How would you know if links were no longer important to Google?

[goodroi]

Ok, right now links are an obvious part of Google ranking algorithm (paraphrasing Google's John Mueller). Recently John Mueller also said that his friend's new site with no backlinks was indexed and gaining traffic from Google. I have personally experienced similar situations of websites performing well without significant backlinks. We also know that Yandex stopped using backlinks for certain serps. It is plausible that Google might move away from link data and instead use other metrics like monitoring usage via browsers, toolbars or isp data logs.

How would you know if Google stopped counting links and started stressing another metric for the rankings?

For example if they switched from link data to user traffic the serps would probably look similar. The better backlinks tend to drive real traffic. If you reverse engineered the top ranking websites you would see many good quality backlinks (which wouldn't be why they were ranking) that were driving high traffic (which could be the real reason they were ranking).

What do you look at when trying to decipher Google's preferred ranking signals?

[ganzojin]

Some time ago Matt Cutts said that Google tried to make an index without link metrics, and that its quality was fairly lower than the "normal" one.

I think that we will never "notice" a sudden change, because they never will "turn off" istantly link metrics: they are currently becoming less important or, to be more precise, more sharply judged imho.

[aakk9999]

Some time ago Matt Cutts said that Google tried to make an index without link metrics, and that its quality was fairly lower than the "normal" one.

Good memory, ganzojin. Yes, it was about a year ago and we discussed this here:

Matt Cutts: Search Results Look Worse Without Links As Ranking Factor
Feb 19, 2014
http://www.webmasterworld.com/google/4646820.htm [webmasterworld.com]

On the other hand, I do think Google may be experimenting with it in narrow niches (or within a sub-niches of more competitive niches). I had a few cases where a new site with no links (or only one nofollow link) starts to rank for a low competition keyword (which was not a long tail kw and where there were other sites going for it).

[netmeg]

I've never paid a lot of attention to links. First because I didn't know how to do it, then after a while because it didn't seem to me that it could scale and keep its quality, and finally because I was too lazy. Turns out, that wasn't as bad a move as I thought.

I think if you're unique enough, and people engage with your site, you'll still do fine without a lot of high profile linkage.

But if you're just another site doing essentially the same thing as a thousand other sites (particularly in a competitive niche), well. You really will need the links.

So to answer the question - I probably wouldn't notice if Google stopped paying so much attention to links.

[austtr]

if they switched from link data to user traffic

That would just make the collateral damage of the past into a permanent fixture from which there can never be an escape. So many worthwhile sites were slammed by Panda, Penguin, authority bias, top heavy filters, and all the other penalties of recent years... and the nett result was huge numbers of sites lost traffic.

Imagine a scenario where "now that all those sites have lost most of their user traffic because of our need to protect our own commercial interests, lets make user traffic an important criteria for ranking sites".

Yep... that's the way to foster good relationships with publishers.

[RedBar]

I'm very much aligned with netmeg, I've never once asked anyone for a link even when link rings were popular.

But if you're just another site doing essentially the same thing as a thousand other sites (particularly in a competitive niche), well. You really will need the links.

I agree with this statement however I seriously wonder whether G is really able to sort/filter/judge what is quality relevant or not? Looking at their SERPs it wouldn't seem so to me.

[ronin]

I think if you're unique enough, and people engage with your site, you'll still do fine without a lot of high profile linkage.

I'd tend to agree.

The orthodox wisdom in SEO circles is that off-page factors exert a lot more influence than on-page factors.

However, to what significant degree can SEO agencies upgrade on-page factors? They can tinker with headings and add footer navigation, sure. They can restructure the site and rewrite browser URIs with .htaccess.

But can they tell the company to ditch whatever cruddy CMS that company is using and switch to a CMS which outputs clean, lean, semantic markup? Probably not.

So... isn't it in the interest of SEO Agencies to keep re-iterating that off-page factors remain a lot more influential than on-page factors?

[n0tSEO]

I agree with netmeg and ronin.

Experience tells me it's possible to rank a site pretty nicely with only on-page SEO complemented with incoming traffic from social media and relevant blog comments.

At this point, that's how it's working for a client who hasn't built a single link over the past year. The website may have 5 backlinks at most and they are not even from authority sites.

So I believe backlinks may add to the rankings, but they don't make the rankings.

[Jez123]

This is very interesting. Surely this is as easy to manipulate as links? If not even easier? I recall that there are loads of places that you can "buy" endless traffic. Does google know if it's quality traffic or not? I suppose they do from WMT data etc?

[n0tSEO]

Sorry, Jez123. :) I didn't explain myself well.

I meant to say that on-page SEO was enough to rank a site, and that the site thrived on social media and comment traffic that complements traffic from Google.

[netmeg]

Does google know if it's quality traffic or not?

I believe they do. In fact, they smartprice AdSense for traffic issues. I don't know what signals they use, but I'm sure they use them. If I have a site targeted to a specific US state and suddenly I start getting a ton of traffic from India, they're probably going to think that's not good quality traffic for my site.

I always assume Google profiles everything. Because they can.

Bottom line, if you held a gun to my head and said I could have only ONE of the following: great inbound links, perfect on-page SEO or stellar user engagement, I'd pick the user engagement every time.

On a related note - please don't anybody hold a gun to my head.

[ganzojin]

As a SEO consultant myself im more and more inclined to just propose my clients to just work together with me on a very informative and spider-ready information structure.

To structure logically the site to answer users questions, and to do it using several layers of taxonomy deepness. It works like a charm for non-spammy querys ( the ones were Google Zoo seems absent, and spam tactics reign )

[Planet13]

@ ganzojin

...and to do it using several layers of taxonomy deepness...

Could you clarify / elaborate what you mean by that?

Thanks in advance.

[rish3]

This is very interesting. Surely this is as easy to manipulate as links?

There are, in fact, at least three services that sell visitors that act "real", specifically for SEO purposes. Like visitors that search google for a search term, click specifically on your site, and navigate internal pages. Supposedly, with realistic user agents, pause intervals, dwell times, etc.

I've never used them, so I don't know how well the services match up to their own claims.

[ganzojin]

@Planet13

I meant to say that to optimally build a site you should structure the site so your taxonomy are vertically aligned as topics

example:

site selling design furniture
- kitchen furniture & bedroom furniture etc. etc.
- kitchen chairs & kitchen table etc. etc.
- tags like : wooden chair, plexiglass chairs, rounded chairs etc. etc.

i know it seems pretty easy but, in fact, i see a lot of mistakes that cost dearly in terms of SEO

[Nutterum]

@rish3 - there will always be ways to cheat the system. Remember the days of xyz Page Rank quality links for x USD ?

I think the only plausible way for Google to minimize the impact of backlinks on SEO and move on other signals is for Google not only to gather and parse the data from Analytics to judge your website SEO rankings but enforce all websites to install Google Analytics and similar other "ranking tools". And even then the game can be rigged. (imagine paying for e-mail direct visitors that act naturally and then go to google to search for similar product browse a few pages but interact with your page the most)

All in all, sooner or later in my personal view, it will not matter how many hundreds of thousands of backlinks you have or how many directories you are listed at, but rather how many high quality/trust links you have. And we all know that those can be tricky to get especially in the local SERP space.

[n0tSEO]

it will not matter how many hundreds of thousands of backlinks you have or how many directories you are listed at, but rather how many high quality/trust links you have.

That's already happening. Quality over quantity, and it adds to the ranking benefits of good on-page SEO.

I think the only plausible way for Google to minimize the impact of backlinks on SEO and move on other signals is for Google not only to gather and parse the data from Analytics.

That would work to a certain extent, but not so much, because not every website uses Analytics. That would wipe out millions of sites from the index and impoverish the DB. Not sure Google may want that. :)

[Nutterum]

@N0tSEO - I do believe that if they make an official announcement everyone will make an account, whether they want it or not. Those who opt-out of the service are not services Google would like to work with in the first place.

As for my first comment about the backlink flooding (white hat of course) it might be happening but for the most part non-sitewide backlinks are still very effective tool to rank on top on 3 keyword phrases and other long-tail variants.

[minnapple]

The way I see it.

Google would prefer to use more user data in their search algo and lessen the amount inbounds plays in the algo.

That works if Google has x-quanity of user data for a certain type of search.

If they do not, it is a entirely different situation.

So what is the best way to get the value of inbounds?

Check your Analytics for high value, low traffic terms that perform well in terms of conversions per sessions, and back those up with one or two inbounds.

[Wilburforce]

That's already happening. Quality over quantity

Not "already happening": it was ever thus.

[martinibuster]

I can only speculate but I believe the concept of "non-manipulated freely given links" explains netmeg's situation where netmeg doesn't directly encourage inbound links but I believe what netmeg may be doing is indirectly encouraging those links by simply giving site visitors what they want. What do visitors want? I speculate that they want information, they want it fast, they want it in a manner that is platform independent (tablet, phablet, phone, desktop, regardless of OS).

Another member in this discussion mentioned social media attention. That's another way to indirectly encourage "non-manipulated freely given links." My experience has been that those kinds of links have always worked well. They continue to work well. What's changed is that some of the other manipulated links don't work as well as they used to. But old fashioned freely given links, the signal Google's always been after, those still work. I am confident of that because that's my experience.

Q: How would you know if links were no longer important to Google?

A: One clue is that there would no longer be announcements of manual or algorithmic actions targeting link schemes.

[wheel]

Not "already happening": it was ever thus.

No, it wasn't.

Only in the last 2-3 years has Google figured out how to reasonably well define a decent freely given link. I think that's why many sites are struggling, and why brands are dominating. Links still count, but now only really really good ones count.

I'm sure link schemes still work on a limited basis - there's always exceptions. but it wasn't that long ago that link schemes worked on a broad, public scale. Nobody remember digitalpoint? Directories? And any of a million other schemes.

I don't think Google needs to get rid of links anymore. They've got it reasonably well figured out how to determine a decent link, they've got half the market crazed about the very thought of giving someone a link because it might damage their site, and the other half of the market paralayzed from even asking for a link because it might be seen as unnatural.

I've tightened up my link building noticeably recently, and I've always been more concerend about quality and relevance than most SEO's. Even then, where I would maybe target a dozen links a few times a year, now I might target a handful a year. It's that important to cherry pick quality.

[Wilburforce]

@wheel

No, it wasn't.

Are you saying that link-quality has not always been a factor?

I am not talking about how well Google assesses link-quality now compared to five years ago: I am saying it has been a ranking factor much longer than that.

My point is that a few high-quality links have always been worth more than hundreds of low-quality links. The measures Google uses to assess quality may have changed, but that fundamental fact hasn't.

"I've always been more concerend about quality and relevance" implies that you, too, have always thought so, so I am not at all clear about what you are disagreeing with.

[EditorialGuy]

Is it possible that Google treats links in different ways for different broad categories of sites?

It's easy to imagine Google thinking that links to an informational site are a reasonably strong indicator of that site's quality, because links to informational sites are (or ought to be) citations. An informational site of decent quality should attract "freely-given editorial links" over time.

On the other hand, links to an e-commerce or thin affiliate site may be less meaningful, if only because such sites have a stronger economic incentive and direct ROI for buying or cultivating inbound links. Also, it's probably tough for commercial sites to attract significant numbers of "freely-given editorial links" unless they're huge sites like Amazon or Expedia that get featured in the media regularly, so the presence or absence of such links may not say much about a commercial site's quality.

Finally, there must be vast numbers of links that can simply be ignored. Our editorial site gets hundreds (sometimes several thousands) of inbound links every month, and most of them are automated links from sites that are completely worthless. Google may not be able to tell every piece of #*$! from Shinola, but the most obvious dreck should be easy to identify and toss.

[martinibuster]

Is it possible that Google treats links in different ways for different broad categories of sites?

I did some research with Link Research Tools and found that what passes for normal in one category isn't so normal in another category. Inbound linking patterns can differ according to topics and even within those topics. For example, a strictly local site with one office serving a single city can tend to accumulate a specific pattern of inbound links that is vastly different from another site in the same topic that has offices in multiple cities. Statistical analyses can be graphed, for example, along the percentage of inbound links and the number of pages in a site (which will reveal natural patterns specific to a site serving a single city versus a site serving multiple cities, both sites within the same topic bucket). That will create a baseline for normal linking patterns specific for a topic. Anything outside of that baseline could be considered abnormal. To compensate for natural "unnatural" patterns these can (and often are in research papers) combined with the results of on-page factors, for greater accuracy and to minimize collateral damage.

Research papers by search engines and university scientists have shown that categorization of sites can improve algorithms for finding spam and identifying authoritative non-spam sites.

Also, it's probably tough for commercial sites to attract significant numbers of "freely-given editorial links" unless they're huge sites...

It is tough. But there are strategies that can be used to overcome that difficulty. Think about the last SMB you may have purchased from online and why you did that. Once the dross is removed the remaining link signals can be used for ranking and relevancy purposes regardless of niche (since all niche topics have varying patterns for what passes for so-called natural).

[Nutterum]

In local, link spamming is still a thing. Not with the same strength it used to be, but it is still there. More often than not I see websites and listings pop-up that have several thousand or more backlinks from all corners of the known internet - and it works. Might not work for every niche or the really big cities(think NY size) where local brands still dominate the SERP but pretty much everywhere else its the same old deal.

Oh and one more thing, this is SEO for the US. In Europe - for some reason(though I have few ideas) this strategy (as in backlink stuffing) does not work in the slightest. Sites with 200 backlinks are outranking other websites with 10k+ backlinks - go figure.

[martinibuster]

How would you know if links were not important anymore?

[EditorialGuy]

How would you know if links were not important anymore?

You wouldn't, unless Google said so (as Yandex did when it said it was no longer using links as a ranking factor for commercial searches in the Moscow region).

[Nutterum]

I think there is an easy way to check this by the way. If you are willing to suffer the loss of SERP positions for some of your keywords. Hell I did this by mistake.

I have several good websites with relevant dofollow links linking to one of my landing pages. However at the backend of the website I mistakengly changed the year from 2015 in to 2016 . Changing this is a big deal as it screws my tags and ult paths as well as dates on the page etc. etc. It more or less changes a good chunk of the entire on-page SEO. The only thing related to 2015 is the content and product listings.

Two weeks later I noticed the change, looked at the SERPs to see if anything happened and noticed that I got -0.6 average position on the search term + year queries. Change that can very well be a "noise" change not related to the change I did.

I believe that the good anchor text and relevancy of the backlinks kept my landing page afloat. If it was on a page with no partial backlinks pointing at it - Google would have placed me on the bottom of the SERPs almost immediately.

So backlinks are important but not the way they used to be. Its not about filling a bucket with water to say you have a full bucket of water, rather than having a bucket full of water guarantees you have a bucket in the first place...hope you got the notion.

[dethfire]

We'd stop seeming total crap sites rank well that oddly have great link profiles.