Latest G algo changes regarding link count and placement

[emils]

Greetings everyone,

I monitor about 50 websites of different types which all draw traffic mainy from Google. Some of these have been affected by the latest algo changes occuring after Google visited these sites in mid December and it appears to be tightening in the first days of Jan. We have addressed the issues and now the sites are recovering nice from these losses.
So I thought it would be useful to share with everyone what we observed, and managed to fix:

1. It appears that G has increased filtering on pages having more than 100 links on them. We had pages with roughly 100 links on them (internal links) and all did ok. Everything else going up like 150 or 200 links (again, all internal) have been dropped. These pages where ignored, dropped from index and those links on them also not followed. When simply decreasing the number of links to 100, woops - they are back in index and all is ok. So, from now on, myself I would really begin taking more likely Google's recommendation to have NO MORE than 100 links per page. Please note that this applies to the total number of links, including internal and external ones.

2. Some of our sites where a bit interlinked using links placed at the bottom. Nothing excessive, just a couple links here and there. Some sites have higher PR and this boosts the traffic for a couple other smaller sites. After the Mid december update, those secondary sites have lost a lot of traffic and their pages going deeper in index. What happened, is that Google no longer valued those links because they were placed at the bottom of the page. What we did, was to take them upwards in the page (in the left hand menu). This did the trick, and at the next G crawl traffic was restored and everything is fine.

So, conclusions and my recommendations would be:
- From now on, stick firmly to the 100 total links per page limit.
- Do not place links at the bottom of the pages if you care about them pasing PR and boosting up the sites you link to.

Comments are welcomed.

[modernsavage]

Interesting. When you say the bottom of the page, what does Google consider to be the bottom of the page? Do you mean visually at the bottom of the page or the bottom your HTML Code.

For example if you have a table with 3 columns and you have links the bottom of the first column this might be at the bottom of the page visually but not in the code.

Do you think it's worth putting a small paragraph of text at the bottom of you pages to push the links up.

[emils]

Originally Posted by modernsavage

Interesting. When you say the bottom of the page, what does Google consider to be the bottom of the page? Do you mean visually at the bottom of the page or the bottom your HTML Code.

For example if you have a table with 3 columns and you have links the bottom of the first column this might be at the bottom of the page visually but not in the code.

Do you think it's worth putting a small paragraph of text at the bottom of you pages to push the links up.

The position in the HTML file is what appears to count. I am not sure if putting a small paragraph of text will help, never tried that, but I believe it could help if the paragraph isn't that small. Placing links within the text would be best. It is hard to draw a line in the sand here, like with many other SEO guidelines, and of course this is affected by other inpage factors as well -- but it appeared to us like as higher in the page the link is placed, as better the effect. For example there are pages with a 3-columns layout, 2 small columns at edges and a large content column in the middle. If I want to place a link that appears at the bottom of the page, but still working, then I place it at the bottom of the left column because in HTML that would be somewhere in the first portion of the page. And it works. If I place that at the bottom of the right edge column, chances are high that it will be ignored.

Another thing - the negative effect increases if there are more links placed one after another without text content in between them. Google appears to isolate that area somehow as an 'inpage links area' and at cases completely ignores it, although all other links scattered over the page appear to make effect.

A third effect that has been noticed in case of outbound links, (I am not 100% sure about this yet but it really appears as such) is that if there are such 'ignored' bottom links, the sites they target seem to get slightly worse in performance. I have obtained greater results by removing the links completely than having them but a the bottom and ignored. And of course, when moved upwards, links work and performance goes up, due to the passed PR.

[activeco]

Emils, thanks for sharing your experiences.

How big are your html files? Is it possible that Google don't spider them completely due to length facor (>101kb)?

[crankydave]

Originally Posted by activeco

Emils, thanks for sharing your experiences.

How big are your html files? Is it possible that Google don't spider them completely due to length facor (>101kb)?

My question as well. What factor, if any, did page size play based upon your observations.
THANX!

Dave

[cbp]

How do you know the changes you made were responsible for the sites coming back up?

I have seen a number of sites hit by the mid-dec change, come back on there own with no changes and others that did not.

CBP

[jackson992]

I don't know about the second comment but the first one is patently false. Some of the only pages where I'm getting google traffic is my sitemap pages that have from 100-1000 links on them.

[emils]

Originally Posted by cbp

How do you know the changes you made were responsible for the sites coming back up?

I have seen a number of sites hit by the mid-dec change, come back on there own with no changes and others that did not.

CBP

Well there are a couple details that come to answer this. The changes were applied only to part of the sites, and sometimes not to all sections. Did not have time yet to review and modify all of them. All that has been modified has been already restored in G's index. Everything that was not patched yet, is still down. You are btw right about the mid-dec change comeback (saw that too here and there) but i'm positive this isn't the case here.

Furthermore, is the exact timing. Google crawls our sites real often; more or less on a daily basis (this goes on like 3 weeks our of 4 each month). From the moment when the site was modified, it took precisely 2 days overall from the following google regular visit, for the spider to rescan our pages and then pages appeared in index back ok. The timing was too accurate in all cases following the page update moment (changes have been applied at diff dates to diff sites) to believe of anything else. We also have realtime Google spider monitoring, and watching it closely btw.

Originally Posted by jackson992

I don't know about the second comment but the first one is patently false. Some of the only pages where I'm getting google traffic is my sitemap pages that have from 100-1000 links on them.

Good point, but I personally never stated all this as an absolute rule, but perhaps as a guideline which should be taken into debate. Obviously, it may apply different to different sites. Nobody can really say where the line in the sand is really drawn, but we may figure out that it exists. The fact that it didn't happen in your case doesn't necessarily mean its false. This appears to be an algo tuning, not a fixed limit; other factors usually do come into place as well deciding if a page is discarded or not. High PR sites for example appear to have less problems in several aspects including this one. We just had a too broad impact of this matter, and it has been noticed in several different sites making it very hard to be just a coincidence. That's why I decided to make this post.

So, if I would be a forum member here, and read this, and feel that it may affect me, I would simply try to avoid having more than 100 links from this point, or at least test if so, if I can do that. We have sharply lost thousands of dollars in revenues accros various sites these days due to this 'minor finetune' and it was menacing all our projected income from this line of activity for this year, so this was no joke for us. I do not intend to spawn yet another myth here cause we don't need that. Therefore, comments and other experiences are always welcomed.

Regarding page size, we don't have any pages nowhere close to the 100k limit, they are all less than 70kb with no exception.

[emils]

Originally Posted by activeco

What factor, if any, did page size play based upon your observations.

I did not notice any difference based on the actual page size.

[28overpar]

We also saw an (adverse) effect late December, on some of our pages with more than 100 links on.

I'm going to leave it until the end of the month to see if these pages bounce back, before I look to reduce them.

In general - How long should a site be left after an algo change has been implemented, to see if the effects on our SERPs position is temporary or permanent?

Would it be easier to ask for next week's winning lottery numbers?

28overpar