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?

[aakk9999]

This is the way I saw it happen in a medium-competitive SERPs with a site that had a good user engagement:

- a new site, good on-page SEO, no links.
- after the launch, the site ranks on top of page 2 for the main KW, does not rank well for secondary keywords.
- over the next three months the site slowly creep up to page 1 for the main KW. The secondary KWs follow ranking increase, but the "distance" between the main to secondary KWs is still approximately the same
- one year on, the main KW is now top 3 positions, the secondary keywords some at the bottom of the page 1, some on page 2 for their searches
- the number of secondary KWs driving traffic keeps increasing
- end of the second year, the main KW is #1, secondary KWs are all on the first page for their searches. The site started to show for searches for the "BIG" KW on the 2nd page (BIG KW is in a wider niche, much more searches, and initially this KW was not targetted. It was seen as being too hard to achieve ranking for because of competition from strong established sites)
- end of the third year, the site is #1 for main KW, #4 for the "BIG KW" and #1 - #3 for secondary keywords.

Increase in traffic followed increase in ranking.
No link building at all, however some forum links/fb links and similar picked up on the way as people talked about it / referred to the site
The content was added infrequently, but when added, pages were good. In fact very good - visitors bookmarked them and kept returning back to them. No fluff. No walls of text text.
As the site matured, its ranking and traffic spread out of the niche the site had initial focus on, onto a wider niche.

So there must be this initial traffic that would allow the measure. But it can be done in a different way - target a smaller tighter niche and then with the time spill out of it into a wider defined niche. I am not sure the results would be the same if we targetted a wider niche straight away, which was much more competitive. I guess the site would be too far down in SERPs then to get any measurable traffic.

[netmeg]

That would result in the situation that Google's SERPs would rely on the SERPs of other search engines.

You don't think they take information about how users interact with a site from Google Analytics?

You mean from links?

Yep.

An established site can do that. A new site can't. This might skew the rankings toward established sites and against newer sites. It might also skew the rankings to sites that can pay for advertising based traffic. OR it might skew to sites with aggressive link building programs including paid links, advertorials, guest posting, reciprocal links- basically links and advertising would be back on the table, I think...

Depends on how you define 'direct' - if you generate the buzz, you will get the type ins. Just launched an ecommerce site for a client who had done nothing but catalogs for the past 50 years. They put it in the catalog and on their fax list (yes, they use faxes, even in 2015) and boom - instant type in traffic.

[Nutterum]

@aakk9999 - when a read your post, I immediately thought - is he talking about my niche?!

I am in a middle-competitive pure organic niche (b2b) and the main Keywords are somewhat reachable like <niche>+<year>, <event>+<year>+<country/city> . However only recently (and through more than 120 pages of revised and/or new content) we began creeping on the generic niche keywords (which is good because the content is relevant!) . The more broad our reach got the more natural backlinking I began to see, there were months with over 50 purely onatural backlinks.

Our main traffic is still direct as we have a big network of subscribed customers and good e-mail campaign set up, but more and more organic is catching up and people convert from organic to direct visitors (via bookmarks I assume).

Once Google saw this, since Jan. 2015, we skyrocketed on some more competitive keywords - rank #3-#5 on some content verticals.

To make it even interesting the SERP "leaders" have close to 300k backlinks. We have a few thousand - and we are still competitive. Granted the strength of our links is way more relative, than the shady forum comments and what-not I see the other websites were doing a few years ago, but in the end links for the sake of links is not that relevant for google - case and point.

[toidi]

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

the results would be relative?

[Nutterum]

OK Guys,

We did a quick experiment with a friends blog. He had around 100 links, nothing big and usually gained traffic through social and some niche keywords revolving around book reviews. We removed every single link. 5 days later SERPS dropped 0.5 on average with one of the more popular query dropping 10 places.

So are links important? - YES! Can you do without them in general? - YES! Do they loose value the more you hoard them ? YES!

I know these are generalized assumptions from a small blog occupying a small niche but in the end it is a pure demonstration of what happens when you remove all links. If the content is good and people like it on social - Google sees it.

[Wilburforce]

We removed every single link

Do you mean that ALL backlinks were self-posted? How, if not, did you remove them?

[martinibuster]

We did a quick experiment with a friends blog.

Thanks. Good information. However I just want to clarify that there's no doubt or question that links are important.

The topic under discussion is,how would you know it if Google stopped counting links, and what metrics would Google use in the place of links?

[Nutterum]

@wIlburforce - 80 or more of the links were placed by myself on various directory type of websites and such. The rest I have contacted the few local news websites to change the link. Since I do work with them on other projects they replied quickly and changed the links knowing full well it is for the purposes of the experiment. In the end I believe we managed to remove all backlinks associated with the blog.

@martinibuster - I understand the topic and I have discussed it quite a bit, but my point is that in order to weight what Google will use if there are no links pointing to the site, I had to make an experiment where Google literally had no links to work with. And the change was not that significant because in the end, do-follow links are just one (even though big) signal.

I need to clarify, that I did not remove the no-follow links (eg posts from social media websites and such) only do-follow links and I suspect that Google segregates no-follow links in several categories and is using them for trust and thus ranking.

A good example was a small tools info website a colleague of mine works on. They managed to get 3 links from Wikipedia and the results were very positive on the SERPS in just two weeks time even though these backlinks were no-follow (as all links in Wikipedia are as far as I know)

If we go into vacuum and make the statement "How will Google rank websites if they did not take into account _any_ type of links" then we can speculate until the end of eternity, something I am not too interested in.

What I am interested in answering is "How will Google rank websites if they acknowledge no do_follow links on the websites?" something I think people are more interested in, as it directly impacts link building, spam and the vast majority of black/grey-hat SEO techniques.

[Nutterum]

Arriving just in time for our discussion, what do you think about the new research paper by GOOGLE wanting to introduce Knowledge-Based Trust (KBT) as alternative to link-based SERP ranking algorithm?

[Wilburforce]

@Nutterum

I'm not sure how relevant it is to this particular topic (I suppose it might form the basis of an alternative ranking formula if it worked), but in my own view it is a long way from viable.

[webmasterworld.com ]

[Kratos]

Backlinks will always be the foundation of Google to rank documents (i.e. pages). Unless they completely revamp their algorithms (I don't see this for at least 5 years in time).

The internet was founded on links (how do you think we browsed AOL or Yahoo Directory back in the day?) Links have always and will always be a good gauge of a vote of confidence, but just like with any elections, they can be bought or deceived.

Personally the only time when links are not need to rank on the top 3 are with low competition keywords or when the first page of a medium competition keyword has all the ranked pages demoted due to spam (and the site with no links thus increases its rank). I've seen the latter happen a lot of times when a webmaster thought that his increase in rank was because his site was awesome and had no links. Remember, it's not just what you do, it's what your competitors are doing, and ranking is highly dynamic.

It's like that dude the other day on a G Hangout complaining to John Mueller that he had kept his site super white hat, yet he wasn't ranking (while his competitors were using the churn & burn method to occupy the top 3 real estate). This guy went and removed all good links he had (including a BBC link) to let the G bot know they're as white as snow. No wonder the poor guy's site wasn't ranking (well, his on page SEO was terrible too, but he had no powerful links whatsoever, so even the spammers blasting their GSA spam were out-ranking him).

Great on-page SEO but no links = maybe you will rank somewhere on the low end of Page 1 to Page 2

Mediocre site with great links (from real sources) = you will rank on the top 3

The above is somewhat simplified, but it's what it comes down to in my opinion.

[martinibuster]

what do you think about the new research paper by GOOGLE wanting to introduce Knowledge-Based Trust (KBT) as alternative to link-based SERP ranking...

Are you talking about the science news site or did the SEO bloggers catch on? I had given the SEO bloggers a two to three month lag, lol.

Here's the deal about KBT:
1. It's not a replacement for links.

2. It's an experiment with interesting outcomes for identifying accuracy. If you read the PDF download, the actual paper, you will find that what they call trust is explicitly defined near the beginning of the paper as accuracy. It's somewhat confusing though because the word trust is used more often than accuracy.

3. The study stated that more work is needed because, presumably, the algorithm is not scalable to the web. What they actually stated was that their methodology used a lot of resources. Between the lines I take that to mean it's not scalable in its present form.

4. I didn't check on all the authors but at least one of those authors works in the Knowledge Base division of Google. So it's possible that the study might have more implications for the knowledge base part of Google than it does for an actual Internet wide algorithm.

Most importantly, any author reporting this as an imminent algorithm change or something that Google wants to introduce is either overstating the study (lying) or is ignorant.

[Nutterum]

To further clarify, my question was more in the lines of - can you see this as a possible algo signal down the line? Because I believe that Google are trying to avoid links as a ranking signal in general, the point is that they can`t. So the alternative is to approve only super-trust link signals with other signals and algo "approvals" to rank a site in the future. Thoughts?

[Kratos]

Nutterum, Google already does a pretty decent job at ignoring links that are thought (by Google) to be deemed spammy or manipulated. If not, you would not have those blackhat noobs with their shiny GSA SER crying every time Google refreshes Penguin (and even Panda, as the spam target pages drop in perceived value). That's not to say that spammy scripts like GSA SER can be used effectively to fool the Googlebot.

Approving only super-trusted sites' links would not be good either. First, what do you define as trust? High PR? A database from Google? Do you know that many perceived-to-be-trusted sites have been caught selling links and this crap continues to this very day and for the foreseeable future? Google is very aware of this all.

It's very simple. Google will continue to cull the spammy links as it improves their algorithms. However, links from sites that are not perceived as spammy (from a new blog on cats to the CNN) will continue to count just as they've always done. Not only that but links are still and will always be (unless major revamp which I don't see happening yet) the most-important signal when ranking documents.

Links are Google's foundation and their Achilles heel.

[martinibuster]

Well said, Kratos. Google has made it more difficult to manipulate links. Some black hat noobs may dispute this but they are obligated to dispute it to justify sales and membership in their link selling, link trading, and content sharing networks- until they're closed down by Google and then another big mouth noob and her/his affiliate cronies step in to take their place. The truth is that even an off the shelf tool like Link Research Tools can unravel an entire link selling/link trading/content sharing network. All the crowing about stealth, hiding ownership and using different IP subnets is BS, smoke and broken mirrors. The truth is that the most lucrative SEO business right now is not actual SEO but reconsideration requests.

Let's get this out of the way
Here's a thumbnail background history of Google and links. The history of Google's relationship with links has been one of refining the link signal. So almost from the beginning Google was actively depreciating links in footers, links from irrelevant pages, links with surrounding text that said Sponsors, links from reciprocal link pages (that kind of blew up when first implemented round 2003/2004). In the early days, mid 2000's I suspect, they implemented statistical analyses and that changed the landscape tremendously. In recent years Penguin has made enormous strides at combatting manipulative links resulting in a visible shift in the SEO industry from performing SEO to non-link building activities like Social Media, PPC, Local Search, and reconsideration requests. Virtually nobody wants to build links because there's so much risk involved. It was difficult to do before Penguin. It's so much more difficult now.

[fathom]

The whole premise of links or not is dumb.

If your website is built on content then you won't notice any difference if your website is built on some scheme using links you will most definitely see a difference.

Does it really matter if you don't see any difference?

[Nutterum]

@Kratos - I fully understand your point but the discussion is are links important and how rapidly do they decline in value and by proxy are they needed at all today or in the near future. And I firmly believe that sooner rather than later websites with no or small backlink profile will have no trouble competing with websites that have millions. Hell I already see this in the hotel booking sector where small family hotels with 10 backlinks on some god-forsaken blog but with fresh monthly content get above booking/expedia/hotels and the rest of the big dogs in local and still do fairly well on the more broad <keyword>+<country>/<year> phrases.

No one is disputing that links are not vital anymore as we have have all seen a million times that a good dofollow link from a relevant source can set the average position arrows to green. What I and some other participants are saying is that slowly but surely you can do well without being "obligated" to do SEO in its traditional form and I would not be surprised that in a few years what people perceive as SEO will have nothing to do with what we are doing today.

My prediction - in 3-4 years there will be an enormous demand for website coders with online marketing/social networking/content writing skills as this will be all that is needed to have a working and well ranking website.