Are the ranking effects of new links changing?

[MrFewkes]

The way links are being used to rank sites has changed.

Its taking longer for them to be effective - if at all.

I think old existing links are still holding most of whatever power they had whilst still in place.

I think newly found links are being looked at differently.

I cant explain it - but I know its true.

I remember a couple of months ago some google engineer saying "we will be changing the way we handle links soon" - well - its happened I am certain.

Anyone?

[almighty monkey]

Page quality (Panda iterations)
Business local citations (linkless citations)
Business trustability
Brand development / permeation
Domain ownership / history
Social signals
Authority of inbound links
Other (still to come)

I agree with most of this list. Can I ask what you mean, though, by 'Business Trustibility'?

[triggerfinger]

I may be wrong, but I tend to think that the precedence of link metrics in the ranking algo has dropped, significantly.
I see on-page content quality and user engagement metrics as the #1 ranking factor. If you fail this panda test, no amount of comment spam and/or authoritative links are going to help you.
However, if you and your competitors pass the panda test, then it will come down to other factors, links being one of them.
If you take a look at overstock, or any of the other big link buyers, I suspect the reason the link buying still works is because the domain performs very well with panda metrics, allowing the links to be factored in as an additional beneficial metric.
This would make sense if Google were trying to avoid showing sites that performed poorly (user bounce, etc) but still rank high due to links, DA, keyword targeting, whatever.
Any thoughts?

[wheel]

I see on-page content quality and user engagement metrics as the #1 ranking factor.

Can I get a half dozen examples of sites that rank with user engagement and onpage content quality, but no backlinks?

[MrFewkes]

I gave a site a superb quality PR4 link a few weeks ago. Index to Index - inparagraph - "keyword1 keyword2 here" - only about 3 other outbound links on the source page.

The target page jumped up 8 places within days of the link being crawled. Then - two days later - it dropped 8 places again.

Linking is dying folks - be on the ball and see whats replacing it.

In my arena - the ranking pages are totally "splattered" with about 100+ 2cmx2cm images of products, ads, buttons, accolades (usually bull#*$!), 100+ broken up blocks of text (say 15-20 words) rather than a handful of paragraphs of 120+ words.

Well - i'll be experimenting with my own "splatter pages" for ranking experiments - but I dont see how all the fussy looking in your face little graphics help the user. Sure - onpage theres more choice - but to be honest I find it all befuddling.

Conversions will be the tell tale.

I saw this kind of site ranking in the casino arena years ago - just an observation.

Summary - SPLATTER pages are the new name of the game - not links. (in my arena - niche electric goods)

[triggerfinger]

Wheel, no I don't have examples of people ranking in competitive arenas with 0 backlinks, but that doesn't provide an argument against the precedence of panda factors. Being the #1 ranking factor doesn't imply there aren't other ranking factors.
Its just a thought, perhaps the actual logic behind panda is much more complicated and/or sector dependent, but it does seem that the pre-panda, classically accepted model of SEO has changed.

[wheel]

No, nobody has any examples. That was the point behind my comment. THere aren't any examples. It's wishful thinking. Yet here we've got people telling us that link building is dying.

As for panda factors, hey, I'm no expert, but it's the same thing - people speculating with stuff that sounds good but is a load of made up horsepucks.

They've talked here about too many ads on the page, pushing content too far down, etc. bounce rates and so on. Yet have you had a look at amazon lately? If the speculation was correct, amazon would be the poster child for being slapped by panda. No decent content, thin ecommerce pages, top 2/3/s of the page filled by some ginormous ad for a tablet computer, etc.

Being the #1 ranking factor doesn't imply there aren't other ranking factors.

that's not what I'm saying. I'm saying show me a site that ranks using the #2 factor. Just one. And it doesn't even have to be in the top spot. Heck, show me a site that ranks on something decently competitive on even the front page using these other ranking factors that exclused backlinks. Surely some sites are ranking using the 2nd most important ranking factor right? Wrong.

[MrFewkes]

Wheel - no disrespect at all but Amazon is clearly treated differently than joe blogs sites.

In my arena - linking is not working.

I have tried the following - all to no avail.

Press Release
Facebook links/likes
Social Bookmarks
Bought Homepage links
My own links (strongest - used to be anyway)
Mass spam forum profiles
Gentle spam forum profiles
Video submissions with links to domain
Postings via onlywire from an onsite blog.
Tweeting

Nothing is moving in my serps.

(Testing covers 6 domains)

[wheel]

Wheel - no disrespect at all but Amazon is clearly treated differently than joe blogs sites.

What? Says who? No it's not. My site ranks alongside companies that make amazon look tiny. That's an excuse. Size of company and annual sales are not ranking factors. Backlinks are. The reason amazon is treated differently than you is that they have better backlinks.

In my arena - linking is not working.

No, low quality linking isn't working.

Press Release

crap

Facebook links/likes

crap

Social Bookmarks

crap

Bought Homepage links
My own links (strongest - used to be anyway)
Mass spam forum profiles
Gentle spam forum profiles
Video submissions with links to domain
Postings via onlywire from an onsite blog.
Tweeting

crapcrapcrapcrap-crappity-crapcrap.

Nowhere in there do I see anything that looks like you've done the work to go get a link from a gov't body or regulatory authority.

It's not that links don't work. It's that low end crap links don't work as well as they used to in all cases. Frankly, I'm not surprised.

But the conclusion to draw is that Google's getting better at finding the right attributes about links to rank - not that they don't use links anymore, or even links almost exclusively.

I showed you personally and specifically where to get a good/great link for your site. And you didn't do it. So I guess you didn't test those kinds of links.

[MrFewkes]

Wowzer Wheels - take it easy.

You are forgetting perhaps that we don't all have access to your kind of links.

My list above does include my own links - and some of them have been excellent in the past (pre-January 2011 or thereabouts) at ranking boosting.

Ok - I will "test" the kind of links you want to go with if you like - as per our PMs (i'll pm you re the specifics).

It will be a test for me for certain - and I am sure you would be interested in the outcome aswell.

Look - I am all ears - the kind of things I have been testing have been around a long time and were fairly well established - but they are now not working.

Yes - time to move on - which is why I started this thread.

RE Amazon - I think the domain - like ebay - has a lot more google love than the vast % of sites out there. Why this is - I do not know - but it ranks all over the show.

I can happily say that I will totally and absolutely DUMP every strategy I have tried - in favour of the next thing. You have no worries on that front - I am guessing that I am just lower on the curve than you appear to be.

[Planet13]

RE Amazon - I think the domain - like ebay - has a lot more google love than the vast % of sites out there. Why this is - I do not know...

Probably because Amazon has lots, and Lots, and LOTS of inbound links, I am guessing.

[martinibuster]

For some competitive SERPs, the formula certain spam sites appear to be using to exploit Google's latest algo is to buy one or two of the biggest most authoritative links they can find, then build an army of low quality inbound links from footers, WP templates, and forum spam.

The low quality links, standing on the broad shoulders of the paid authority links, are then able to lift the site into the top ranking.

Is this happening because the few authority links are enough to get a free pass on the spam links? Or is there a Panda/content algo loophole that is being exploited?

What do you think?

[MrFewkes]

Martini - I had heard a year or so ago - that garbage links were given credibility by the algo if there was also a presence of high quality links in the overall set of links.

So what you are saying there is possibly true.

This however relates to well before panda.

[martinibuster]

I think you're right, that something similar relates to before Panda, but I'm not sure this is the same thing. I think this is a variation. I have heard that, too and have seen it. But that was in the context of big brand sites and established sites- and out of the context of ultra-competitive niches with rampant negative SEO, like gambling.

These are not big brand sites or sites with history. And this is not the case of the presence of many authoritative links. Only a very few.

[almighty monkey]

For some competitive SERPs, the formula certain spam sites appear to be using to exploit Google's latest algo is to buy one or two of the biggest most authoritative links they can find, then build an army of low quality inbound links from footers, WP templates, and forum spam.

The low quality links, standing on the broad shoulders of the paid authority links, are then able to lift the site into the top ranking.

Is this happening because the few authority links are enough to get a free pass on the spam links? Or is there a Panda/content algo loophole that is being exploited?

What do you think?

That 'Trust' is being included in whatever Goog's PR algo looks like these days.

So, lots of PR but no trust = Spam. If you have a couple of what Goog judges as 'high quality' links (even if these links are in reality paid - It's still an algo with holes, after all), you have greater 'trust' and therefore Google is more likely to give you the benefit of the doubt, and you start seeing the more traditional rise followed by a slow descent from spammin

Has anyone considered that onsite factors might contribute to 'trust' as well? If anyone has ever done Adwords, they'd know that Goog looks at things like an existant phone number, a privacy policy, a physical address and the like as part for their scoring algo there. I wonder if these give you more leeway if you have these sorts of indicators?

[petehall]

One of the sites I notice that Google is disregarding my internal link text on the indented sub links on the listing (above 'More results from ...')

Instead of using the link text Google is now picking out the H1 tag of the target page (or could possibly the first line of text).

I'm sure this is quite new - has anyone else noticed this?

[tedster]

Yes, Google is quite inventive these days in creating their own link text for all kinds of applications, even including sitelinks. They trust their own algorithms bout what's going to draw the click more than the webmasters - and they will change the link text depending on what the original query terms were. That's to the website's advantage when they get it right.

[Planet13]

@ MrFewkes:

I gave a site a superb quality PR4 link a few weeks ago. Index to Index - inparagraph - "keyword1 keyword2 here" - only about 3 other outbound links on the source page.

What does the backlink profile look like for the PR4 site from which you got that link?

I ask because I was thinking of buying a link in a directory from a PR5 site, for example, but then I looked at the backlinks to that directory (at the behest of martinibuster and wheel) and realized that all the inbound links to that PR5 directory site were from lots of SEO directories, VBulletin profiles, and the regular spam links.

I decided that I could probably spend the $25 on better things (like beer).

[kellyman]

Been doing a bit of testing throwing all kinds of links at my site and nothing happened, have some real juicy links from authorative sites too but no movement.

I then went through my internal keywords and where i just published a lot of data it has diluted my top keywords, my top internal keywords dropped down, i adjusted my text on pages so now one the keywords i am trying to rank for have moved back up according to WMT, my ranking has gone from a very stubborn position 14 to position 3 7 days after i made the changes, this is a keyword where its not in title or my description but just casually mention a few times on page of 700 words with 52m competing pages

Im going to change a few more things over this weekend and see where i end up, it could be a coincidence or the on page factors are being given a lot more weight as previously mentioned

[lfgoal]

I fear the age of a link is of paramount importance - moreso than the trust / PR / whatever reputation of the link giving site.

Wouldn't that be a good rationale for doing what it takes to get those high quality trusted links NOW so they can begin "aging in the wine cellar"?

[lfgoal]

crapcrapcrapcrap-crappity-crapcrap

Wowzer Wheels - take it easy.

In the new paradigm, the only movement obtained from crap is past the bees wax and into the sewer line. So saith Goog.

[martinibuster]

crapcrap-crappity-crap is king in Google, right now. Once they fix it I will take that back. But right now these are the oddest and most disappointing Google SERPs I've seen since Update Florida.

I cannot say that crap doesn't work, that would be denying what my eyes see in the SERPs. Right now crap works and Google's SERPs are full of crap sites.

[wheel]

Absolutely. Some version of garbage almost always works in Google. If not this,then that.

In my industry, people are polarized over product. Consumer advocates rave over one type of product or another and if you buy the other, you're getting screwed. What kind of product is right all the time? That's the wrong approach. I advise my clients on all the product choices, educate them on the differences, make a recommendation, and then they tell me what they want.

This is the same thing. Garbage works. But it doesn't work for ever. 'Whitehat' works too, and has a greater probability of lasting longer. Saying white hat is the only answer is provably wrong - just look at the serps.

But we need to educate ourselves on the risks. Don't be taking the easy route and then complaining when it's not permanent. If you're going to run a garbage site, you need to be ready with another one right behind it for when Google uncovers you. That also holds true for sites that may only be marginal - if you're too close to being the same as everyone else, there's precious little guarantee that your site will be around forever.

[seoN00B]

For me links are not powerful nowadays.

Imagine i was outranked by just one LINK from a PR 2 farm Drupal site while i have 20 quality back links. =P

[walkman]

Links are not going away any time soon. I am in panda land so nothing will work meanwhile but I joined two directories. I read from several places that they are sort of G approved since they supposedly manually review the sites for a nice fee. At least they can help with site categorization.

Measuring "engagement metrics" like Panda is supposedly doing might work but trusting them causes...what already happened. They are just too many sites, too many categories, too many in-between /hybrid sites to have a standard metric to compare sites. Especially if you are going to use that to determine if a site survives or not, which a SE with 70%+ of share can do. So eventually, after destroying many businesses, Google will come to its senses. Links, with all their faults will be with us for a long time.

crapcrap-crappity-crap is king in Google, right now. Once they fix it I will take that back. But right now these are the oddest and most disappointing Google SERPs I've seen since Update Florida

I read the Floria thread here, but a month after it some people started to come back. I wonder what statistics G is looking at, especially considering the upheaval they created.

[Freedom]

After 10 years and lots of different link approaches, my conclusion, my 2 cents is to play it super safe.

1. Forget about keyword anchor text to interior pages. It'll boost you for awhile, then you are forever punished.

2. Forget about links that link companies get you. They are junk links anyone can get and they do not have your best interest at hart, only in selling you links.

3. Build all your links to your home page with the domain name as the anchor text, and hope the benefit flows down into your site. Google can't punish a site that has it's domain name as the anchor text. That's ludicrous.

4. ONLY get links from SUPER relevant sites, even if they have a low quality score and look like junk, if it's relevant to your topic, that means the visitors at that site are interested in your site, and google will see that.

5. Get links from authority websites. Newspapers, .edu, organizations. If it's a hard to get link and a recognized authority, it will pay off.

The harder it is to get a link from a website, the more value it will have. Nothing is easy. There are no free lunches, and gaming the system is a cheap trick that won't last.

Getting the right links is a long, painful process and test of endurance. It should not be done by anybody else but you. The harder it is means that you are doing it the right way and it will pay off.

And seriously, forget about keyword anchor texts. Keep it to the domain name and let it all filter down and it will eventually lift those interior pages up in the serps, slowly, gradually, naturally, and when Google sees fit to do it.

Now, here comes the... "Not true!..It worked for my site that way so it must be true for everybody..." - type comments.

I've learned the hard way on some things, and I'm sticking to my Keep it Super Safe (KISS) plan.

[Crush]

Spent a few grand in the last month on really nice blog links, like alexa under 100k in the sidebar, decent pr. No shift in rankings, if not slightly worse.

Last year this would have boosted us to page one, now....meh

[Planet13]

3. Build all your links to your home page with the domain name as the anchor text, and hope the benefit flows down into your site. Google can't punish a site that has it's domain name as the anchor text. That's ludicrous.

That does seem like a pretty safe way to go.

Maybe building links to internal pages with anchor text like, "click Here," or "Buy Here," or "On this site" are still ok.

Or, use a phrase that is found on the target page. Like if your page has some particular statistic on it, even if it is not the main topic of the post.

I say this because I have one page that ranks really well, and for a long time I never figured out the reason. Then I realized there was a link from a site that was totally natural (had no idea they were linking to us) that was from a subject that was totally off topic, and they used "buy them here" as the anchor text when they linked to us.

4. ONLY get links from SUPER relevant sites, even if they have a low quality score and look like junk, if it's relevant to your topic, that means the visitors at that site are interested in your site, and google will see that.

While I can't argue with relevancy, I also am now double checking the backlinks to the site I wish to receive a link from.

If the site is relevant, and the backlink profile looks relevant and clean, then I think it will have more value.

On the other hand, I have seen a couple of niche-specific directories with a toolbar PR5 that would link to / list us for a modest fee (like US $20).

Even though the directory is on topic, their own backlink profile is crap. I wouldn't really want all their backlink "baggage" to be associated with my site.

[Freedom]

All my best ranking internal pages were from natural links. When I tried to "copy" the effect myself by "getting" links - it blew up in my face.

That's why I build to my home page, let it all filter down, and hope find someone finds an internal page they like and they can link to it any way they want, and they do so in some weird ways, but it works.

[hannamyluv]

yet, links are what makes the web work

I know I am late to this, but actually - links are not what make the internet work anymore. The flow of traffic through connections is what makes the web work. In the past, G did not have insight into this so links were the only way to make a "best guess" at connections on the web.

That has changed drastically in the past few years. The obvious is that G has insight into 80% of ALL web traffic now. They don't need links to see where traffic and connections are flowing. Social media has also altered the landscape for connections. If you have a catchy/memorable name, a significant number of your social mentions will not even include a linking trigger like .com or www. I mean, when was the last time you said or typed ebay.com or flickr.com? And yet, they will get traffic from those mentions (because duh, even my Grandpa knows to add a .com on the end). Why do you think G is SO desperate to get into the social game? Because they have limited insight into FB connections and FB could make a VERY competative search engine, just based on what people talk about on FB.

Links are dead. Connections and traffic flow are the key, IMHO.

[Bennie]

crapcrap-crappity-crap is king in Google, right now. Once they fix it I will take that back. But right now these are the oddest and most disappointing Google SERPs I've seen since Update Florida.

I cannot say that crap doesn't work, that would be denying what my eyes see in the SERPs. Right now crap works and Google's SERPs are full of crap sites.

I cannot agree more.

Right now I have new sites with barely any content and a handful of hand built comment spam on exact match domains out ranking old sites with site links that have been ranking forever.

Old Trustrank PR is back, age is back. Parts of the SERPs look like 2003-2005. Facebook likes and fan pages are working well and also supporting 'social' media link twitter links etc.

Seems like the editorial knobs have been turned up but in a country like Australia, where the sources are gaming Google quite hard, it's really distorting the SERPS with irrelevant out of date stories ranking very well in the regular serps. They don't link out properly and no url fields in comments at all, so the networks just horde PR. Pretty much an amatuer SEO getting happy on trust juice, *coff* in house SEO's *COFF*

Seems to be the social factor of blog spam and on topic forums and yahoo answers picking up the slack. At first I really expected user data of page views and time on site to be a factor, or semantics, but seriously, I just think it's plain bull#*$!.

Never have I seen such an unpredictable update with such random results. News stories ranking for ecommerce product queries, that are out of date and not relevant.

In Australia, we have the major news sites all crosslinking duplicate stories, and it's working. Wouldn't be the first time someone had to step in and fix the Australian serps.

Another theory I've been strongly thinking about is the X factor of Eval team and editors doctoring the serps with trust and non-trusted sites. It's just too random to explain what is going on without assuming someone's playing games, be it Google, or some dodgy eval teams.

It's honestly beyond me at the moment and this is the first time I have been genuinely stumped by Google in a long time. Not only that, but the results are junk, seriously crap.

If Google wants to keep this up we may as well all start spamming the crap out of them fulltime and roll with the punches.

Well that was a good rant. Sorry if it was slightly disjointed, I am too :D

[edited by: engine at 8:35 am (utc) on Jul 19, 2011]
[edit reason] specifics [/edit]