Forum Moderators: Robert Charlton & goodroi

Message Too Old, No Replies

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

         

goodroi

11:31 am on Feb 9, 2015 (gmt 0)

WebmasterWorld Administrator 10+ Year Member Top Contributors Of The Month



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?

wheel

7:57 pm on Feb 23, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



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

That's exactly what I'm saying. The old digitalpoint network is a perfect example of how link quality didn't use to matter.

I think link quality is exactly what's going on with the rankings these days though - I think Google has a very good handle on exactly what a quality link is. I think that ability alone accounts for much of the 'brand' talk that goes on. Not social media, not all the other signals people are talking about - real, absolute top-quality links.

It's a pretty good working hypothesis anyway. It's the basis for everything I do these days.

I have some suspicion that on page SEO, in some cases, may help as well even in competitive niches, but don't have a lot of data or experience there. In fact that's probably something I could use some further investigation on.

rish3

8:48 pm on Feb 23, 2015 (gmt 0)

10+ Year Member Top Contributors Of The Month



I think Google has a very good handle on exactly what a quality link is. I think that ability alone accounts for much of the 'brand' talk that goes on.


Depends on your definition of high quality. At least in the niches I frequent, there's a ton of manipulation at the top.

Google does seem to have figured out how to combat the really low end spammy links, poorly done blog networks, etc. I don't see evidence that they are interested in doing anything about the high-end paid links, even when it is painfully obvious.

wheel

8:58 pm on Feb 23, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



There's always exceptions. Look at the serps where the brands are. In my niche, there's almost nothing left but brands - hardly any manipulation.

Except for the one site owned by a national media empire; they've got a commercial site they link to from all their sites. It'd be nice if Google figured that out and slapped them into last week - but I don't see that happening soon.

rish3

9:24 pm on Feb 23, 2015 (gmt 0)

10+ Year Member Top Contributors Of The Month



Look at the serps where the brands are. In my niche, there's almost nothing left but brands - hardly any manipulation.


Ahh, yes, I see some of this too. Of course, "brands" benefit from the "Google Brand Dictionary" :)

Trades links with other businesses
small business -> link network / manipulation -> penalty
brand -> co-marketing -> win

Footer links to other owned sites
small business -> unnatural linking -> penalty
brand -> vertical segmentation -> win

Exact match anchor text on mommy blogs
small business -> guest post -> penalty
brand -> community outreach -> win

repost interesting content from another site
small business -> duplicate content -> penalty
brand -> curation -> win

goodroi

10:48 pm on Feb 23, 2015 (gmt 0)

WebmasterWorld Administrator 10+ Year Member Top Contributors Of The Month



Let's be careful to stay on target - How would you know if links were no longer important to Google?

rish3

11:07 pm on Feb 23, 2015 (gmt 0)

10+ Year Member Top Contributors Of The Month



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

Well, that is the very essence of SEO, right? You're in the business of reverse engineering a black box that you can't open...and further, a black box that gets its internals adjusted on a very regular basis.

So, you experiment, accept the fact that you can't really isolate variables, and make your best guess.

In this case, it seems logical that traffic / engagement are the only thing that could really displace backlinks.

If I suspected user interaction was now the #1 factor, I would take a currently ranking website and tweak it such that it is a horrible experience to visit. Perhaps some javascript to remove the ability to click on links, done in a way that isn't visible to googlebot. Then wait and see if rankings plummet.

seoskunk

3:26 am on Feb 24, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



Links determine how the big boys rank still but authority determines who the big boys are. It's not right , its morally devoid but hey that's the current system on google.

I personally have thrown in the towel on seo and am retraining. There only so long you can bang your head against a brick wall.

Oh and agree with above, without discussing brands and other factors the conversation is pointless.

Nutterum

8:25 am on Feb 24, 2015 (gmt 0)

10+ Year Member Top Contributors Of The Month



Hasn't anyone made the experiment to just cross one of his landing pages by disavowing all partial backlinks pointing specifically to that landing page and see what happens. This will directly answer the question of knowing whether your backlinks are of any importance to Google Ranking.

To have a control group change the content of another landing page in the same vertical but keep the backlinks.

I bet 10 bucks that the landing page with backlinks will see almost no change in short term (10-14 days) while the disavowed page will sink like a rock.

martinibuster

12:36 pm on Feb 24, 2015 (gmt 0)

WebmasterWorld Administrator 10+ Year Member Top Contributors Of The Month



The way I see it, links serve two purposes in the algorithm.

1. Links establish authority
The description of this function has always been that links tell us how important a site is.

2. Links establish relevance
The topic a link originates on, anchor text, and the chosen page linked to sends a signal of what the linked site is about. Lacking anchor text there is still the context of wording around the link and the title tag.

Any conversation about whether links are important or not is pointless without discussing how the algorithm determines a web page's importance and relevance.

If links are removed from the algorithm then you must think in terms of what defines a web page as important and what makes a web page relevant. Importance and relevance are ranking factors. Brand is merely a symptom of a site that has the signals of importance and relevance behind them. Brand is not a ranking factor.

[edited by: martinibuster at 12:50 pm (utc) on Feb 24, 2015]

Nutterum

12:46 pm on Feb 24, 2015 (gmt 0)

10+ Year Member Top Contributors Of The Month



+1 to Martinibuster.

When we stop thinking about technicalities and look at the big picture trusted and secured sources will still be on top of the SERPS even if they disavow their entire backlink stack. (I hope you realize I speak more metaphorically here!) . It`s just that many people, partly even including myself get too granular in hopes to find something "that works" and say "AHA!"

rish3

12:50 pm on Feb 24, 2015 (gmt 0)

10+ Year Member Top Contributors Of The Month



Brand is merely a symptom of a site that has the signals of importance.

Perhaps, but that is just a guess on your part. Google could establish "brand" another way. The knowledge graph seems to have some idea what a brand is, for example.

It's hard to debunk or prove theories when you can't look in the box.

Edit: Also, the advent of machine learning, and bolt on filters, means Google doesn't even reliably know which signals are used, or not used. "Too noisy to be used" doesn't apply to a signal that is visible to a machine learning algo. And, the machine learning algo might get the signal unintentionally, or indirectly.

martinibuster

1:30 pm on Feb 24, 2015 (gmt 0)

WebmasterWorld Administrator 10+ Year Member Top Contributors Of The Month



Google could establish "brand" another way.


Why use the word, Brand? Why not simply say authority?

Google could establish "authority and relevance" another way.


That's really what the algorithm is concerned with identifying, authority and relevance. Do you see how your statement makes sense when the words authority and relevance replace the word, brand?

rish3

1:41 pm on Feb 24, 2015 (gmt 0)

10+ Year Member Top Contributors Of The Month



Why use the word, Brand? Why not simply say authority?

Because they aren't the same thing. I can find a [specific widget] on Amazon [a brand] but I can tell with 100% certainty that they have -0- authority and relevance for said widget.

martinibuster

2:47 pm on Feb 24, 2015 (gmt 0)

WebmasterWorld Administrator 10+ Year Member Top Contributors Of The Month



Please define brand. The word brand keeps getting mentioned as if it's a thing like PageRank. What do you mean when you mention brand?

I am a member of several forums on another topic. eBay is one of the top referenced sites. Why is that, if eBay is not the expert company in that niche?

Authority and relevance = Ranking. That's what all the algorithms have ever aspired to rank.

Brand is just a marketing goal that allows a site to sell a product or service at a premium. Brand is not a ranking factor. It's a marketing goal. To get to be a brand a company must develop all the signals that often lead to links. Think of Apple.

rish3

3:05 pm on Feb 24, 2015 (gmt 0)

10+ Year Member Top Contributors Of The Month



What do you mean when you mention brand?

Well, the important question is "what does Google mean when they mention brand?" As in "Brands are how you sort out the cesspool."

Your contention is that they mean "domain authority" + "relevance". Maybe you're right, maybe not. They could, for example, derive "brand" from data that they have access to that has nothing to do, at all, with links. I have no idea how they identify a brand. I can say, however, that something has changed...and if "domain authority" + "relevance" has always been the big driver...what changed?

Separately, even Amit Singhal has conceded, in the past, that there might be problems.
Google’s Mr Singhal calls this the problem of “brand recognition”: where companies whose standing is based on their success in one area use this to “venture out into another class of information which they may not be as rich at”.



Added Note: For clarity, should probably say "link authority" or something else instead of "domain authority". The term fits, but is easily mistaken for MOZ's DA metric...that's not what I'm referring to.

[edited by: goodroi at 11:58 pm (utc) on Feb 24, 2015]
[edit reason] Added note from rish3 [/edit]

Wilburforce

4:16 pm on Feb 24, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



Brands fall under the same area of intellectual property law as Trade Marks. The two concepts have broad similarities and may be, but are not necessarily one and the same: Jaguar, for example, is both a trade mark and a brand, while Kindle is a brand owned by Amazon (which is a trade mark).

Rights in a brand or trade mark are not absolute, but have relative weakness or strength based on a combination of factors that include authority and notoriety (as well as use, history of use, and registration), but conflating these elements is probably best avoided: we should be clear about what is happening when I search for a brand (e.g. Nike) and the #1 result (which, according to Google, is based on authority and relevance, although some might argue that nothing could have higher authority and relevance that the brand owner) is eBay, which has notoriety and is a trade mark.

When posters on WebmasterWorld say that the top positions in their sector are occupied by brands, it isn't clear what they mean. The fact that I have strong rights in my own registered trade mark does not of itself guarantee generally good results in Google SERPs, although I would expect it to be at #1 - which it is - in searches for the trade mark.

samwest

5:30 pm on Feb 25, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



If Simon says "stand on your head" will webmasters do it?
(looks like this topic veered to brands)

rish3

5:44 pm on Feb 25, 2015 (gmt 0)

10+ Year Member Top Contributors Of The Month



looks like this topic veered to brands

That isn't surprising to me, given the topic:
How would you know if links were no longer important to Google?

If you net out links as the primary factor, there's not much left other than:

  • Traffic / User Engagement
  • Determine "Brands", and prefer them

    Is there something other than those two that would be a suitable substitute for backlinks?
  • goodroi

    6:11 pm on Feb 25, 2015 (gmt 0)

    WebmasterWorld Administrator 10+ Year Member Top Contributors Of The Month



    Determine "Brands", and prefer them

    That statement without details and/or facts can be seen as a bit of conspiracy paranoia which I do not think was your intent. I think it might be more productive to discuss specific brand identification methods that Google might utilize to give preferential rankings. For example:

    -How often other web pages mention you in plain text could be an alternative to counting backlinks.
    -Filing for a registered trademark could be a signal you should get preferential ranking.
    -How often your site name or company name is entered into the Google search box could signal you are important and should rank higher.
    -Occurrence of your content being scraped. If you were the original publisher and 100 sites stole your content, then it probably is above average content worthy of higher rankings.

    rish3

    6:50 pm on Feb 25, 2015 (gmt 0)

    10+ Year Member Top Contributors Of The Month



    That statement without details and/or facts can be seen as a bit of conspiracy paranoia which I do not think was your intent.

    Sorry, I am confused. This is a speculative thread by definition, right?

    We're pondering how we would know if Google stopped caring about links.

    Since links are the very core of their foundation, that means we're speculating about what they would use instead, to figure out what's authoritative, right?

    Why would a white list be paranoid? They use them today. For example, Google News. You have to apply for, and be approved to show up there. A whitelist.

    There are many ways to build a white list. For example, the knowledge graph already knows a bit about brands [i.imgur.com].

    fathom

    11:15 pm on Feb 25, 2015 (gmt 0)

    WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



    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.


    Hmmm... I know the merits of the links and the merits of on-page SEO but what "exactly is" "stellar user engagement"

    If I landed on your website and then walked away from my laptop for the afternoon and the cat got up there and push the mouse around a bit would Google see that as "stellar user engagement"

    While I can see Google experimenting with everything I doubt they will ever drop link value as the cat could easily be your worst critic.

    Nutterum

    7:51 am on Feb 26, 2015 (gmt 0)

    10+ Year Member Top Contributors Of The Month



    Hence ( @fathom ) why we are full circle to the point made earlier - if Google is to lessen the strength of links or remove them as a direct ranking factor (emphasise the word direct!) then it needs to accumulate relevant data from via other means like Google Analytics data, where it can be easily seen that the "worst critic cat" is a single phenomena that can easily be placed as noise to ratio.

    In the end is all about relevant data and I would not be too surprised that in the near future Google will consider Google Analytics data as a ranking tool, just like they tried to shove G+ in our faces with the promise that the social signals coming from their network will score the highest.

    netmeg

    1:12 pm on Feb 26, 2015 (gmt 0)

    WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



    what "exactly is" "stellar user engagement"


    I know it when I see it.

    martinibuster

    1:26 pm on Feb 26, 2015 (gmt 0)

    WebmasterWorld Administrator 10+ Year Member Top Contributors Of The Month



    The glaring problem with the user engagement metric for determining authority and/or relevance for ranking purposes is that a site must first receive relevant traffic in order to send user engagement signals. But the site cannot obtain traffic to send user engagement signals until it first sends user engagement signals. You see how the user engagement metric does not fit and does not work for the purposes of ranking a site?

    User engagement signals are used for improving/demoting the SERP positions of sites that are already ranked. For example, if a site is relevant for a SERP but it is determined that the user has a preference for that site, has visited that site, or even has associated a search query with that site, then Google will juggle the SERP order to better serve that user. That's an example of how user engagement metrics are used to improve the SERPs.

    1. User engagement metrics cannot be used to produce the SERPs because it's impossible to generate those metrics without first ranking the site.

    2. User engagement metrics can only be used to improve the SERPs.

    guggi2000

    2:49 pm on Feb 26, 2015 (gmt 0)

    10+ Year Member Top Contributors Of The Month



    Maybe links are also bad for Google's business?

    Less links, means less referral traffic, means more time and influence on THEIR web properties.

    Any thoughts?

    EditorialGuy

    3:53 pm on Feb 26, 2015 (gmt 0)

    WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



    Maybe links are also bad for Google's business?


    On the contrary. They're essential to Google's core search business--and to the repeat Google Search visits that generate ad revenue day after day, month after month, year after year.

    guggi2000

    4:07 pm on Feb 26, 2015 (gmt 0)

    10+ Year Member Top Contributors Of The Month



    @EditorialGuy Yes, links are essential to their algorithm and thus to their core business.

    But more website links and more direct navigation between sites means less time spent on Google.com
    (in average)

    fathom

    4:15 pm on Feb 26, 2015 (gmt 0)

    WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



    Yes, links are essential to their algorithm and thus to their core business.

    But more website links and more direct navigation between sites means less time spent on Google.com
    (in average)


    That isn't a bad thing.

    Less time on a search engine that can only allow searching is GOOD!

    netmeg

    5:21 pm on Feb 26, 2015 (gmt 0)

    WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



    User engagement metrics cannot be used to produce the SERPs because it's impossible to generate those metrics without first ranking the site.


    Not sure I agree about this. What if the site ranks well in other search engines, or gets measurable social traffic, or referral traffic, or even direct traffic?

    Might be an interesting experiment. Keep a site out of Google for a time, develop all the other channels and signals (and run GA of course) and then open it to Google and see what happens.

    martinibuster

    5:54 pm on Feb 26, 2015 (gmt 0)

    WebmasterWorld Administrator 10+ Year Member Top Contributors Of The Month



    What if the site ranks well in other search engines, or gets measurable social traffic, or referral traffic, or even direct traffic?


    Good inquiries! :)


    1. What if the site ranks well in other search engines?
    That would result in the situation that Google's SERPs would rely on the SERPs of other search engines.


    2. What if the site gets measurable social traffic?
    Is it possible that social shares can be verified to be real shares and the traffic genuine and not spoofed? Google analytics data would play a larger role for this one?

    3. What if the site gets referral traffic?
    You mean from links?

    4. What if the site gets direct traffic?
    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...

    Any thoughts?
    This 77 message thread spans 3 pages: 77