| 5:28 am on Dec 15, 2010 (gmt 0)|
It's an odd symptom that we see ranking URLs that are duplicates, auto-generated spam, don't resolve, contain malware, etc. The continued SERP shuffles during holiday buying also are something we haven't seen much of in past years.
Here's a thought experiment (it's NOT established fact.) Imagine that for some time the complete algo has included a degree of automated self-adjustment. Then imagine that there are certain key personnel who really know and monitor this self-monitoring, automated adjustment facet. Further imagine that these key personnel are temporarily on another assignment and the self-tweaking modules are a bit out-of-whack right now.
Might that describe some of the inscrutable observations in recent days? Some kind of an uncorrected feedback loop gone unstable?
| 5:47 am on Dec 15, 2010 (gmt 0)|
The number one site for business (filler words for fluff) directory has sitewides across the bottom that are dubious. What does that say about the state of the algo when that site outranks manta, jigsaw, linkedin etc? What does that say to would be link sellers and spammers?
Searching for kids stuff tonight I see that google will display the same product offered by the same merchant on every site. In this specific case vendor xyz was selling product abc on amazon, sears and five other sites using the same text on all sites. They all rank between 2-20 displacing unique sites. It adds no value, is duplicate content and "thin/affiliates" but they seem unable to filter.
Any search with numbers is a problem. Five digit part numbers are a mess. Brand plus the part number is a mess if the brand has any geographical base. Think "Miami green unicycle.". Even if that were a huge product we all knew we would get a lot of results for unicycles in Miami, Miami motorcycle shops but not exact matches.
Specific matches are dead. I could sell "unique product tango whisky bravo 54321" and target that with the page name etc. I may be the only person licensed to sell this super popular product and could have backlinks from al gore himself but I won't be first if people searched for that term. Someone selling tango whisky, the bravo channel and other stuff would come first in what appears to be an overaggressive filter.
The results just keep getting worse and I do wonder is it deliberate manipulation by outside sources (to harm our economy) or is it as it appears, linked to and related to caffeine and mayday. Is this a case of more is not better?
| 6:19 am on Dec 15, 2010 (gmt 0)|
A niche I follow pretty closely has had a couple of new sites jump into the first page results. As best I can determine, one did a recip program, then dumped their recip page making all their links one way. Kind of old time, but it evidently provides results in the current environment. Other site seems to be using the forum profile approach. That'll probably work for some time period.
If we look at the niche, it could be similar to adult incontinence products. The niche isn't medical related, but it's not really likely to get FaceBook mentions or Twitter follows. It's just one of those things where it's not going to fit into the social network graphs. In a practical sense, who's going to talk about adult diapers on a social network?
I certainly don't believe Google is broken, I think it's providing results based on user experience. So perhaps it's not "Google" per se, but results people see based on their search history/location.
Your "vacation, or other assignment" theory is possible, but given the amount of information/pages churning on the internet, it doesn't seem probable Google as a corporation would say "Cool, take a couple of weeks/months off".
Search is the core, I can't really envision them ignoring that.
| 7:19 am on Dec 15, 2010 (gmt 0)|
Tedster we were posting at the same time.
I've never seen such poor results. These bad results aren't being snuffed out over a period of days as is usual but are instead manifesting into more bad results. The only ones I see ranking high are 100% of the time buying links. This forces us all into a corner, cheat or die.
I believe either you are right or my theory has merit. The problems are so bad it either has to be deliberate manipulation or something that has gone off the rails. I just cannot believe a product that twelve months ago was a miracle of science and that today is the Internet version of the delorean is an intentional course of action. This is not what they want, this is not what we want.
The results today are not what google wants. Is it possible they grossly miscalculated what would happen when their bot absorbed so much material?
Think about it wasn't there talk that all of the new upgrades would allow them to index more? Instead they indexed less. Look at this site as an example. I think google had a vast amount of good info pre feb 2010. As they ingested all this new data post caffeine a lot of good old pages got dumped in favor of scrapers and junk. If you take 20% of this sites pages and replace them with junk and this happens on millions of sites the web is a lesser place. That is what happened here.
When caffeine first launched new stuff was indexed in minutes. By last month there were sites not being indexed at all. There was the blog outage and there have been other similar unreported problems. There are weird messages on 3 year old pages in wmt.
I think they underestimated the fact that there was good reason a lot of stuff was not indexed prior to 2010, it was bad content. The spammers are winning and they cannot adjust fast enough. In the meantime legitimate sites are the roadkill.
| 12:05 pm on Dec 15, 2010 (gmt 0)|
In our sector I'm seeing nothing more than I saw last year. I'm assuming that after the new year we all return to normal. We're just not going to bother so much on Christmas on Google next year. There you go :-)
| 12:35 pm on Dec 15, 2010 (gmt 0)|
|Then imagine that there are certain key personnel who really know and monitor this self-monitoring, automated adjustment facet. |
"Pay no attention to that man behind the curtain....the great and powerful Oz (Google) has spoken!" - lol
I'd hope with Google's infinitely deep pockets that they could keep a full time staff operating the big Google "Busy Box".
| 2:47 pm on Dec 15, 2010 (gmt 0)|
|I'd hope with Google's infinitely deep pockets that they could keep a full time staff operating the big Google "Busy Box". |
And Coke should keep full time staff keep re-reading the recipe?
No, a few people know the secret sauce, have access to the data, can understand it, and can manipulate it.
If the pool of "insiders" was bigger, then Cuil would have worked.
We're doing well relative to last year, so this part of the ecomomy isn't currently being wrecked.
| 3:28 pm on Dec 15, 2010 (gmt 0)|
Lots of foreign traffic so the daily ten am algo update must be rolling through. It should stabilize for a few hours before the 2pm update rolls in that's applied to some sites some of the time depending on sunspot activity and the moons orbit.
Even wiki is taking a hit on my searches. A search for a broader match to the earlier example returns not work friendly results in the 1 and 2 spot. Synonyms again where business means two different things.
The post holiday SERPS aren't going to be any better IMO.
| 4:51 pm on Dec 15, 2010 (gmt 0)|
Coke don't change the recipe twice a day ;)
| 5:06 pm on Dec 15, 2010 (gmt 0)|
|And Coke should keep full time staff keep re-reading the recipe? |
I'd say that analogy is like comparing a Tinkertoy to the Space Shuttle.
| 5:40 pm on Dec 15, 2010 (gmt 0)|
Looks to me like something in the serps changed overnight from the 14th-15th, and pushed the "Shopping Results" 1-2 positions higher.
Up through last night, (and for several months) we had a 2nd organic placement, just above the shopping results.
Used to see:
- Other Competitor
More Organic Results
- other competitors...
When I woke this morning, I see:
- Other Competitor
It was definitely an overnight change in our market (things people wear on their feet)
Its a ghost-town on the site.... where Google referrals are concerned.
(edited for spelling an to add a Google reference in the last sentence)
| 7:46 pm on Dec 15, 2010 (gmt 0)|
O boy, on the right hand side for adwords ads, when i searched "brand widgets" I'm seeing "related widgets" for ads. This is BS. If I wanted to target that keyword I would have chosen to do so.
| 9:33 pm on Dec 15, 2010 (gmt 0)|
my rankings climbed back a bit (after tanking for almost a month) since this Monday, although not as good as before it all started.
Has anybody else observed this?
| 10:03 pm on Dec 15, 2010 (gmt 0)|
Starting from past monday all of my posts are ranking for the exact keywords. It looks like the long tail is gone or at least it has only been reserved for well established websites and blogs. If I make a google search for the exact keywords my post will rank first but that's all of it.
Is anyone experiencing this behaviour?
By the way. Older posts are not affected by this meaning tha their long tails have been preserved...
| 2:07 am on Dec 16, 2010 (gmt 0)|
Sites that are high pr selling text links now end up ranking for those text words vs the target. Funny in a way!
Frank google seems to have adopted the what have you done for me lately approach. Rankings are relative to the day. I firmly believe if we could add the same exact page each week we would find it would rank differently every time based on up to date data.
Twitters role is comedic. We asked a writer that does our technical editing to tweet w product/retweet. He was an early tweeter and has thousands of followers. We are now first after only five hours from 14th. That is ridiculous and if this is part of what is going on I see why things are a mess.
Pay per tweet. Be my friend for $5.
| 3:50 am on Dec 16, 2010 (gmt 0)|
|if this is part of what is going on I see why things are a mess. |
The ranking effect of Twitter and real-time-activity does seem to be way out of proportion at this time. I do think a large degree of the volatility we see is hooked up to that factor. At this point I've seen many examples of amazing rankings flux tied to social activity. The SERPs are not exactly information retrieval in the classical sense any more.
| 5:22 am on Dec 16, 2010 (gmt 0)|
Its almost time to break out SERPS as informational v commerce.
I am seeing the biggest shifts tonight I have ever seen. Its now clear to
me a major change is rolling out. They may claim it effects only some sites, but something big is in the wild. IF the hints I see are the final results we will be happier people soon.
| 6:27 am on Dec 16, 2010 (gmt 0)|
|Its almost time to break out SERPS as informational v commerce. |
This is where I think my site is taking a hit, as it's an informational site, but also has an ecommerce section. It may be confusing to Google. That wasn't a problem in years past, but in looking at the results for searches on "Acme widgets" in my sector, it's the commerce sites that are ranking. Search for "Acme widget specifications" and informational sites dominate.
I'm now ranking 5-7 for some phrases that I held the #1 or #2 position on for years. I don't think any changes Google will make now will get those rankings back for me.
| 6:39 am on Dec 16, 2010 (gmt 0)|
We lost our rankings for our main keyword which was our main source of traffic. Just checked yesterday we were ranking on 85th position as compared to 6th. But the url that is ranking is an internal url not the home page.Majority of traffic was coming through homepage. Now is seems it has been completely eradicated. I checked if the page was banned but everything seems to be fine. Also the traffic from the related keyword has been completely eradicated.
One more thing we noticed is that new content has not been cached or indexed for few days. It's kind of peculiur as earlier it used to be cached & indexed in just 15 min. Sites that have started ranking in our place do not have enough content or even proper meta descriptions.
Any suggestions will be of great help.
| 8:05 am on Dec 16, 2010 (gmt 0)|
How can a nofollow twitter link pass juice, I really don't get it. It seems Google has some sort of access to twitter intranet or API.
Just image negative or positive votes with social bookmarks, this could also be a ranking factor, but quite noisy. I can hire people at mechanical turk and get hundreds positive bookmarks, so I think they should not use this signal.
| 8:31 am on Dec 16, 2010 (gmt 0)|
Our one site that died a death on the 2nd came back in a big way yesterday. Traffic up 60%. Something started on the 13th for us...
The other site has gone the other way & is full of junk foreign traffic.
| 9:01 am on Dec 16, 2010 (gmt 0)|
SEOPTI - Google and Bing both recently confirmed that they use Tweets as a ranking factor above and beyond their value as a link.
| 1:01 pm on Dec 16, 2010 (gmt 0)|
What surprised me about the twitter usage is the recent report on demographics and usage. I'm not sure twitter merits the type of bump it's giving the SERPS. If anything Facebook seems far more relevant but I think the importance of both as a signal is overblown. A lot of the businesses we deal with block both at work and on work computers....so why exactly again should I be punished if borat hires his relatives to tweet their site thousands of times a day? This is going to be a bigger source of spam soon than anything theyve ever seen prior.
I really don't get it yet.
| 1:13 pm on Dec 16, 2010 (gmt 0)|
How can a nofollow twitter link pass juice, I really don't get it.
I seem to recall reading somewhere in the past week that the firehose of data coming from twitter is raw and natural. The nofollow and any other html type markup is added real-time. So the only way a link would be nofollowed was if you added the code to your tweet.
From Danny @ Searchengineland [searchengineland.com]
|In that firehose, links do not carry nofollow attributes |
Re: Abuse... My guess is that Goog & Bing use the data from Twitter or other social nets as a very temporary ranking factor, that fades off over the course of time, maybe even as quickly as a day or even a few hours, when the social hype fades away. Just a guess though...
| 1:19 pm on Dec 16, 2010 (gmt 0)|
At least with domain names there is a cost to spam. With twitter you could create 10,000 accounts and setup a network to spam boost. Nobody checks anything. They might as well use our junk mail folders as a ranking factor.
The data shows most use twitter a few times and never use it again. How does that gel with the average google user? It doesn't!
| 5:34 pm on Dec 16, 2010 (gmt 0)|
links : blogs : twitter : youtube: facebook
All have a purpose, but as soon as spammers realize search engines assign some weight to these elements, you have an proliferation of junk.
I look at a couple of my e-commerce competitors: blogs that say very little of value, facebook accounts for their business that say even less, tweets that are just worthless noise, videos that nobody would want to watch. Blah-blah-blah just to look popular.
Some SEOs swear by this stuff, but I don't see these guys ranking any higher than they did prior. While they are busy posting garbage on "MyTwitFace" I'm busy improving my site and adding new products.
I can see a SE placing some value on these things for very topical subjects, like a sporting event or a natural disaster. But a business? Lets hope the buzz only gets counted for a very short time (like 24 hours at most).
| 5:49 pm on Dec 16, 2010 (gmt 0)|
Sometimes you just gotta laugh... its easier and more manly than crying.
Today, all our Google search referred traffic is coming from french visitors. Google.fr
In addition, we dropped 5-7 positions in the US engine on a target phrase, and the only changes made during the holiday season, was adding fresh products.
Hmmm... Algo shift maybe? Gotta love gray area.
| 7:46 pm on Dec 16, 2010 (gmt 0)|
We got outranked today on one of our big screw products by a popular adult toy site. Tell me that's not relevancy gone crazy!
We should all just shut down our sites and list products on wiki or amazon case closed.
We too are seeing a big difference with foreign rankings being great domestic terrible.
| 4:00 am on Dec 17, 2010 (gmt 0)|
Sometime midafternoon the converting international traffic diminished and was replaced by domestic conversions. At about the same time the screw SERPS went back to normal. We sell industrial components, some for aerospace. Parts that we sell as screws and those that were ranking...use your imagination!
| 4:39 am on Dec 17, 2010 (gmt 0)|
Making conversions every 20 minutes today, right up to about 2:30pm CST, then site traffic stopped for the rest of the day. Today we also lost a #1 that we have held for 10 years to an MFA "link building" site (they say it right on the site). I took someones advice and filed a feedback report against it the other day...the result? they pushed it two positions higher to #1, replacing us. It's insanity.
| 4:42 pm on Dec 17, 2010 (gmt 0)|
Had a client call me in a panic that he was gone.
I checked the serps from proxy from several computers and he is still there.
I then had him sign out of Google, remove his cookies and then block google from being able to cookie his browser and run the search again.
He was right where he was supposed to be.
Then had him turn cookies back on, search again and he was gone again.
That is one of the reasons I don't allow Google or doubleclick to cookie my browser.
Any others see this type of activity?