'm up from #2 to #1 today.
Apart from that small but happy change, I can see 4 'news' style articles in the top10 where there was only 2 yesterday.
Something has happened but it's a very minor something.
Whatever it is, it is still in motion this morning.
The SERPs for the keywords I mentioned above, have changed some more. Happily for me, the client's money keyword is back up to above the fold on page one again. The snippet is still old DMOZ. Other keywords bring up SERP's with client pages having recovered a few positions/pages since yesterday. Many of these are now showing the "proper" page-specific snippet, replacing that old DMOZ stuff that had recently appeared.
The pages that appear to be recovering are very tightly defined pages, including the index and one click deeper. (These probably have the "best" inbound links) Pages that are still AWOL include mostly directory style content, deeper in the site that carry lots of very on-topic outbound links, and few inbound links.
Duh...I used to feel like I had some control...
This really is an entertaining sport we play here.
I particularly like the way the goalposts keep moving and how the rules change without any announcements.
It does not suprise me that now big daddy is rolled out that we see subtle changes. They are probally trying to clean up the spam..
These dont seem to be that subtle. Google must just love this game ha, turn the knobs and see us crap ourselves. I think it is good really keeps us busy, and the industry moving.
Yep, it's Internet Cleaning Day, so the Googlers make the Tune Up too! They clean their temporary logs and databases. All the servers will be disconnected from the Internet for several hours, so the scavengers could move the server racks away from the walls and aspirate all the dust behind them.
It's a spring cleaning day, so no worries, your SERPs and your indexed pages will be back pretty soon ;)
number of pages indexed decreased.
what to think?
"It does not suprise me that now big daddy is rolled out that we see subtle changes. They are probally trying to clean up the spam.."
BD has been an unmitigated disaster and Google has yet to produce a decent index out of it.
The only thing keeping this from causing real damage to their reputation and being a financial disaster for Google is - while the webmaster community is aware of it and its becoming a topic on industry blogs, mailings, etc. - the general media hasn't picked it up.
Hmm, I just tried there a search on a camera model and a mobile (cellular) phone. Usually this returns gazzillions of junk pages. This time it actually led to some pages with information.
>> the pages that advanced seem more loosely related to the query, and the pages that declined seem more precisely relevant. <<
I see that in spadefuls on [126.96.36.199...] for some queries. In fact, some of the results look more like those you might get if you had hit the "related" button somewhere, except the function had been changed to be "vaguely related".
I don't have a ton to add, other then I noticed the changes today and yesterday as well on on area of my site. It seems in my case that areas where I had few backlinks but high quality, relevant content have been pushed down, and sites with tons of backlinks but more general and less specific content have been pushed up.
|BD has been an unmitigated disaster and Google has yet to produce a decent index out of it. |
For the keyphrases and keywords that I watch, Big Daddy's index looks about the same as its predecessors did. I used to see more day-to-day variations in the pre-BD days than I've observed since Big Daddy was rolled out.
Is it right to call any of Google's updates (or infrastructure changes) an unmitigated disaster?
After seeing Google's updates for a a few years, I am pretty much clear that quite a few problems with Google's index will remain always.
I would like to trust the more experienced members here - and if one goes by what they say, at any given point in time, there have been those who suffered but didnt deserve it and those who came back from hell and thankful for that. It is always good and bad.
What disappoints me is that at any given point of time, there are those who didn't deserve to be hit.
|Big Daddy's index looks about the same as its predecessors did |
Thatís been our experience, just havenít seen much difference at all (and I thought that was the plan?)
As to April 11th, a definite, albeit subtle change for many of our sites. Generally sites that are well established and have been around for a while, most down a few notches unfortunately!
Half of the "large" index is AUG-05 vintage - lots of 404's and lots of pages Google decided were supplemental because their own index was so old. That's not just me, many have noticed and commented on it.
The "small BD index" is extremely limited.
Isn't that a disaster?
And I am not talking about me or anyone else doing well, doing badly. Its the quality of the index and results.
bobmark is right on. The quality isn't the greatest for a lot of queries. I'm seeing a ton of 404s.
| This 32 message thread spans 2 pages: 32 (  2 ) > > |