Google Updates and SERP Changes - July 2013

[spreporter]

---

System: The following 3 messages were cut out of thread at: http://www.webmasterworld.com/google/4580257.htm [webmasterworld.com] by goodroi - 9:35 am on Jul 1, 2013 (utc -5)

---

RIP Tedster, though I never knew you personally for 8 years now I've learned a lot from your great contributions

[CaptainSalad2]

@god

"same ranking positions but less traffic" possible answer..

Has Google purposely, for lack of a better word “dumbed down” the algo so that it can no longer map plurals to singular words/phrases?

A lot of searches where I type in “widget in area” are vastly different to when I type in “widgets in area”. Same goes for stop words like “in” “and”. “widget in Area” will now often deliver different sets of results, 3 different sets of results for the underlying exact same query.

My understanding was (and certainly seemed to be the case until recently) was that stop words were ignored by the algo and plurals never used to be a problem, from a programming point of view easy to fix, very easy.. if there is a will. Of course by changing the algo to now view plurals as different and credit "stop words" it makes organic results even more un reliable and PPC more attractive?

[edited by: CaptainSalad2 at 6:38 pm (utc) on Jul 11, 2013]

[bsterz]

I seem to keep reading statement like "my rankings are unchanged, yet my traffic has dropped significantly, which just does not add up.

I see this with an informational site I have. The query in question results in very few adwords ads (1 or 2). The two biggest possibilities in my mind have been mentioned on this page. Either it is personalization or auto suggest. It could be user behavior changing, but the numbers look to dramatic to back that up. Any other theories?

[diberry]

dibery, could it not be increased personalisation of results?

I'm not sure how to tell. I'd assume if personalization was the cause, not everyone is seeing my pages at #1, right? So I've checked Google on several computers, IPs, etc., as well as ranking tools, and I can't ever "catch" it not being positioned at #1. Am I missing something there?

Diberry, if there are no Adwords ads Google isn't making any money from those searches, so maybe they started directing people to searches they can show Adwords ads for (via the search suggestion dropdown).

While I'm definitely seeing Autosuggest alter searcher behavior in some cases, I don't believe it is in these cases. At least not based on my own testing of auto suggest with these queries, on several computers, on several IPs. These queries are also kind of unique (hard to explain without violating forum rules) and nowhere near any query that is monetized.

How is your (not provided) entry doing? Up or down compared to the same point last year?

It's up a very little bit, but nowhere near enough to account for this. Where it's REALLY up is with Google UK and AU, which I can't begin to explain. Especially since I rank #1 for these queries in those countries as well, without logging in.

I can see no evidence that there is improvement in visits and then a ranking increase. It's just visits and rankings go up at the same time.

That's every bit as weird as what I'm talking about.

[Wilburforce]

@rhyd

I have just tried a personal search for a UK-specific tax question (I couldn't find what I wanted on the first few pages of Bing/Yahoo, so tried Google), and with Pages from the UK selected over 3/4 of the results are from the US.

They will need to sort that out pretty quickly!

[brianp]

"Something" is in the works. Today we've seen very very low conversion rates, huge bounce rates (10-20% higher than normal), and ranking flux like we see before a big update.

E-commerce site, USA.

[Dymero]

That's every bit as weird as what I'm talking about.

Yeah, I know. It's as if Google is interpreting an industry trend to mean fewer people see the site as quality, but then later reverses that when visits go back up. It's the wrong interpretation. The site visits go up or down, but the site itself has remained mostly unchanged.

I realize I speak as though the algorithm has sentience. I'm not crazy, I promise!

[Wilburforce]

I realize I speak as though the algorithm has sentience.

It has been suggested by several people - including Tedster a while back - that Google is at least experimenting with, if not actually incorporating, AI.

[EditorialGuy]

Just remember that, like most knock-offs, artificial intelligence is seldom as good as the real thing.

[Shepherd]

The site visits go up or down, but the site itself has remained mostly unchanged.

So during your visitor "downtime", do the visitors go somewhere else? Did some other site get created/changed/briefly more popular (possibly from a media mention or social media buzz)?

[nomis5]

It has been suggested by several people - including Tedster a while back - that Google is at least experimenting with, if not actually incorporating, AI.

Nothing personal but the subject of G using AI is, in my view, absolute bollacks.

Far, far too much intelligence is attributed to the G algo. Just look at some of the clangers they produce in their SERPS and it's clear they can't even get the basics correct let alone embark on incorporating AI into the SERPS.

Face it folks, Matt Cutts and G in general are very good at persuading us that their SERPS have all sorts of bells and whistles in them but in truth they haven't been able to solve basics yet. Not surprising really given the volume of data they are attempting to number crunch.

Knowledge stuff, AI, predicting what you are searching for before you even know yourself is all a smoke screen to make G appear to be far more clever than they are.

There is are very basic factors which G use to present their SERPS and it's basic and obvious. User types in a query, user selects a SERPS result, user stays there longer than a few seconds and that means to G the site is worthy of inclusion in their SERPs.

Other factors also apply of course and the most important of those is that if your site challenges G's interests then no matter how good it is you will be demoted in the SERPS.

So my rules are, first don't compete with G, second produce a sticky site, preferably one where users click on G ads.

[SEOGuyAU]

[mozcast.com...] is reporting a bit more flux that seems to be gradually increasing day by day; I'm going to say Panda is starting to roll out.

[backdraft7]

started two days ago for me...this one is initially killing me, but things will prob pop back up before they do it all over again, and again, and again...I think by now we all recognize the unpleasant patterns.

[SEOGuyAU]

Look at the mozcast.com graph; exactly 30 days ago it was roughly the same flux in the SERPs; I'm confident to say that it's Panda rolling out.

[SnowMan68]

Panda hasn't rolled out since the end of May. Matt Cutts was quoted about this.

[SnowMan68]

Edit: End of April

[SEOGuyAU]

Do you believe Matt Cutts, or do you believe readily available data available to us?

I personally trust the data that's available to us then what MC says...

[SnowMan68]

At that point I don't see a reason to lie. I know that a lot of the stuff they spew is FUD, but there was no reason during that conference. Not to mention no one really has said much about panda in the last 9-10 weeks. If its running now I'm betting its the first time since the end of April. I have been anxiously awaiting it.

[backdraft7]

Just complete the most recent keyword check...ALL downward movement in -7 to -9 range, lost many recently regained terms...looks like they just cranked up the bias...and too high again.

[Martin Ice Web]

@backdraft, we see the change. We were doing very good the last 3-4 days. And today it seems like they hit the "bad" button again. Hours with nealry no traffic are becoming usual, unfortunately.
At least i have to say that Amazon and some big brands do not have this precense in the serps anymore. Hope it sticks. IMO the lifted a Little bit the Panda restricitions so that niche sites come up again.
But it could be that it is due to prepanda rollout, like every time before a Panda roll out.

My personal watch cast "bing" is from yesterday 417.000.000 down to 379.000.000 results.

ecom, germany

[kanetrain]

Seeing fluctuations tonight as well. Amazon and other big retail sites have absolutely dominated our nich for the past 5 months and that seems to have backed off a little tonight. It's been almost breathtaking how all of the niche sites (experts in my industry, top stores, etc.) have been completely pushed out for amazon, target, walmart, etc. But maybe we're seeing G back off of that somewhat.
Fluctiations tonight seem to be seeing some of those expert niche sites return to higher rankings... but I would not be suprised to see it reverse course within a few days again. Has been the trend lately.

[turbocharged]

that Google is at least experimenting with, if not actually incorporating, AI.

Artificial yes, but intelligence? I just don't see it.

A shift in user behavior may have negated some of the benefits of general keywords. Having stuck to their guns for months, Google's serps have been so out of whack for so long that people still need to refine their queries to find what they are looking for - or to be presented with a diversified set of results. That's assuming that they can even find it in Google and don't go elsewhere.

To be honest, unless it is for a work related task for a client, I personally have not used Google search in a while.

[Alex997]

MASSIVE drop in SERPs across the board for my site in the last couple of days on google.co.uk. Between Penguin 2.0 and this new hit we have lost over 75% impressions and clicks to the site.

Top of the rankings is now a site that has 12 links from unrelated cookery website comments.

Utter madness Google - what are you doing?! (rhetorical statement)

Bing has become my new default search engines - and I'm surprised at how much better it is.

[bsterz]

Seeing quite a few new 'Searches Related To" for a few queries I follow.

[Wilburforce]

Artificial yes, but intelligence?

It isn't a question of how good it is (indeed, it is how bad it is that makes it more likely in my view that AI is involved): as soon as it incorporates some sort of "learning" (i.e. self-modifying behaviour on the basis of observed results), we're moving into AI.

As I said, several people have suggested this may have already been or may soon be incorporated in Google's algorithm, and I wouldn't personally reject out of hand the possibility that they are right.

[ColourOfSpring]

Seeing quite a few new 'Searches Related To" for a few queries I follow.

I'm seeing more and more "Results for Similar Searches". To me that's a tell-tale sign that Google aren't confident in the results they're showing for the original query. It's an artificial attempt at producing diversity of results through two queries instead of one.

[Dymero]

It has been suggested by several people - including Tedster a while back - that Google is at least experimenting with, if not actually incorporating, AI.

I think they're trying, but haven't quite gotten there yet. At the very least they're doing machine learning, which is different, and could explain the increase in work Matt is giving webmasters in terms of nofollows and disavows: creating data points for the machine to learn (but this could be entirely left field).

So during your visitor "downtime", do the visitors go somewhere else? Did some other site get created/changed/briefly more popular (possibly from a media mention or social media buzz)?

As far as I can tell it's an industry dip. There's just fewer people looking for products in our niche at the time. It's a consistent trend for the last three years. Why rankings seem to follow is the mystery here. If anything, mini- and affiliate sites in our niche have been knocked out to some extent, but this was a few weeks ago.

[TyMax]

In some ways it has not bothered my main sites, but in other ways in highly valued key word phrase I have gone from first or second to last on first page on first on second page. I think Google sees duplication on two pages. But one page defines types of the product, the other lists the most common type of the product.

No matter what minor tweaks I try Google still has demoted a page that was a benchmark for defining the types of this product for well over 12 years. Other websites copy and slightly alter the definitions. This Google upgrade is the first time it has caused problems. Mean while it does extremely well ranking wise else where.

I had a lot more hair before I started SEO work. This change is not helping.

[Shepherd]

Observation: for a keyword that I follow, first organic result is wikip, numbers 2 and 3 both are set up to make it difficult to return to the serps, you have to click the back button repeatedly to get back.

[SnowMan68]

Whatever is rolling out it doesn't feel like Panda. We didn't get the typical deep crawl leading up to it. This crawl was similar to the other updates that have been rolling out of the past couple weeks now. Pre Panda crawls for our site are typically 2-3 times the norm.

You would think that when the update to Panda hits they will do a really big crawl of the web. It's been almost 10 weeks now since the last known Panda.

[backdraft7]

Domain crowding in my vertical is right back to where is was months ago. Ho*zz is again being listed for countless unrelated or shallow related terms. Looks like they renewed their sweetheart deal with G. I do not find anything I'm looking for, only thousands of keyword loaded comments on their pages. Keyword spam apparently still works if you're well funded. I've dropped from #1 on many pages to page two or worse. Stock price reflects record profits again, $917.92 /share.