Algo Change Targets Low Quality Exact Match Domains 9-28-2012

[martinibuster]

Matt tweeted [twitter.com]:

small upcoming Google algo change will reduce low-quality "exact-match" domains in search results.

[Kelowna]

I think Google is only telling a small part of the story by saying this update goes after emd's.

It appears to me and the sites I watch that possibly some sort of penguin refresh was thrown in as well to go after inbound links that look shady. A few weeks ago we noticed someone bombing one of our sites with profile links all using the same main keyword for about 3 days. The site is not an emd but like most sites on the net it does have one partial keyword in the url and now has tanked in the listings with this update. Now inner pages all rank lower in the serps but the main url is nowhere to be found, this smells a lot like a OOP for that keyword now. If I search for the domain name minus the .com it comes up as number 1 so the site is not penalized, just the main keywords. This would explain a lot of the mixed results people are seeing.

[WebPixie]

One thing people should realize before ditching or switching strategies. EMD's carry value beyond Google.

I agree. Our highest revenue site that was hit is still doing decently from number 1 positions on Bing, Ask, and basically every other search engine. While we certainly miss the traffic from G-, I am not about to start making major changes to appease Google that could hurt rankings on other SEs.

[crobb305]

I think Google is only telling a small part of the story by saying this update goes after emd's.

I have to agree. I am seeing non-EMDs that took big hits. Also, a couple of my own sites saw sitelink reductions (from 6 to 4, non-EMD).

[plc90210]

I think the length of the EMD domain also might have been taken into account in the algo update. For example, although cars.com is an EMD, it is not spammy and is brandable but if you have something like paydayloansforpeoplewithbadcredit.com, it's likely you're just trying to spam and get the keyword-rich domain ranking boost.

[MrSavage]

I haven't experience in life a situation where I'm segregated. Up until now that is. Because I'm a certain way, people (Google) and some people here are making assumptions. Oh, you must be a junky site and deserve a different set of rules because I am a keyword domain. I can tell you this sucks and it's not even about me as a human. It's opened my eyes and not in a good way. Gives me a new perspective on racism. Sound odd saying that, but I'm feeling a bit tarred and feathered right now.

[SevenCubed]

I've discovered an interesting side effect happening, more traffic coming in via long tail search for my affected site. Looks like searchers must be getting disappointing results using the previous "servicekeywordname-citykeywordname" search term so they're refining searches by adding one more word "servicekeywordname-citykeywordname-provincekeywordname" and for that the affected site comes up in spot #1 :)

For those who have been affected have you tried searching by adding one, or two at the most, more words to see how well you rank for longer tail terms. If you still rank your site might not be too bad off?

[zarathustra2011]

I see a lot of spammy EMD's, but plenty of decent ones too. Even though the site they destroyed isn't an EMD (and has always been high quality/authoritative in its subject matter), I now seem to have a penalty set against me. I believe even if I wrote an amazing page, full of unique original, and innovative content, it would never appear beyond a certain place within the SERPs because of what has transpired. I've spent today searching certain unique key phrases, and sentences from pages that were always ranked highly, and I either do not appear at all, or I appear below the most spammiest scraper sites I've ever had the misfortunate to come across, that have taken just a few sentences or a paragraph from my pages.

This update can't be just about EMD's, but at the least, why don't Google simply prevent websites from having a boost in the results for choosing keywords in their domain name, rather than going to the other extreme of seemingly penalising them? Also, why the inconsistency, because I am see plenty of very poor content EMD's (and non EMD's) at the top of the results I've tried today.

I have a lot of one way backlinks from a very diverse range over many years, including many authoritative websites, yet scrapers are kicking my arse now. What a mess google. There must be better means for determining "quality" than this.

[edited by: zarathustra2011 at 7:08 pm (utc) on Oct 1, 2012]

[mhansen]

I think Google is only telling a small part of the story by saying this update goes after emd's.

We have a 4+ year old, 8-page website that sits on an EMD for a phrase that generates about 25k monthly exact-match searches. It was untouched during this update for the most part. (Lost 1 position from 2nd to 3rd on the exact match query, personalization off, verbatim, etc)

I have however seen a few non-emd domains suffer with what looks like a -950 (end of results) penalty (more info [webmasterworld.com]) based on over optimization, or at least a high saturation of keyword phrases.

What it all leads me to believe is that:

1 - G* rolled out an EMD algo change. (affecting low quality emd's)

2 - Because of that EMD change, G* also recalculated the algo that relates to over-optimization (maybe related, or close to a Penguin signal?)

3 - The results we are seeing are only partially related to the EMD drops, but other changes can be related to any number of other factors and algo updates, NOT related to EMD, but a result of the complete data recalculation.

MH

[zeus]

Once again I just dont get it, what the point is with this update, of cause a lot have keywords in there domain, be cause there site is about that. If thats a problem then just dont calculate with the domain in the algo. Another crazy update like Panda which hit so many innocent sites.

[Sand]

@zeus,
Hopefully, they're just removing any boost than the EMD may have given you in the past (rather than devaluing the entirety of a site because of its domain). It's too early to say, but I sure hope they're just removing the benefits to level the field.

[zachary]

@plc90210 search in Google UK for Viagra and you will see keyword1viagrakbkeyword2viagra .co .uk coming up in the result
.

[edited by: Robert_Charlton at 7:10 am (utc) on Oct 2, 2012]
[edit reason] Obfuscated specific [/edit]

[zarathustra2011]

I don't think google.co.uk has been affected in the same manner as google.com (not say something isn't forthcoming). I still rank okay on UK but on COM I've been condemned.

[zachary]

.co.uk have also been affected i have seen too many changes in .co.uk

[plc90210]

@plc90210 search in Google UK for Viagra and you will see keyword1viagrakbkeyword2viagra .co .uk coming up in the result

Cutts said the algo targeted low quality EMDs, not just EMDs, in fact some EMDs even moved up after the update. The EMDs that ranking high now probably have good backlinks.

[edited by: Robert_Charlton at 7:13 am (utc) on Oct 2, 2012]
[edit reason] Obscured specific domain & added quote box [/edit]

[zachary]

@plc90210 i agree that domain might have good backlinks but the thing is by looking at domain name itself one can see its spam, that domain name doesn't make any sense. just to increase keyword they have added viagra twice in the domain :)

[SevenCubed]

just to increase keyword they have added viagra twice in the domain :)

That's assurance the site will never drop!

[lewis1]

@plc90210 yes, but have you seen the result at number <"n">? I'm seeing some <widget> website called <Doing Widgets>. Not sure why they are ranking for viagra!

.
Mod's note: Removed specific site ID here. Chances are they're ranking because they've been hacked.

[edited by: Robert_Charlton at 7:19 am (utc) on Oct 2, 2012]
[edit reason] Removed specifics, added comment [/edit]

[zachary]

@lewis1 this algo seems to be a joke to me :)

[lewis1]

I also checked on Google.com and did a search for "search engine". It seems Google doesn't see itself as a search engine any more. Searching for "pay per click advertising" does bring up Google.

I think I'm starting to understand this update now ;)

[superclown2]

One of my EMD (and high earning) sites has been demoted from it's number one position for it's main keyphrase to number 5 and I'm relieved. Why? For months I've been aware that it could be greatly improved in ways that could make it rank for more keyphrases but I couldn't touch it because I was afraid of losing that top spot. Now I can get to work on it and I'm confident it will emerge a more useful site and a better earner than ever before.

Over the past decade and more my sites have been hit or even wiped out more times than I can count and every time I've improved the product which has resulted in even higher earnings than before. This time will be no different.

Ditch EMDs? No way. The public trust them more than they would trust www.my-favourite-square-widget-20002.info and that can't be bad for business. Much better spending our time taking a long cool look at the demoted sites and working on ways of making them even more relevant and useful for visitors. Too many of us are blinded by the brilliance of our professionally written copy when we should be thinking; "is this really useful to the visitor?" "Is the copy 100% relevant without any fluff whatsoever?" "Does this site answer the problem that visitors want to solve in a way that is more efficient than the competition?" "Have I stuck to the KISS principle (Keep it simple, stupid?)"

Maybe this update was insane, maybe it was a brilliant improvement. The question is irrelevant because either way we are probably stuck with it for a while at least so there is no point in our complaints, predictions of personal disaster, accusations against Google. None of them will make the slightest difference.

[edited by: superclown2 at 8:12 pm (utc) on Oct 1, 2012]

[randle]

An "EMD" doesn't make a site good or bad, low quality or high; its just a bunch of characters. To be honest, it was a little surprising to see Google, via Matt Cutts, put the two terms, "low quality" and "exact match domains" in the same sentence.

One has nothing to do with the other. Do "EMD's" have a higher percentage of "low quality" (googles term not mine by any stretch) sites than non "EMD'S"? Absolutely, but whatever letters were configured to make the domain name dont have anything to do with whats in the site, or the user experience.

The actual identifying of the "EMD" has got to be a very small part of whatever formula the Googlers cooked up to attack a genre of site they would like to see leave their party.

(people link to bluewidgets.com with the text "blue widgets")

The single biggest gaming technique employed in Google since the very beginning has been people pounding away at "anchor text" to produce a resultant key word ranking.

This thing reeks of a classic over optimization penalty - the whole things pretty unfortunate, another ill conceived stab at improving the quality of their results by using a broad brush to penalize a type of site they think is "low quality", instead of concentrating on scoring the "quality" sites higher, and letting the weaker ones fall to a natural death. (which technique do you think is easier to accomplish?)

Intentional or not, if its lots and lots of this:

<a href="http://www.bluewidgets.com/" target="_blank">blue widgets </a></p>

pointing to your site, and all in between every page, Google has decided your trying way to hard to get them to know what your sites about.

[incrediBILL]

This thing reeks of a classic over optimization penalty - the whole things pretty unfortunate, another ill conceived stab at improving the quality of their results by using a broad brush to penalize a type of site they think is "low quality", instead of concentrating on scoring the "quality" sites higher, and letting the weaker ones fall to a natural death. (which technique do you think is easier to accomplish?)

I think they do both all the time. While attempting to raise quality sites you also need to push low quality sites down to eliminate a level of noise.

Obviously their approach isn't perfect and some of the collateral damage being done is just as ugly as Panda and Penguin making many unhappy but those that take their places are probably quite pleased.

Right now I am more interested in discussing the changes this so called EMD update has produced.

We all are but my point was simply that those hit hardest should reevaluate their long term strategy which is perfectly a valid point.

I won't be tossing any of my EMDs just because Google doesn't like them and I definitely won't be reacting to the change until they perform an update. At this point flailing in the dark without seeing the big picture of what Google is attempting to accomplish could cause more harm than good when they tweak the EMD update IMO.

[edited by: incrediBILL at 8:25 pm (utc) on Oct 1, 2012]

[Fiver]

I keep an exact match 4 word domain which is undeveloped (one pager, one paragraph on-topic, handful of three-four year old links) which acts as a barometer for me on this - No change, still in the top 5 for my low-volume (but high profit) phrase.

Two adsense units on the page, no way to leave but adsense or the back button, and I assume I get a lot of back-outs... but still have my ranking.

[lewis1]

randle, that's a good post there. However, the site of mine which was worst hit (it is literally nowhere for anything) was the one with the most diverse link profile. It's the one I always felt would sail through any update as it's the one that is and looks very natural in it's look and feel and in its links. Nearly every link has a different anchor text, and plenty of links to inner pages.

But you may have something in your post though....

[ethought]

I am seeing what appears to be a bit of a recovery on sites that got wiped out.

The following pattern is present across quite a few sites that got hit. Here is a few examples.

Site 1:
28/09 = 200 (uniques)
29/09 = 77
30/09 = 5
01/10 = 71

Site 2:
28/09 = 138
29/09 = 49
30/09 = 3
01/10 = 31

Site 3:
28/09 = 261
29/09 = 224
30/09 = 39
01/10 = 302

Traffic seemed to dip unusually on the 29th. But the 30th was by far the worst day for me (Australian time - so 29th evening in US). But many sites seem to be showing uniques going back up to 29th levels.

Anyone else seeing this?

[nonstop]

@JamesMacf it looks like your domain is back!

[randle]

those hit hardest should reevaluate their long term strategy which is perfectly a valid point.

They (and all of us at one point in our business lives) already know that, its just a really sensitive time to be hearing it in the very community they come to for a little help, venting and understanding.

[Leosghost]

@JamesMacf it looks like your domain is back!

as far as #5 ( effective #8 due to 3 adwords )..improved,( and no longer subjugate to the dotcom) but..you don't get many clicks at #5 ( effective #8 )..top 3 ( under the ads ) is what is needed..

Given the parent co..I'm surprised that James didn't see this all coming :)

Personally, I'd wait for the dust to settle before making an offer..for any domain for sale..;)

[RedWine]

Good info, thanks ethought. Hopefully we see some improvement in "high" quality domains that have been effected by this "minor" update.

I personally have been at this since before adsense even existed. I did it because I loved it and was fascinated with the potential and power of the web (I was not making any/much money in those days).

Google has done many great things for the web (esp adsense). But it is updates like this that reminds me how much I dislike the amount of control and influence they wield today.

It seems clear that keyword domains should have neither preference or penalty (other than what they collect naturally via other algo metrics... such as click through rates).

It is healthy to remember that Google is a corporation and business first. I am not so naive as to believe that EVERYTHING is always done to benefit the user/customer or improve their experience.

The quickest way to put the focus on the customer is competition! I think everyone would be better off if the search market share was split evenly three + ways today! Can you imagine if we had three different adsense caliber programs to choose from as well? I hope we get there one day.

[Color7]

Mine has been jumping around today. I was at position 11 pre algo change, then nowhere to be seen post change. This morning I noticed I was back at spot 13. Just checked again, gone.