Google Updates and SERP Changes - September 2012

[backdraft7]

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

Historically we never saw drastic drops due to holidays...except this year, when the now daily everflux drops just happen to coincide with a holiday.

[edited by: tedster at 5:33 am (utc) on Sep 1, 2012]

[ohno]

and if suggestions are localized, too.

That's an interesting one which may explain why we see sales of one product to the same area at times. Maybe a total coincidence but could also be explained by machine learning of a search phrase locally. Does not explain why that would come & go though..that may be down to traffic rotation/throttling.

[Martin Ice Web]

@ohno,

i see the same, IP ranges are "crowded" then nothing then "loose" IP range -> crowded.....

After 3 good days, it is thursday and changing day for google. All rankings lost for our many information site.
But our affiliat site with no "EXTRA" information and copied description from the manufacturer ( and we are not the only one with this description ) is doing very well again.
edit: the main site gets only image search and foreign traffic.

Because of the ongoing development in real life the internet is changing at a daily basis. If it is machine learning - and this machine learns realy bad - then it will be behind this development for ever.

[tedster]

I remember last year - or even earlier - someone from Google mentioning to the press they were considering an evolutionary change. This was something that might take years to accomplish, and that during the development they might have to sacrifice some of the quality they had developed. But they also said that in the longer arc of things they thought it was the best direction. Imagine that - thinking beyond quarter-to-quarter!

I don't remember if that comment came from Matt or Amit or even someone higher up. However, I took notice of the storm warning at the time and now it seems the storm is on us. The overall algorithm is doing things that we can barely describe, to say nothing of reverse-engineering. And updates are so frequent that we can't even identify the date of minor ones.

[tedster]

IP ranges are "crowded" then nothing then "loose" IP range

Martin, thanks for sharing that pattern. I've quoted you in the Zombie Traffic analysis thread.

Are other members seeing a similar shifting pattern of traffic sources? If so, how long does each phase seem to last?

[IanTurner]

Nice spot xcoder, tedsters post also makes good reading with the use of quotes and plus signs (something that no 'average searcher' will do)

My experiences have been similar to tedsters, but I don't do any large corporation work, just small medium business and our own sites.

What I have seen is most sites steady or growing slightly and just a few suffering problems.

From my position all our informational sites have been okay and growing - unless they were hit by penguin and one other very specific SERP layout change -> but I know the reasons for each one and can address the problem.

Small business sites have generally remained steady or grown when they have been targetting a local market and are in service industries.

Our ecommerce sites and ecommerce clients have been suffering unless they are 'extremely niche markets' with all down significantly over the last couple of years.

The interesting thing is that these small ecommerce sites have generally been reliant on long tail search terms for their traffic, they are never going to compete with 'corporate' sized businesses on major keywords (which are also those with Adwords dominated SERPs)

Changes affecting long tail and how it is handled by Google are therefore going to have a much more significant impact on small businesses - Google diverting traffic from some long tail terms to others could have a catastrophic effect for some because as what will happen is that bigger business will have more capactiy to target suggested phrases and these will then become more competitive further driving out the small independent sites.

The conclusion is that if Google is reducing the number of queries that it returns distinct results for in the long tail, then this policy is harming small/independent ecommerce sites disproportionately when compared with larger businesses.

[ohno]

IP ranges are "crowded" then nothing then "loose" IP range


Martin, thanks for sharing that pattern. I've quoted you in the Zombie Traffic analysis thread.

Are other members seeing a similar shifting pattern of traffic sources? If so, how long does each phase seem to last?

My this do you mean the IP ranges are close together? I've certainly seen that for converting traffic, so much so that I have wondered if the traffic source was a forum post for an ISP but that didn't hold any water, it just seemed that an IP range for a certain ISP was "switched" on to our results for a short period of time. Like I say, maybe total coincidence but it makes you wonder. (Geo targetting in the UK by IP would be impossible anyway, my IP claims I'm over 50 miles away from where I am right now, the UK isn't big!).

[Martin Ice Web]

I donīt know if this would not be a to short sight or even a shoot in the foot. Because more and more small webmasters quit working on their pages and there will be more and more webmasters. In near future the work disburse only by putting large information in short time, this can only be done by autogenerated information systems. Only big will survive and I doubt they will share their information with others. Internet will be much less informational then now. Google will have less pages to get "Good" information from.
But all this machine learning does not explain why some pages suffer while other ( low content pages ) get a burst.

[Martin Ice Web]

ohno,

i see very close IP ranges that are "local/close" dail-in points. This behaviour seems to interact with the dial-in provider. Sometime we see a lot of german telecom ips, then it is vodafone and then another provider. But all this with "crowded" ips.
What i am wondering about is that we are industry and business based not so much private persons. But all we get is privat IPs, we donīt see the business IP- only in short periods. I think it is not only localization but although something like our site has been classified for private and not for business?!

[Splugged]

@Martin

If will be this "scenario"... G will fall into the deep. People want informations, not monopoly. G will die.

[ohno]

I need to check our foreign traffic, I do know we get spurts of one country sometimes 2 at a time but I've not drilled that down to IP's yet........

[Martin Ice Web]

Ohno, would be nice to know what kind of traffic you get - is it business ( like BMW, Mercedes, BASF.. ) or is it private?

[ohno]

Mostly private here. I've just gone back over our logs & can't see any obvious grouping on the foreign traffic.

[edited by: tedster at 2:25 pm (utc) on Sep 13, 2012]

[sunnyujjawal]

Got drop in some "A" category sites. Is there any update on 12 September?

Keywords which were on 1 to 3 page are not in 100 even.

[netmeg]

(FWIW, none of my sites are down and some are very up - by mine I mean my own sites and the client sites I manage. And yea - I mean sales, not visits. )

[backdraft7]

netmeg - just curious, are your sites CMS based? or straight up html? The reason I ask is that from my observations, wordpress, joomla, drupal based sites seem to now do much better than hand crafted html code which for many older sites is sloppy and prone to compatibility issues.

It seems a kid with 1 photo, 1 sentence and a wordpress site can easily outrank older "authority" sites.

[Martin Ice Web]

It seems a kid with 1 photo, 1 sentence and a wordpress site can easily outrank older "authority" sites.

No that is the "extra" MAtt Cutts Information that brings this sites up.

I see pages with internal search results from brands coming up, but they do not have any information on it, just
"we did not found the item your are looking for"

Our site is next do dead, yesterday big day, today dead.

[netmeg]

netmeg - just curious, are your sites CMS based?

All of my own sites are WordPress; those aren't ecommerce. The client ecommerce site are all made up of various packages, some extremely heavily modified - Magento, OSCommerce, even a couple of <cough> Zen Cart sites. Plus some client WordPress installs with shopping cart plugins. There are two client ecommerce sites that are database driven but our programmer wrote all the code - they're like ten or more years old, and they're doing fine, but they're not terribly competitive.

I have a hard time thinking it's just code though. Have you ever looked at any of these ecommerce package's source code? Spaghetti would be an upgrade.

[bobsc]

but they're not terribly competitive

If those sites were in a competitive market - you'd be singing the blues too.

[netmeg]

No I wouldn't. I would be doing something to ameliorate the problem. But that's not a topic for this post.

[backdraft7]

I believe if we were to group all the sites by daily traffic volume, we'd see a pattern emerge.
This time of year, my niche site typically gets 1500 to 2500 uniques per day. If you are seeing problems and you're near (or below) that range, then maybe that is a clue. If you are doing 10k or better per day, I'll bet you're not noticing the problem as much, if at all.

[bsand715]

"seems a kid with 1 photo, 1 sentence and a wordpress site can easily outrank older "authority" sites."


"No that is the "extra" MAtt Cutts Information that brings this sites up."

Don't understand answer MIW

The reason I ask is in one of my niches where a site of mine was an original authority #1 for a number of years now floats from 3-7. The sites that control 1-3 are 1 photo, 1 sentence wordpress sites.

But by any analysis of web sites these new sites simply should not be ranking.

[zeus]

bsand715 - welcome to Panda,you will see a lot of site that dont belong on the first pages of google. That what I dont understand when I look at what google says you have to improve or "dont do that" guide, those that ranks are sites that ONLY do what google dont recommend. Especially link exchange, almost no real content, those two really makes me wonder, why sites that focus on that ranks on page one for MANY keywords.

[SnowMan68]

The sites that control 1-3 are 1 photo, 1 sentence wordpress sites.

With thousands if not tens of thousands of backlinks for the keywords they rank highly for?

[bsand715]

Snowman they have very few backlinks, one even has broken u-tube link to photos. Simply not good contructed sites. I don't know.

[bsand715]

Maybe my site is a lot worse than me or google though it was for 8 years.

[SnowMan68]

What is the search volume for the phrase or phrases?

[SnowMan68]

*monthly

[bsand715]

25k

[bsand715]

That is for main keyword / phrase. over all of site approx 90 - 95K

[Martin Ice Web]

@bsand

it was meant cynical/ironical. Sorry, but I think there is no good explanation why this site ranks. I see this a lot of this sites. All my competitors and myself have been replaced by such low content sites, thanx to panda.