Link Spammers Beat Yandex, Forcing Yandex to Change Algorithm

Dec 9, 2013 • 9:18 am | comments (22) by twitter Google+ | Filed Under Other Search Engines

yandex logoAndy Atkins-Krüger posted an awesome news story at Search Engine Land covering how Yandex is changing their algorithm for some types of queries to no longer count links as a ranking factor.

Yandex is the leading Russian search engine, way bigger than Google in Russia. They announced last week that they will no longer use links in their ranking algorithms specifically for commercial queries in its Moscow region. This accounts for about 10% of the queries Yandex sees, which is a lot.

Alexander Sadovsky, who is the Head of Web Search at Yandex, told Andy:

There is a lot of noise around the links signal particularly for commercial queries and especially in Russia. We see a lot of paid links and even automated paid links where there is no human actually involved. The problem with these links is they’re frequently off-topic and are effectively cheating users.

Alexander added that "there is a bigger problem of spam links in Russia than elsewhere because Russia has a lot of programmers who have been turning link building into a new profession."

The truth is, there are tons of link networks in Russia that spam and manipulate Google, as well as other engines. In fact, Google publicly went after two well known ones including Ghost Rank and SAPE Links. Heck, even this morning we covered another link network Google went after.

It seems like while Google actively goes after the link spam, Yandex is simply going to remove links as part of the algorithm for the query types that get spammed the most? It feels like when search engines stopped using the meta keywords tag but a heck of a lot more shocking.

The question is, how long can Google keep manually breaking into link networks and then reverse engineering them. Are they fighting a battle they can get ahead of or will they have to take Yandex's approach and remove links from the algorithm for certain query classes?

I did ask Google to comment on this and they declined.

Forum discussion at Google+.

Previous story: Google Voice Search Answers In International Languages



12/09/2013 02:53 pm

This is well-known fact, that big companies have the budgets to be spent on SEO. This is also the fact, that there is almost no methods of the results control. Budgets must be spent and that's it. This is also the fact, that most of so-called SEO agencies are pure link builders. So, we have money to be spent no matter what and those who is always ready to spend, pretending they do work. This is endless run. Matt Cutts can't win. It's just the matter of time when Google will follow Yandex approach. In fact, I believe Google tries already to disregard links as much as possible, but the lack of the other signals doesn't allow to disregard totally. Anyway, the direction of SE development is pretty clear. Happy SEO )


12/09/2013 04:41 pm

First time I've heard this story put quite like this. Most places seem to be viewing the news as "Yandex are doing something cool to revolutionise search" rather than "Yandex have failed to beat spammers". Maybe this is just their method of 'winning' rather than perceivable failure? Sooner or later, time will tell...


12/09/2013 05:23 pm

Hm, you're right ... I even didn't pay attention how Barry put it upside down ) Barry, could you please explain how did you make conclusion that " Link Spammers Beat Yandex " ?


12/09/2013 05:53 pm

because google already using links as small factor. It widely using white lists and manually reviewed sites in organic serp


12/09/2013 06:00 pm

yes, it really clear. people need to find our sites without using any search engines, such as google/bing/yandex. We need to find another way, like open source (non-commercial) search engine for example.


12/09/2013 06:20 pm

Algo is just the small part of any search engine. Search Engine is about the Big Data first of all. I can't find it right now, but I read somewhere on DuckDuckGo, that they can't afford crawling and storing data, so they have to use Google's data. There are just two Big Data holders on this Planet - Google and Microsoft (NSA chases them)). Any search engine will be dependable on them anyway. It will be this way until the hardware will be significantly cheaper.


12/09/2013 06:24 pm

I'm curious what does it mean for DuckDuckGo, keeping in mind this :


12/09/2013 06:26 pm

may be cloud databases can be the answer. something like shared database hosted on users computers (not now, but it can be in future). To use free search, need to provide some space on your hdd. When index is counted (every day for example), it can be stored in one central or cloud location for quick access.


12/09/2013 06:34 pm

Yup, probably. Anyway, there is no reason to expect new noteworthy players on the search engine market, until the cheaper way of collecting and storing Big Data will be found.

Durant Imboden

12/09/2013 06:51 pm

One could argue that the introduction of link popularity and PageRank were the beginning of "Search 2.0," while Yandex's baby steps toward an algorithm that doesn't depend on links is a forerunner of "Search 3.0." In words, what Barry sees as a sign of failure could just as easily be viewed as a sign of innovation. It's all a matter of perspective.


12/09/2013 07:16 pm

"It widely using white lists and manually reviewed sites in organic serp" How do you know this? Where is the evidence?


12/09/2013 07:58 pm

lot of evidences. just look into wikipedia spammy backlinks. or youtube spammy backlinks, or amazon spammy backlinks, here is also lots of another examples. However this sites feel good and stay at almost every top10.

Michael Martinez

12/09/2013 10:32 pm

"The question is, how long can Google keep manually breaking into link networks and then reverse engineering them," I'm not sure that is exactly what they are doing but it's a good question in principle.

Michael Martinez

12/09/2013 10:33 pm

No, no. Don't say "whitelists". Say "learning sets". :)


12/09/2013 11:12 pm

until they will able to sell organic top 10 via adwords. All going to this way.

PM Fiorini

12/10/2013 12:04 am

I'm seeing that "traditional" links are counting less and less. Google is definitely moving in that direction


12/10/2013 01:02 am

in this case they will count faked behavioral/social signals, citations and faked business directory listings (may be even local). I think they are unable to repair their search engine after animal updates and have nothing to do, but search for similar "no links as votes" solutions. But that solutions will worse than with links.

Jitendra Vaswani

12/10/2013 05:17 am

Goolgle tying to catch these spam networks but still I can see some bad websites are getting good page rank. I dont know from where they get such a good page rank.

Jitendra Vaswani

12/10/2013 06:03 am

I agree with you durant


12/10/2013 10:02 pm

Interesting to see how this will go. But the truth is that if the programmers are that skilled to make Yandex abandon links in their algorithm, what stops them from going after whatever else they come up with for ranking the terms? Isn't there a saying that goes "if there's a will, there's a way"?


12/11/2013 03:27 pm

You have to understand that the game in Russia is different. There is no such thing as the clean site. I asked matt at SMX advanced last year about link buying and link networks in Russia, his answer was - you have to detect what's NOT spam Russia, rather than catch them spam.

Saleh Stevens

07/23/2014 04:54 am

Google is gearing toward being the Matrix, always trying to govern and define reality. Neo is inevitable. The question is: who, what, or which network is Neo?

blog comments powered by Disqus