| This 265 message thread spans 9 pages: < < 265 ( 1 2 3 4 5 6 7  9 ) > > || |
|Google Updates and SERP Changes - August 2013|
System: The following 16 messages were cut out of thread at: http://www.webmasterworld.com/google/4589243.htm [webmasterworld.com] by engine - 4:03 pm on Aug 1, 2013 (utc +1)
|There is certainly a disconnect/time lag between Google spidering site changes and those changes taking effect, and with the "rolling Panda" and Penguin (in which remote changes may impact your site) it really is getting impossible to get a clear idea of what is going on. |
The results are fluid but not as much as we would believe.
When updates are rolled out they change the value of pages/links. The change in values makes the current rollout instantly obsolete which will impact the next rollout, and the next. While the changes are current they impact page rank which sends ripples out that won't be accounted for until a future update.
The trick is to not add more flux to the ripples hitting your own site by making too many dramatic changes. Reduce spam pointing at your site, create great content and acquire solid backlinks where possible... but don't turn the puzzle upside down too often with dramatic changes or you keep starting over.
I am seeing strange things. I had been recovering slowly. I deal in a market where we sell ecomm (main part of the business) but we also at times buy the same type items that are older. Sales are okay but not back to normal but the people inquiring to sell to us is through the roof. It is several per hour which is much higher than normal.
The other strange thing is we have structured language. There are about 9K pages I would guess. I had about 2K being picked up then WMT dropped and said there was none. I didn't change a thing, then a few weeks later we had 3.5K picked up. Now I am back to zero. Like a yo-yo.
The following 11 messages were cut out to new thread by robert_charlton. New thread at: google/4604838.htm [webmasterworld.com]
1:12 pm on Aug 25, 2013 (PST -8)
Mod's note: I've split off the discussion of TBPR, which was developing into a good discussion of its own and was off topic wrt SERPs changes in any event.
New thread here...
The utility of Google Toolbar PageRank?
[edited by: Robert_Charlton at 9:23 pm (utc) on Aug 25, 2013]
I know we all like to jump on Google when traffic is off, but I still suspect some odd global network issues. I know of a few others who have observed the "spam volume" phenomenon - when sales / conversions are high, traffic level is normal but spam volume is high, when sales / conversions are low, traffic again seems normal, but spam volume is very low. Not sure if it's even worth mentioning, but it is highly repeatable. Today started off with very little spam, and no sales...last Sunday, the day started with a flood of spam and a ton of sales. Both days started with similar traffic levels.
[edited by: phranque at 6:46 am (utc) on Aug 26, 2013]
[edit reason] clean up system tracks [/edit]
|Martin Ice Web|
I donīt know what they did but i see very odd results with 5-6 times the same domain on page #1. Amazon is #1 again for every term, ebay #2, brands #3-#10, seems like a dial up for brands again.
Spam volume was exceptionally high yesterday evening through early morning. I usually see this when an update is in progress, but this was much higher than anything I usually see during an update.
After a record week last week, my search referrals from Google have been lower than usual since Saturday. I do see some search queries where I've dropped a few spots, but nothing severe.
This is a US-based informational site, but so far, it looks like most of the loss has been foreign traffic.
|I donīt know what they did but i see very odd results with 5-6 times the same domain on page #1. Amazon is #1 again for every term, ebay #2, brands #3-#10, seems like a dial up for brands again. |
Martin - one of my terms Amazon takes 1st - 8th! please tell me how thats offering ANYTHING for the surfer ...Bing on the other hand I'm 2nd and the top ten all relevant with no Amazon in sight
So I wonder if spam volume has any correlation to serp updates or traffic quality or patterns. It doesn't seem like it would, but this might reveal a network quirk affects traffic quality at certain times.
You know me, always looking for patterns in the chaos.
BTW - the email under observation is over 13 years old...so it's a pretty good antennae picking up a broad range of sites, both good and bad. I doubt you'd see this phenomenon on recent "young" email accounts.
Regardless, something substantial appears to have happened over the weekend.
|Martin Ice Web|
tigger, no extra use for the Users! I though g$$gle fixed this problem, but every time they change something, it seems like they dropped/dumped some other code for the new one. Mayby they are working with old x386 with only 512KB RAM for source code?
backdraft, bsterz, email spam? or referral spam or zombies?
Something definitely seems amiss today. Check these out, the 'Marijuana Possession Phoenix' has the first 17 results about the same story.
And the 'Phoenix <snip1> <snip2>' screenshot shows Google serving up a local branded result for an exact match domain for a partial match keyword search. Are these glitches or an update gone awry?
Marijuana Possession Phoenix: [i.imgur.com ]
'Phoenix <snip1> <snip2>': <screen shot link removed>
[edited by: Robert_Charlton at 9:04 am (utc) on Aug 27, 2013]
[edit reason] removed search that was IMO too commercial to discuss [/edit]
nsauser - Generally, we don't allow posts referencing specific searches, and I've removed your second search and screen shot because it's too specifically commercial in nature. But the first search is to a current hot news story, returning basically news sites, and I think we can make an exception for that one.
It's an interesting example, I think, of what might be considered the distorting influence of notoriety... be it fame, pop status, or brand, ... and how that notoriety is so strongly associated in current news with the query, which is...
[marijuana possession phoenix]
...that it's clearly affecting what Google is showing.
|the first 17 results about the same story |
My emphasis added in my quote.... While the results are about the same story, the several I looked at have all been substantially reworked by their particular newspapers. So, they're different, but the same topic, which clearly has a lot of interest.
The topic is that a Phoenix Sun's star basketball player, Michael Beasley, has gotten busted for pot for the third time, and he's been bad. Hot stuff. The stories are drawn from a wide spectrum of publications, playing on various angles of the interest in Beasley. Beasley, in effect, is the brand.
Essentially, the screen capture shows that the serps are being skewed by "buzz" to such an extent that Google is assuming, probably statistically, that any one of the top 17 news stories containing the vocabulary searched is likely to be more popular and relevant to searchers using that query than other sites would be (eg legal or informational sites) which might normally be brought up by the search.
The search query strongly associates with the Beasley story and the search, in such a way that, in Phoenix, the query is effectively a query for the Beasley name... in much the way that [iPhone] becomes a branded search for "Apple"... if that makes sense... or "raging fire" might currently become a search for "Yosemite". Eventually, after the 17 news stories on Beasley, Google does return a legal article.
I note that you can search other citynames in combination with the same search terms and get either legal/information sites, or celebrity cases, depending on the local climate and currency of the stories.
What gets interesting, IMO (and I don't want to get into the details of a specific site), is that an attorney's site that does appear among the Beasley stories actually has another story that discusses another guy named Beasley somewhere in the site. I have no idea whether that's affecting the choice of which attorney Google returned in the mix. Probably not. I think it's more likely that the attorney just had a good article matching two othe search terms in the query.
Anyway, to use very sloppy math... I think it's quite likely in the current news climate the 17 news stories on the topic were, in the aggregate, at least 17 times more popular for the query than a legal article would be... and that's the way I think Google is returned many of its results now, at least until Google sorts them out and gets enough data to further refine the serps statistically. The more long tail, the longer it will take to sort them.
So, I don't think it's "amiss". I do think it's interesting, and for me it describes how the serps are likely to continue to evolve. Hope I haven't taken this update thread off topic, but that's what I'm seein in this particular example.
The question of exact and partial match domains in the other search is a whole other topic. We generally don't discuss specific quereis in this forum, so we're not going to discuss those results. But to add a thought that should stand without the specifics... I think that just because a few sites come up containing query words in their domains doesn't mean that Google is favoring or not favoring EMDs or PMDs. It's much more complicated than that. Not sure that fully addresses the question.
|Martin Ice Web|
Once again: Since 09. August I see the same results like before this multiple week rollout? All the pages that gained some places are gone again?
Did they roll back this update?
I noticed that after this update, g$$gle did not ignore one or more keywords or searched for "similar" words. Results have been good in our niche. But since 09 this is all back to old behaviour. Very frustating.
Nothing much has changed for me. Just slowly losing traffic since April 2012. About another 5% this month. I don't have much left to loose anymore...
|Martin Ice Web|
@JM, this is scary.
Do you loose constantly old keys or other, jet not touched, keys take the way down ?
I see cicles of uppopping keys in a 4 weeks turn. They stay one to three days and then they are gone to make place for a new keyword.
Always to have the same amount of traffic a day.
Googles internet world gets smaller, itīs a very itelligent way of make the user used to have less choices. In one or two years google will make money only by showing 5 domains and poeple are lucky.
Plus, they donīt have to invest in new storage hardware. Nice new internet world.
|Nothing much has changed for me. Just slowly losing traffic since April 2012. |
Apart from a brief spike in the middle of last week, that about summarises it for me, too.
Down, down and down since February 2011...
About 87% loss of income from apprx $100K
1,000 hours trying to fix
Information site USA
|Down, down and down since February 2011... |
...which poses the question of where it is all going. Is it at all clear who, if anyone, has picked up your own referrals/traffic/business?
I still see quite a few sites (brands, but not Big Brands) that see pretty consistent increases in Google referrals. The increase are much lower percentage wise than they used to be, but I attribute this to the internet becoming much more like other traditional mediums.
However, newer sites and older sites that are just now being optimized seem to be able to make much larger gains than older sites that have been optimized for years.
Has anyone who tracks ecomm SERPS seen an increase in news stories appearing in what was previously exclusively ecommerce site mix? It is not happening to a great extent in my niche, but seems to be something that's increasing in prominence.
And, after having the best two weeks we've ever had earlier this month, it's starting to look a lot like a typical April or May, not an August where things are on the rise prior to the holiday season, as usually happens.
|Has anyone who tracks ecomm SERPS seen an increase in news stories appearing in what was previously exclusively ecommerce site mix? |
Yes. Today I noticed a similar phenomenon to what Nsauser reported above: a long way down the SERPS for my key term (on about page 30-something) there was a solid block of news stories - all different pages/sites, but all the same story - that had no relevence whatsoever.
It was roughly equivalent to "Hi-tec Widget" bringing up a block of stories about theft of hi-tec trainers from warehouse.
This looks like a new phenomenon, and although Nsauser's example might not have been "amiss" (and I am not so sure), in this particular case it was a contiguous block of about 8 results sticking out like a very sore thumb.
I guess what I should have added to my post was that the news stories showed up at some point in the last week as I do ranking reports for our company every Monday. I understand the concept of 'buzz' but it seems like 46 of the first 50 results being related to the same story from the first week of August hardly seems like buzz. But I get it. If in fact it's a fresh buzzy type of thing it seems like it's a bit glitchy, that's all. I've never seen a SERP like that without using a long-tail search.
I also saw a lot less domain diversity as well and a change in location based results.
I'm in the centroid of Phoenix and a search for a term including Mesa gave me first page results of approx 55% Mesa, AZ related and the rest were matched to Costa MESA, La MESA, etc. The typical distribution is typically 85 - 90% Mesa, AZ results.
@Wilburforce: This is different. These news stories appear on the first or second page. Usually only one or two at a time. So not many, but they end up taking a spot that might have otherwise been held by an ecomm site.
|Martin Ice Web|
|Is it at all clear who, if anyone, has picked up your own referrals/traffic/business? |
In our niche amazon did eat 90% of the lost traffic. But amazon began to get popular in germany two years ago. They got an extra traffic boost from google since then.
80% of friends i am talking to are going straight to amazon or one of the two big price compare engines. They use google to find this sites. Like typing in search box "amazon".
This now makes it extra silly because if google does use search metric, this isnīt realy a search because the users knows what he wants and is to lazy to write down the url.
Users get used to favorite sites. I think the overall search for for google begins to fade. The search statistic just shows the disrelevance of the current serps and that users arenīt sufficient with this.
All the ecommerce sites are down because Google loves serving up Amazon, Walmart and all the other big boys first.
@Dymero: yes, I see (although I am not sure a different phenomenon accounts for it).
I have since had another look. There are 5 two-day-old reports on the same story (all from different publications, and phrased differently) on the same page, in a position that would ordinarily be occupied by a commercial site.
The news stories that nsauser reports are almost three weeks old at this point, but clearly the story he points to has a lot of traction. I'm thinking it isn't so much a glitch as a new way of weighting results that Google is testing... and yes, it's a very weird result.
Dymero and Wilburforce - Are the stories that you're seeing in any way high profile? And, what's your default results per page setting when you look at these? IMO, using anything higher than 10 results per page can cloud the issue. If we're talking about results seen at different settings, that would definitely be clouding communications.
Trying 10 results per page vs 100 results per page, though, and comparing them to see what's different might indicate whether Google is using a new type of clustering. Traditionally, clustering has been per domain. Here, it might be per "perceived topic", for want of a better word right now.
|if anyone, has picked up your own referrals/traffic/business? |
I'd have to say big brands/names sites in my industry. Hardly a sign of smaller sites nowadays...all gone to page two and three, but doing well on Bing/Yahoo.
Local also taking a bigger share compared to three years ago on key phrases.
|Are the stories that you're seeing in any way high profile? |
Yes, very high profile. So I suppose it's possible that the story was big enough to break into an ecommerce SERP. And I have it on the default 10 results.
Strange search today. Website deals with lets say "widget training" we have a subdomain on the site that deals with "widget jobs" a search comes in from Bing "special needs widget training" and the main domain is ranked.
I enter the same in Google and the sub-domain is ranked. The sub has no text for "special needs widget training" it is all about "widget jobs" and has no relevancy to the search. The main website is on the 3rd page and the sub-domain "widget jobs" is on the 1st page.
Just not a good result showing at all, actually for the search on Google it is really poor quality search with the other sites showing on the 1st page.
Are the stories that you're seeing in any way high profile?
I wouldn't say so (and they are not high in the results, either: it is just the fact that they are in a cluster that drew my attention to them).
|And, what's your default results per page setting when you look at these? |
| This 265 message thread spans 9 pages: < < 265 ( 1 2 3 4 5 6 7  9 ) > > |