homepage Welcome to WebmasterWorld Guest from 54.211.235.255
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Home / Forums Index / Google / Google SEO News and Discussion
Forum Library, Charter, Moderators: Robert Charlton & aakk9999 & brotherhood of lan & goodroi

Google SEO News and Discussion Forum

This 172 message thread spans 6 pages: < < 172 ( 1 [2] 3 4 5 6 > >     
Google.com SERP Changes - July 2008 part 2
cj94111




msg:3698118
 4:53 pm on Jul 14, 2008 (gmt 0)

< continued from [webmasterworld.com...] >

I run a large, well-established website that has been around for years, generating millions of visitors per month across millions of pages of documents. We have always been diligent about tracking all of our website metrics so that we understand user behavior, where our audience is coming from, and can use the data in order to improve user experience.

Recently we have been experiencing *very* erratic Google organic traffic which jumps up and down by 30%-80%. The cycle has now repeated itself 6 times over about 6 weeks time. While there have been times in the past where our Google organic traffic has increased and decreased, it always has done it in a measured manner; we have never before seen erratic behavior from Google.

Here are the traffic specifics:
•June 3, Google organic drops by 30% vs. normal
•June 4, Google organic traffic returns to normal
•June 9, Google organic again drops by 30% vs. normal
•June 17, Google organic returns to normal
•June 19 , Google organic again drops by 30% vs. normal
•June 27, Google organic returns to normal
•July 9, Google organic again drops by 30% vs. normal
•July 11, Google organic returns to normal
•July 12, Google organic again drops, but this time by 80% of normal
•July 13, Google organic returns to normal

While we are constantly in the process of refining our site, the only major change over the last couple months has been to our “related articles” component which does what it sounds like: if you are looking at article A, here are a handful of other articles that are highly relevant to the one you are viewing. Over time, we have been tuning the algorithm that generates these links so as to improve relevancy.

I have also noticed some other artifacts:
•Google bot spidering activity has increased, reaching a plateau of about 140% of pre-link change levels; on some days approaching 1 million pages/day.
•The number of page indexed in our Google Webmaster site map reports jumped by 12%.

Any ideas about what might be going on here?

Thanks!

Greg

[edited by: tedster at 8:03 am (utc) on July 15, 2008]

 

BillyS




msg:3699400
 12:58 am on Jul 16, 2008 (gmt 0)

Establishing cause and effect is not as easy as saying "this happened after that." It's called the post hoc ergo propter hoc logical fallacy. That fallacy creates magical thinking and mythogloy - including SEO mythology.

I took logic in college, I loved that class. Everyday people break these rules. Cause and effect is not always what it seemst to be...

willybfriendly




msg:3699463
 2:38 am on Jul 16, 2008 (gmt 0)

the last couple of months have seen changes in the Google SERPs that defy explanation from any previous basis.

Well, a couple of obvious things...

1. Currently there is one hyphenated domain in my niche out of the top 50 for 'widget' and 'widgets'. That is a total of one hyphen in 100 results.

2. There is only one subdomain in the top 50 for both 'widget' and 'widgets' and that is a very old text file on a .edu domain (#6 for 'widget' and #7 for 'widgets'.)

3. Checking age of domain via the way back machine shows one domain less than 6 years old in the top 20 for both 'widget' and 'widgets'.

From my perspective, these factors have been getting increased weight in for the past 6 months. The hyphens and subdomains have filtered a lot of garbage, but there has been some collateral damage as well.

Weighting age (whether that might age of domain or age of links) has brought a lot of low quality sites to the top. Just because Auntie Em's Widget Shop has been on the Internet since 1998 does not make it a site that everyone will want to visit when looking for widgets.

tedster




msg:3699468
 2:48 am on Jul 16, 2008 (gmt 0)

The most perplexing new SERP observations are those that report cycling, sine waves, yo-yo, rollercoaster, or pick your favorite synonym. Sometimes these cycles happen down in the deep results pages after a url has dropped from page 1 - an apparent penalty. And sometimes the cycling appears on page one - from 3 to 10 to 3 to 10, day after day or week after week.

I don't have a site under my auspices that is showing this effect, but I've been asked to look at few that are - and so far, I can say that the phenomenon is real, but am mystified by it. I felt this way when the -950 first appeared back in 2006 or so, and slowly some understanding of that has emerged. Sure hope we can get some understanding about the yo-yo phenomenon, too.

CainIV




msg:3699477
 3:17 am on Jul 16, 2008 (gmt 0)

I would agree fully Tedster. I can see it myself not only by direct observation, but by tracking where a stronger two word phrase will rapid-fire the tracking for about 3 hours, then it's gone.

It seems very similar to DC flux, but may not be that at all. Could this possibly be testing of various groups of data that are ordered by various changes including manual input?

willybfriendly




msg:3699488
 3:51 am on Jul 16, 2008 (gmt 0)

The most perplexing new SERP observations are those that report cycling, sine waves, yo-yo, rollercoaster, or pick your favorite synonym.

As I noted earlier, I experienced this for a while, but things seem to have steadied out.

Given that I moved the site from a hyphenated to non-hyphenated (along with a complete site overhaul) domain (along with a complete site overhaul) back on 1/15/08, and was on the rollercoaster until May, I can offer little feedback on the yo-yo phenomenon. For I, and a colleague, the yo-yo revolved around position #6 and #11, or there abouts. The yo-yoing stopped with both of us down on the bottom of page two or top of page three. That was on the plural term 'widgets'.

tedster




msg:3699500
 4:23 am on Jul 16, 2008 (gmt 0)

These days, working to improve your rankings is a lot more like being a gardener than being a mechanic. You think you've got the ideal fertilizer, it always worked before, but now the leaves get brown spots. Has the fertilizer gone bad? Or maybe there's a tiny insect you didn't notice? Maybe the soil is now depleted of some trace mineral that the fertilizer did not ever supply. Maybe that new batch of seed was genetically compromised in some way.

And so on. Getting and manitaining good rankings is really about watching over the health of your website, indefinitely. There's often no such thing as "set it and forget it", no magic formula that is always guaranteed to work. Of course that's business as a whole, anyway.

That name we use - organic rankings - is proving to be more and more accurate.

sahm




msg:3699534
 5:33 am on Jul 16, 2008 (gmt 0)

A drop in ranking may not be "a penalty", but rather just a failure to do well with regard to some new measure or other. In other words, the url just doesn't rank as well as it did, and there may not be a penalty involved at all.

I completely agree. It's the only thing that makes sense and explains why some of my urls have dropped and others haven't.

MLHmptn




msg:3699539
 5:45 am on Jul 16, 2008 (gmt 0)

Anybody else noticing a difference in single word result sets with lowercase and uppercase being different? I have only noticed this within the last hour.

One result set in CAPS...9,490,000 results...

Same word in lowercase....4,780,000 results..

Different case is causing different shuffling of the results...Very Strange!

I should mention this is with a single word search as well.

[edited by: MLHmptn at 5:48 am (utc) on July 16, 2008]

RedCardinal




msg:3699638
 8:16 am on Jul 16, 2008 (gmt 0)

These days, working to improve your rankings is a lot more like being a gardener than being a mechanic. You think you've got the ideal fertilizer, it always worked before, but now the leaves get brown spots. Has the fertilizer gone bad? Or maybe there's a tiny insect you didn't notice? Maybe the soil is now depleted of some trace mineral that the fertilizer did not ever supply. Maybe that new batch of seed was genetically compromised in some way.

And so on. Getting and manitaining good rankings is really about watching over the health of your website, indefinitely. There's often no such thing as "set it and forget it", no magic formula that is always guaranteed to work. Of course that's business as a whole, anyway.


A nice analogy. Now perhaps consider this in the light of Google not wanting SEO to meddle in the ranking of websites. They've stated more than once that they see the future of SEO being about crawlability and site architecture. They dont like the results manipulation that occurs as a result of the inherently flawed link-based ranking algo.

Again, purely speculation on my part, but I have a feeling they are going to try to change 'SEO', and some of what we've been seeing recently would fit nicely with such a strategy.

StaceyJ




msg:3699734
 11:20 am on Jul 16, 2008 (gmt 0)

The most perplexing new SERP observations are those that report cycling, sine waves, yo-yo, rollercoaster, or pick your favorite synonym. Sometimes these cycles happen down in the deep results pages after a url has dropped from page 1 - an apparent penalty. And sometimes the cycling appears on page one - from 3 to 10 to 3 to 10, day after day or week after week.

I posted some of my observations a couple of pages back, so I won't waste your time posting an additional list, as things haven't changed.

Still watching a two word phrase that site used to rank #1 for - from July 12 through July 14, all from one data center/IP address, position has bounced from 34 to 110 with 4 different sets of returned results (1,130,000; 1,150,000; 1,160,000; 1,170,000). In all these results, the top 10 have pretty much stabilized and stayed the same all along.

On July 15 the results were from a different data center (and still are from that same one) and were basically the same as above through this morning.

edit - added further information.

[edited by: StaceyJ at 11:22 am (utc) on July 16, 2008]

Gissit




msg:3699808
 12:57 pm on Jul 16, 2008 (gmt 0)

Firstly the WMT conspiracy thoery - PMSL. What information is google getting from a site that is registered in WMT that it does not already have access too? Absolutely nothing that I can think of. It does not even gain knowledge of ownership as I have a number of sites in WMT that I do not own but just help out with and I'm sure plenty of others are the same. Foil Hat Syndrome me thinks, lol.

One of my own sites has had a rough ride since the middle of june with it comming and going in the serps on a roughly seven day cycle. It is a uk .com site on a UK server but I monitor it for US SERPS as well through a proxy so I have posted here as it is affected in all results.

I have spent a good ammount of time looking at what is happening and I think, in my case at least, it may be down to so algo tweaks that are not as obvious as previous changes. It seems at the moment that google has changed the way it looks at single words and plurals used in phrases. It seems that individual words used in phrases no longer carry any weight for the individual word but only for the phrase and the individual words in a phrase are not stemmed to include the plural (although the whole phrase may be stemmed to plural), but words used singularly are still stemmed to the plural. Many words and phrases that used to rank are gone but searching very similar things with slight changes such as plural and word order still put the site high in SERPS. I think I'll sit it out for a bit before making any changes though as it does not seem to have settled just yet.

c41lum




msg:3699837
 1:54 pm on Jul 16, 2008 (gmt 0)

Hi Gisset,

I am seeing the same as you. My site is a UK site and it is running on a 10 day cycle, in and out. Crazy.

What I am noticing is that I am still getting crawled by G every few hours but my cashed page is older than 30 days. When I go up in the SERPS is when G adds a new cached copy. Very strange. I think they have two different algos running in the UK at the moment.

It also seems that its only my main pages (Catagory Type) that are being hit hard. My product pages are still holding there own.

cj94111




msg:3699983
 4:31 pm on Jul 16, 2008 (gmt 0)

Well, here we go again. After 3 days of great organic traffic, we are again in the penalty box with traffic drop off of 90% since 12:01 AM. So here is the full sequence:

•June 3, Google organic drops by 30% vs. normal
•June 4, Google organic traffic returns to normal
•June 9, Google organic again drops by 30% vs. normal
•June 17, Google organic returns to normal
•June 19 , Google organic again drops by 30% vs. normal
•June 27, Google organic returns to normal
•July 9, Google organic again drops by 30% vs. normal
•July 11, Google organic returns to normal
•July 12, Google organic again drops, but this time by 80% of normal
•July 13, Google organic returns to normal
•July 16, Google organic again drops, but this time by 80% of normal

That is 6 cycles in 6 weeks for a site that has been around in some form or another since 1997 and who has never used any black hat techniques.

I have 2 questions for the group:

1) One thing that happened that may have set off Google alarm bells is that one of our partners appeared to have been linking to us on millions of pages (they were supposed to put a link to us in their partner section as we include their content as components on our pages).

Google picked this up and our external link count went from about a half-million to three and a half million and then our partner fixed it and our link count went back down to a half million. That happened some weeks ago. We only discovered this when we saw our external links report in Google Webmaster console. Since that system will only output 1 million records (and we had 3.5 million), we weren't able to fully confirm that all those links were coming from our partner, but it sure seemed like it from the million we were able to review.

2) Someone had suggested that we look at keyword stuffing. Of course we don't do that, but I am wondering whether some of our cross-linking techniques may look like it to a machine. Here is what I mean: we have millions of documents on our system which is essentially designed as a research tool. When someone comes in to look at an article, it is generally because they are looking for pretty specific content.

Given that they are in research mode, we like to provide related categories and related articles that would be highy relevant to the one that they are looking at. Over time we have invested a lot of engineering in improving these "related content" algorithms, but what happens is the more we improve the algorithm, the more we surface articles that have very similar headlines and descriptions (which, after all the point).

But that in turn could look like a lot of the same keywords repeated numerous times.
Thanks for any feedback!

Greg

[edited by: tedster at 6:42 pm (utc) on July 16, 2008]
[edit reason] removed specifics [/edit]

Gissit




msg:3700113
 7:01 pm on Jul 16, 2008 (gmt 0)

c41lum
If you change the order and plurality (is that a real word?) of words in your target search phrases does it make a lot of difference? eg if "red widgets" is your phrase do you rank better for "widgets red" or "widget reds" or "reds widget"?
Just something I am looking at and some extra feedback would help..

SEOPTI




msg:3700165
 7:50 pm on Jul 16, 2008 (gmt 0)

cj94111, they try to force everyone into adwords by randomizing their SERPs.

signor_john




msg:3700170
 7:55 pm on Jul 16, 2008 (gmt 0)

Warning: Red herrings and sour grapes can cause indigestion. :-)

cj94111




msg:3700188
 8:25 pm on Jul 16, 2008 (gmt 0)

SEOPTI-

If that was true I think we would have seen this kind of behavior at least a couple times in the last several years. As it is, this is the first time it has ever happened. Also, we are already a pretty large AdWords user.

If they randomize their SERPs to any degree, I would think they would be doing it in a less overt fashion (not a 90% decrease in traffic) and they would be doing it to make sure that you can't game their system.

All the above is conjecture of course as they don't come out and tell you.

Thanks for the feedback!

Greg

iInventedtheinternet




msg:3700196
 8:34 pm on Jul 16, 2008 (gmt 0)

I just want to chime in on the whole controlling SERPS to force webmasters to use AdWords.

There is no evidence to support this theory and in fact plenty of evidence to the contrary. AS cj94111 mentioned he is a heavy AdWords user already and his site is still getting jacked around. I too was a heavy AdWords user and have recently had these same 90% fluctuations.

My biggest problem with this recent spate of trouble is that Google is acting borderline irresponsible with these SERP moves. We're not talking about small fluctuations in traffic but rather we are mostly talking about large sites losing and gaining 90% traffic that sometimes reaches tens of thousands of visitors a day.

The infrastructure needed to support this type of traffic isn't simple and it isn't cheap so by shutting it on and off Google is really putting these webmasters in a really tough spot.

tedster




msg:3700221
 9:01 pm on Jul 16, 2008 (gmt 0)

Well said. We still don't know why some sites get the roller-coaster treatment and not others, but I agree that this is a pretty lousy way to run a search engine. Google likes to talk about the "signals" that they get from websites, but right now their own signal sent TO at least some responsible websites is as noisy as I've ever seen.

We can hope it's a mid-summer experiment and they get a handle on it soon. If I want a roller coaster, I'll go to an amusement park. If this behavior sticks, then we've got a lot more detective work to do.

cj94111




msg:3700266
 9:42 pm on Jul 16, 2008 (gmt 0)

In my case we are talking hundreds of thousands of visitors each day.

Excuse me while I run out for another antacid...

tedster




msg:3700337
 11:29 pm on Jul 16, 2008 (gmt 0)

-- THREE TYPES OF CYCLING? --
I've been giving this yo-yo thing some more thought, and we may have three different situations -- at least three.

1. Ranking falls off page 1 into the deep pages, and then cycles down there, 657-430-250-790-etc. But it never goes back to the top. So this cycling follows immediately after what looks like a penalty.

2. Ranking begins to cycle on and off page 1. This can cause dramatic shifts in traffic - up and down 90% as cj94111 is describing. It seems to hit some well-established domains with years of solid history. It's not clear why this starts to happen, and the continuing reappearance of traffic seems to say it's not a penalty, at least nothing like we've ever known.

3. Ranking cyles up and down, but always on page 1 - as in the 10-3-10-3 report above. This does not seem likely to be a penalty either, but it sure is strange compared to traditional SERPs.

-- IDEAS FOR EACH TYPE OF CYCLING --
Some first steps I'd suggest in these situations, if they persist for any significant amount of time.:

1. You've got a penalty. Get cooking to figure out what is wrong. No, please don't get stuck in complaint mode, just get to work and figure it out. The cycling that happens down in the depths is an interesting new wrinkle, and may well be generating some kind of statistical metirc for Google - but understanding that will not remove a penalty.

2. Fine tune your metrics. Understand which query terms are most affected, and whether there are some that are not hit. When rankings drop for any given query, how far do they drop? Is the change geo-targeted? Dig, dig, dig into your data.

3. This may be Google's initiative to offer "diversity" at the top of the SERPs. The best thing I can think of is to improve your site so that Google wouldn't want to experiment by depressing it because the quality is so clearly there.

Depending on the specific query, this kind of front page oscillation may also be a test of user intention. Does your page answer every kind of user intention represented by the query, or just one? Google is working very hard not just to answer the query people types, but what they might mean. There's a new article today on the official Google Blog [googleblog.blogspot.com] that goes into some detail about what Google is trying to do here.

-- JULY 2 CHANGES --
As a further comment, there is a growing feeling among some webmasters in highly competitive markets that Google rolled out an algorithmic attempt to sniff out link BUYERS. This seems to have been roughly around July 2. At that point in time paid links that didn't seem problematic before all of a sudden seemed to be causing trouble.

As I said, this is conjecture but it is supported by a lot of individual reports. How might such an algo work? Well, it could start with seeds of proven link BUYERS, not just sellers. If there are links to that BUYER from some other domain (in addition to a known link seller) then that domain might also be selling links, but under the radar. So all of it's outbound links would then come under scrutiny.

This kind of an approach might cause a bit of collateral damage, but it might well snag a lot of link buyers who have been clever enough to avoid detection so far.

That mechanism is also guesswork right now. In fact, there's no official statement from Google on the July 2 changes, except some vague hints that they might not be evaluating some sites' backlinks the same way as they used to.

cj94111




msg:3700386
 12:34 am on Jul 17, 2008 (gmt 0)

Great post tedster, very helpful.

Actually, we have been digging like crazy since this started 6 weeks ago. The list of things we have ruled out is fairly long at this point. Luckily we put in place very robust reporting systems so we can really see what is going on under the covers. We have looked at traffic by content types (we have many), where the traffic is originating, high-traffic keywords, low-traffic keywords, page distribution on days when traffic was up vs when it was down. Spider behavior, i.e. how many page requests, what pages are they requesting, how do those trends look from a historical perspective.

So far we have yet to find any smoking gun. In fact everything looks insanely normal, only the scale changes. But I will certainly post back to the group if we run across something interesting.

One question I have is that when sites are in the "penalty box" does Google typically stop spidering or reduce their spidering of the site? In our case, spidering in the last 6 weeks has hit unprecedented high levels of about 140% of normal. Is this normal?

Thanks!
Greg

tedster




msg:3700395
 12:49 am on Jul 17, 2008 (gmt 0)

spidering in the last 6 weeks has hit unprecedented high levels of about 140% of normal. Is this normal?

That's been a common report for sites caught up in whatever this #2 phenomenon is. Some penalties, especially for severe loss of trust, definitely do not show increased spidering at all. So the increase in spidering certainly may be another clue here. We'll see how many other sites notice the same thing.

If we're seeing an experimental stage roll-out, then increased spidering might make some sense.

ken_b




msg:3700414
 1:07 am on Jul 17, 2008 (gmt 0)


3. Ranking cyles up and down, but always on page 1 - as in the 10-3-10-3 report above. This does not seem likely to be a penalty either, but it sure is strange compared to traditional SERPs.

Ideas....
3. Depending on the specific query, this kind of front page oscillation may also be a test of user intention. Does your page answer every kind of user intention represented by the query, or just one? Google is working very hard not just to answer the query people types, but what they might mean. There's a new article today on the official Google Blog that goes into some detail about what Google is trying to do here.

This pretty much sums up my situation, and my assumption of the reasons. This has been going on with my site for a few months.

Most of my pages fall into a category that makes it hard to really justify the top 1-3 ranking they've held for a few years.

The "intention" part fits great with my situation. As my ranking has moved up and down the sites that have been moving around with me could be argued to have just as much, or just as little, claim to top ranking as me.

My site is an widget info site. The others are widget parts sites or widget dealer sites.

So why the hunt for the "intent" of the surfer? Poorly structured search queries.

Instead of searching for old widget info, old widget parts, or old widgets for sale, most of the searches have traditionally been simply for "old widget" (no quotes).

What to do?

Teach the surfing public about the long(er) tail?

potentialgeek




msg:3700429
 1:56 am on Jul 17, 2008 (gmt 0)

cj94111/Greg wrote:

> Recently we have been experiencing *very* erratic Google organic traffic which jumps up and down by 30%-80%. The cycle has now repeated itself 6 times over about 6 weeks time.

. . .

> While we are constantly in the process of refining our site, the only major change over the last couple months has been to our [links to] related articles component which does what it sounds like: if you are looking at article A, here are a handful of other articles that are highly relevant to the one you are viewing.

. . .

> Any ideas about what might be going on here?

I made similar changes that you did and got penalized. I was trying to be helpful to visitors so they could more easily find related content. However, it's very easy for Google to misinterpret this as spam, because it can be very difficult to add links to related content which don't have the same keywords in the anchor text.

You didn't mention anchor text but I'd suggest reviewing it, especially when it's at the end of the page (logical place). Big and well-established sites can suffer less extreme versions of major penalties (which makes it harder to decipher). Mine was the 950 Penalty on small sites.

One way round it is to use an icon (like a bullet) with the link but no alt tag. Then the text beside it with no link. That makes the link "neutral," and you could have as many links to related content as you want with no penalty for Anchor Text Spam.

p/g

minnapple




msg:3700449
 2:50 am on Jul 17, 2008 (gmt 0)

I have noticed what might be a change in the relational [~] term weight, along with the balance of inbound anchor variance.

cj94111




msg:3700450
 2:56 am on Jul 17, 2008 (gmt 0)

potentialgeek-

Thanks for the response. I may have failed to note that we have always had related articles, we just have been working on improving their quality. We also increased the quantity of links, though since we were worried about that we also removed some other components on the page that had links so the net increase was 5 links.

The articles are listed as headline with a brief description in the form of the first 2 lines of the article. Since so many of the article titles are brief or not very descriptive, we include this description which really helps readability.

As for the anchor text, it is the the same exact text as the link itself.

Thanks again!
Greg

trinorthlighting




msg:3700452
 3:01 am on Jul 17, 2008 (gmt 0)

Has anyone looked at external links in their google webmaster panels?

I noticed that our sites have not updated the "last found" for any date past July 4th. Typically, ours shows updates at least twice a week.

tedster




msg:3700485
 4:20 am on Jul 17, 2008 (gmt 0)

minnapple: I have noticed what might be a change in the relational [~] term weight, along with the balance of inbound anchor variance.

In todays' Google Blog post from Amit Singhal [googleblog.blogspot.com], we read:

Another technology we use in our ranking system is concept identification. Identifying critical concepts in the query allows us to return much more relevant results.

Sounds like there could be a tie-in there.

cj94111




msg:3700495
 4:41 am on Jul 17, 2008 (gmt 0)

trinorthlighting-

Quick glance at my external links confirms nothing past July 4.

Best-

Greg

cj94111




msg:3700496
 4:44 am on Jul 17, 2008 (gmt 0)

One other detail that may be worth noting...

Whenever our traffic returns from Google, it is at a *higher* level than before. Mind you, not enough to compensate for the huge plunge, but 10%-15% higher that pre-plunge levels. In fact prior to this latest plunge we had our 2 best Google organic traffic days in the history of the site.

--Greg

This 172 message thread spans 6 pages: < < 172 ( 1 [2] 3 4 5 6 > >
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google SEO News and Discussion
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved